MEDICAL Volume 23, Number 2

June 2021

ULTRASONOGRAPHY
A N I N T E R N A T I O N A L J O U R N A L O F C L I N I C A L I M A G I N G

„ Microvascular Doppler US in children with acute pyelonephritis

„ The value of abdominal ultrasonography compared to
colonoscopy and faecal calprotectin in following up paediatric
patients with ulcerative colitis
„ Ultrasonography techniques in the preoperative diagnosis of
parotid gland tumors – an updated review of the literature
„ The ut ility of ultrasound in the diagnostic evaluation of the
posterior ankle joint
 BioAcoustic™
Imaging
Technology

Taking ultrasound
to new heights
ACUSON Sequoia
ultrasound system

siemens-healthineers.com/ultrasound
 Contents

Editorial
To screen or not to screen for NAFLD?
I. Sporea .................................................................................................................................................................................... 133

Original papers
Transfer learning with pre-trained deep convolutional neural networks for the automatic assessment of liver
steatosis in ultrasound images
E.C. Constantinescu, A-L. Udriștoiu, Ș.C. Udriștoiu, A.V. Iacob, L.G. Gruionu, G. Gruionu, L. Săndulescu, A. Săftoiu ....... 135
Contrast-enhanced ultrasound for the assessment of focal nodular hyperplasia – results of a multicentre study
R. Şirli, I. Sporea, A. Popescu, M. Dănilă, D.L. Săndulescu, A. Săftoiu, T. Moga, Z. Spârchez, C. Cijevschi, C. Mihai,
S. Ioanițescu, D. Nedelcu, N. Iacob, G. Miclăuș, C. Brisc, R. Badea ....................................................................................... 140
Evaluation of bowel preparation before colonoscopy by ultrasonographic monitoring of colonic fecal retention:
a case series
M. Matsumoto, M. Fujioka, T. Okada, Y. Naka, A. Amemiya, E. Matsushima, N. Tamai, Y. Miura, G. Nakagami,
H. Sanada .................................................................................................................................................................................. 147
The value of abdominal ultrasonography compared to colonoscopy and faecal calprotectin in following up
paediatric patients with ulcerative colitis
I. Fodor, O. Serban, D.E. Serban, D. Farcau, O. Fufezan, C. Asavoaie, S.C. Man, D.L. Dumitrascu ..................................... 153
Microvascular Doppler ultrasound in children with acute pyelonephritis
G. Choi, B-K. Je, D. Hong, J. Cha ............................................................................................................................................ 161
Comparison of IOTA three-step strategy and logistic regression model LR2 for discriminating between benign
and malignant adnexal masses
J.J, Hidalgo, A. Llueca, I. Zolfaroli, N. Veiga, E. Ortiz, J.L. Alcázar ....................................................................................... 168
The presence of effusions between the volar plate of the proximal interphalangeal joint and the flexor digitorum
tendon is a common phenomenon: a single-center, cross sectional study
L. Xue, Y. Zhang, D. Yan, J. Fu, Z. Liu ...................................................................................................................................... 176
Assessment of testes with two-dimensional Shear Wave Elastography in patients with operated inguinal hernia
M.S. Durmaz, F. Ates, S. Arslan, T. Kara, F.G. Durmaz, M.A. Eryilmaz, K. Arslan ................................................................. 181
Congenital complete atrioventricular block from literature to clinical approach – a case series and literature
review
L. Gozar, C. Marginean, A. Fagarasan, I. Muntean, A. Cerghit-Paler, D. Miklosi, R. Toganel ............................................... 188

Reviews
Ultrasonography techniques in the preoperative diagnosis of parotid gland tumors – an updated review of
the literature
S. Stoia, G. Băciuț, M. Lenghel, R. Badea, M. Băciuț, S. Bran, D. Cristian ............................................................................. 194
The diagnostic accuracy of ultrasound in the detection of foot and ankle fractures: a systematic review and
meta-analysis
J. Wu, Y. Wang, Z. Wang ............................................................................................................................................................ 203
Understanding the role of echocardiography in patients with obstructive sleep apnea and right ventricular
subclinical myocardial dysfunction – comparison with other conditions affecting RV deformation
I.M. Chetan, B. Domokos Gergely, A. Albu, R. Tomoaia, D.A. Todea ....................................................................................... 213

Pictorial essay
First trimester fetal heart evaluation. A pictorial essay
I.C. Rotar, D. Mureșan, C. Mărginean, D.G. Iliescu, Ș. Tudorache ......................................................................................... 220
 Medical Ultrasonography
Official Journal of the Romanian Society for Ultrasonography in Medicine and Biology
Medical Ultrasonography (formerly Revista Româna de Ultrasonografie from 1999 to 2008) is the official publication of the
Romanian Society for Ultrasonography in Medicine and Biology (SRUMB). Starting with 2008 the entire content of Medical
Ultrasonography is published in English, quarterly. The journal aims to promote ultrasound diagnosis by publishing papers in a
variety of categories, including Original papers, Review Articles, Pictorial Essays, Technical Innovations, Case Report, or Letters to
the Editor (fundamental as well as methodological and educational papers). The published papers cover a wide variety of discipline
of ultrasound. The journal also host information regarding the society’s activities, the scheduling of accredited training courses in
ultrasound diagnosis, as well as the agenda of national and international scientific events.
Medical Ultrasonography is now listed in Science Citation Index Expanded/ ISI Thomson Master Journal List, Medline/
PubMed, Scopus, Pro Quest, Ebsco, and Index Copernicus data bases. Impact Factor 1.553 (JCR 2019); 5 year IF=1.598
Editorial Office
2nd Medical Clinic, 2-4 Clinicilor str., 400006 Cluj-Napoca, Romania
Tel.: +4 0264 591942/442, Fax: +4 0264 596912, Email: medultrasonography@gmail.com
Contact person: Daniela Fodor, email: dfodor@ymail.com
Journal web site: http://www.medultrason.ro
Editorial board
Editor in Chief Methodological adviser Editors Assistant Editors English language editors
Daniela Fodor Petru Adrian Mircea Radu Ion Badea Carolina Solomon Sally Wood-Lamont
Sorin Marian Dudea Bogdan Chis Ioana Robu
Oana Serban
Members
Mihaela Băciuţ (Cluj-Napoca, Romania) Richard Hoppmann (Columbia, South Carolina, USA Alina Popescu (Timişoara, Romania)
Boris Brkljacic (Zagreb, Croatia) Walter Grassi (Ancona, Italy) Alper Ozel (Istambul, Turkey)
Ciprian Brisc (Oradea, Romania) Lucas Greiner (Wuppertal, Germany) Adrian Săftoiu (Craiova, Romania)
Vito Cantisani (Rome, Italy) Norbert Gritzmann (Salzburg, Austria) Paul Singh Sidhu (London, UK)
Anca Ciurea (Cluj-Napoca, Romania) Zoltán Harkányi (Budapest, Hungary) Zeno Spârchez (Cluj-Napoca, Romania)
Sorin Crişan (Cluj-Napoca, Romania) Anamaria Iagnocco (Rome, Italy) Ioan Sporea (Timişoara, Romania)
Adrian Costache (Bucureşti, Romania) Adnan Kabaalioglu (Antalya, Turkey) Florin Stamatian (Cluj-Napoca)
Jarosław Czubak (Otwock, Poland) Daniel Lichtenstein (Paris, France) Dan Stănescu (Bucureşti, Romania)
Christoph Dietrich (Frankfurt am Main, Germany) Carmen Mihaela Mihu (Cluj-Napoca, Romania) Iwona Sudoł-Szopińska (Warsaw, Poland)
Dan Dumitraşcu (Cluj-Napoca) Dan Mihu (Cluj-Napoca, Romania) Kazmierz Szopinski (Warsaw, Poland)
Viorela Enăchescu (Craiova, Romania) Daniel Muresan (Cluj-Napoca, Romania) Adrian Şanta (Sibiu, Romania)
Otilia Fufezan (Cluj-Napoca, Romania) Luca Neri (Milan, Italy) Roxana Sirli (Timişoara, Romania)
Odd Helge Gilja (Bergen, Norway) Monica Platon Lupsor (Cluj-Napoca, Romania)
Tehnical staff Instruction for authors: Subscription information:
Iulia China Full instructions are available online at http://www.medultrason.ro/ Medical Ultrasonography is published quarterly.
authors-guidelines/ and at the end pages of the journal. ISSN (print) 1844–4172; ISSN (online) 2066–8643
The annual subscription:
For Romania – 150.00 lei for individuals For other countries (including postage fees) – 60 euro for individuals
– 350.00 for institutions – 120 euro for institutions
Bank account:
– in lei: SRUMB – Filiala Cluj-Napoca (CIF 10232679), BRD, sucursala Cluj-Napoca, Cont: RO62 BRDE 130S V959 5116 1300
– in euro: SRUMB – Filiala Cluj-Napoca (CIF 10232679), BRD, sucursala Cluj-Napoca, Cont: RO54 BRDE 130S V959 5132 1300
How to subscribe: Subscribers may pay in the bank account already mentioned. Orders should be sent online at www.medultrason.ro
“Subscription” or by email to the Editorial Office (dfodor@ymail.com)
Publishing House: “Iuliu Haţieganu” Medical Publishing House, Cluj-Napoca, Contact person: Ioana Robu,
8, Victor Babeş str, Cluj-Napoca, Tel.: + 40-264-592629 / 590832, Fax: + 40-264-598820, E-mail: irobu@umfcluj.ro
All right reserved to the Medical Publishing House of the “Iuliu Hatieganu” University of Medicine and Pharmacy, Cluj-Napoca;
no parts of this publication may be reproduced, stored in a retrieval system, or transmitted in any form or by any means,
electronic, mechanical, photocopying, recording, or otherwise without the permission of the publisher.
 Contents
(continued)

Continuing education
The utility of ultrasound in the diagnostic evaluation of the posterior ankle joint
W. Konarski, T. Poboży .............................................................................................................................................................. 226

Case report
The role of multimodal imaging in the diagnosis of an asymptomatic patient with congenital anomaly
R. Tomoaia, A. Molnar, R.Ș. Beyer, A. Dădârlat-Pop, F. Frîngu, D. Gurzau, G. Simu, I.A. Minciună, B. Caloian,
D. Zdrenghea, D. Pop ................................................................................................................................................................ 231
A rare cause of biliary obstruction – intraductal neuroendocrine tumor of the right hepatic biliary duct:
a case report
M. Danila, R. Sirli, A. Popescu, N. Iacob, A-M. Ghiuchici ....................................................................................................... 235

Letters to the Editor

Dynamic air bronchogram and lung hepatization: ultrasound for early diagnosis of pneumonia
K. Akutagawa ............................................................................................................................................................................ 238
Ultrasound imaging and guided hydro-dissection for injury of the recurrent motor branch of the median nerve
K-V. Chang, W-T. Wu, Y-C. Yang, L. Özçakar .......................................................................................................................... 239
Utilization of diagnostic ultrasound in the detection of hip fracture
Y-H. Chiu, S-G. Shyu ................................................................................................................................................................. 240
When meniscus ‘tears’ make the Baker’s cyst ‘cry’: a story on knee ultrasound
C. Pirri, C. Stecco, N. Pirri, R. De Caro, L. Özçakar .............................................................................................................. 241
Controversies in the management of bowel obstruction in pregnant woman
F. Popa, P. Bernard, E. Georgescu ............................................................................................................................................ 243
An interesting dynamic ultrasound  finding of pharyngoesophageal diverticulum: technical advice
J. Wu, X. Hu, X. Wang ............................................................................................................................................................... 244
Comment to: Effectiveness of contrast-enhanced ultrasound for detecting the staging and grading of
bladder cancer: a systematic review and meta-analysis
J. Wu, X. Hu, X. Wang ............................................................................................................................................................... 245
Authors’ response
X-Y. Ge, Z.-K. Lan, J. Chen, S-Y. Zhu ........................................................................................................................................ 246
Editorial
To screen or not to screen for NAFLD?
Ioan Sporea
Department of Gastroenterology and Hepatology, “Victor Babeş” University of Medicine and Pharmacy, Regional
Center of Research in Advanced Hepatology, Academy of Medical Science, WFUMB Center of Education, Timişoara,
Romania
Nonalcoholic Liver Disease (NAFLD) has become
the most frequent hepatological entity in the clinical
practice in the developed world. The increasing number
of overweight and obese subjects, together with the high
prevalence of type 2 diabetes mellitus (DM) and meta-
bolic syndrome, results in this condition being present in
approximately 25% of the population. A new terminol-
ogy that tends to replace the term NAFLD is MAFLD
(Metabolic Associated Fatty Liver Disease) which seems
to show better the connection between fatty liver and the
spectrum of metabolic syndrome [1,2].
For a long time, fatty liver in obese or type 2 DM
patients was considered a benign disease and not much
attention was paid to this entity. In the last years, many
papers have shown that it can progress to significant fi-
brosis or cirrhosis in a proportion of subjects.
Now the question arises to screen or not to screen the
general population for fatty liver and the methodology in
how to carry it out? A populational study performed in
Korea using Magnetic Resonance Elastography (MRE)
in a large cohort of 8183 subjects, found that 9.5% of
the patients had at least significant fibrosis (F≥2), which
included 2.6% subjects with at least advanced fibrosis
(F≥3) [3]. In this study, multivariate analysis revealed
that liver fibrosis was associated with age, male sex, type
2 DM, abnormal liver function test, HBsAg positivity
and obesity, as well as with metabolic syndrome. How-
ever, at this moment, screening of the general population
is difficult to perform since it would comprise a huge co-
hort of individuals.
On the other hand, screening the population at risk
has been proposed more and more frequently by different
Received Accepted
Med Ultrason
2021, Vol. 23, No 2, 133-134
Corresponding author: Prof. Ioan Sporea, MD, PhD
13, Snagov Street,
300482 Timisoara, Romania
E-mail: isporea@umft.ro
Med Ultrason 2021, Vol. 23, no. 2, 133-134
DOI: 10.11152/mu-3251
authors. Two main categories of patients are considered at
risk and should be screened for fatty liver and fibrosis: pa-
tients with type 2 DM and metabolic subjects. Concerning
diabetic subjects, very often they also have a metabolic
syndrome and are obese, increasing the risk of liver dis-
ease. In a Dutch study [4] performed in a cohort of 3,041
participants (mean age 66.0 years), 5.6% of participants
had significant fibrosis (liver stiffness measurements by
Transient Elastography ≥8.0 kPa). In multivariate analysis
age, alanine aminotransferase, spleen size and the com-
bined presence of diabetes mellitus and ultrasound stea-
tosis (OR, 5.20) were associated with significant fibrosis.
In a study performed by our group [5] in a cohort
of 776 type 2 DM patients, evaluated with TE and CAP
(Controlled Attenuation Parameter), we found that 60.3%
of them had severe steatosis and 19.4% had advanced fi-
brosis. In a review paper [6] results from two epidemiol-
ogy reports from United States documented steatosis in
70-74% of unselected patients with type 2 DM, advanced
liver fibrosis in 6%-15%, and previously unrecognized
cirrhosis in 3-8%.
All these facts are arguments for screening for stea-
tosis and fibrosis in such categories of patients. But how
should this screening be performed? Simple and inexpen-
sive biological tests such as FIB4 are useful for the gen-
eral practitioner for fibrosis assessment. However, what
we propose is to use ultrasound and ultrasound based
elastography for steatosis and fibrosis assessment. The
accuracy of standard ultrasound for a semi-quantitative
assessment of liver steatosis is quite good (especially for
moderate or severe steatosis), with an AUROC of 0.93
[7]. Using Transient Elastography with CAP for the as-
sessment of fibrosis and steatosis, the AUROC’s range
from 0.80 to 0.90 for the detection of significant steatosis
and fibrosis, respectively.
In the last few years advanced ultrasound systems
have been developed that can perform ultrasound based
elastography and also steatosis quantification. Some of
these new systems can also evaluate the viscoelastic
134 Ioan Sporea
properties of the liver tissue, which are considered to be
a measure of inflammation. Thus, multiparametric ultra-
sound systems are now available in the clinical practice
for the assessment of diffuse liver disease. Nowadays,
modules for steatosis and fibrosis assessment are imple-
mented by some companies, in middle range ultrasound
machines (not only in high end systems).
With the implementation of “point of care” ultra-
sound, performed by specialists in their consultation
room, the evaluation of MAFLD patients has become
more feasible and in a very short time. Regarding the
cost-utility and cost-effectiveness of using ultrasound
and ultrasound based elastography for the screening of
NAFLD patients, some studies have shown positive re-
sults [8,9].
Thus, finally, to answer the question from the title,
I believe that this is the perfect time to start screening
patients for NAFLD, in order to evidence this disease in
the early stages [10]. Using ultrasound machines and ul-
trasound based elastographic methods, we can undertake
this screening with good results and with a low cost. Not
only screening in NAFLD patients should be proposed,
but also in other categories of patients, such as alcoholic
subjects (for alcoholic liver disease: ALD). And if this
screening can be performed as a “point of care” method,
it is a better solution, both for patients and physicians.
Thus, my advice is to start screening such patients
with steatosis and fibrosis assessment using modern ul-
trasound systems, for an early diagnostic and a therapeu-
tic solution.
References
1. Eslam M, Sanyal AJ, George J; International Consensus
Panel. MAFLD: A Consensus-Driven Proposed Nomencla-
To screen or not to screen for NAFLD?
ture for Metabolic Associated Fatty Liver Disease. Gastro-
enterology 2020;158:1999-2014.e1.
2. Lin S, Huang J, Wang M, et al. Comparison of MAFLD
and NAFLD diagnostic criteria in real world. Liver Int
2020;40:2082-2089.
3. Nah EH, Cho S, Kim S, Chu J, Kwon E, Cho HI. Preva-
lence of liver fibrosis and associated risk factors in the
Korean general population: a retrospective cross-sectional
study. BMJ Open 2021;11:e046529.
4. Koehler EM, Plompen EP, Schouten JN, et al. Presence of
diabetes mellitus and steatosis is associated with liver stiff-
ness in a general population: The Rotterdam study. Hepa-
tology 2016;63:138–147.
5. Sporea I, Mare R, Popescu A, et al. Screening for Liver
Fibrosis and Steatosis in a Large Cohort of Patients with
Type 2 Diabetes Using Vibration Controlled Transient Elas-
tography and Controlled Attenuation Parameter in a Single-
Center Real-Life Experience. J Clin Med 2020;9:1032-
1045.
6. Zoler ML. Widespread Liver Disease Missed in Patients
With Type 2 Diabetes. Medsearch 2021 Jan 25. Available at:
https://www.medsearchuk.com/widespread-liver-disease-
missed-in-patients-with-type-2-diabetes-widespread-liver-
disease-missed-in-patients-with-type-2-diabetes/
7. Hernaez R, Lazo M, Bonekamp S, et al. Diagnostic Accura-
cy and Reliability of Ultrasonography for the Detection of
Fatty Liver: A Meta-Analysis. Hepatology 2011;54:1082–
1090.
8. Zhang E, Wartelle-Bladou C, Lepanto L, Lachaine J, Clout-
ier G, Tang A. Cost-utility analysis of nonalcoholic steato-
hepatitis screening. Eur Radiol 2015;25:3282-3294.
9. Tapper EB, Sengupta N, Nunink MG, Afdhal NH, Lai
M. Cost-Effective Evaluation of Nonalcoholic Fatty Liv-
er Disease With NAFLD Fibrosis Score and Vibration
Controlled Transient Elastography. Am J Gastroenterol
2015;110:1298-1304.
10. Sporea I, Șirli R. Nonalcoholic Fatty Liver Disease and the
Need for Action. J Gastrointestin Liver Dis 2020;29:139-141.
Original papers Med Ultrason 2021, Vol. 23, no. 2, 135-139
DOI: 10.11152/mu-2746

Transfer learning with pre-trained deep convolutional neural

networks for the automatic assessment of liver steatosis in ultrasound
images
Elena Codruța Constantinescu1, Anca-Loredana Udriștoiu2, Ștefan Cristinel Udriștoiu2,
Andreea Valentina Iacob2, Lucian Gheorghe Gruionu3, Gabriel Gruionu3, Larisa Săndulescu1,
Adrian Săftoiu1

1Research Center of Gastroenterology and Hepatology, University of Medicine and Pharmacy Craiova, 2Faculty
of Automation, Computers and Electronics, University of Craiova, Craiova, 3Faculty of Mechanics, University of
Craiova, Craiova, Romania

Abstract
Aim: In this paper we proposed different architectures of convolutional neural network (CNN) to classify fatty liver
disease in images using only pixels and diagnosis labels as input. We trained and validated our models using a dataset of 629
images consisting of 2 types of liver images, normal and liver steatosis. Material and methods: We assessed two pre-trained
models of convolutional neural networks, Inception-v3 and VGG-16 using fine-tuning. Both models were pre-trained on
ImageNet dataset to extract features from B-mode ultrasound liver images. The results obtained through these methods were
compared for selecting the predictive model with the best performance metrics. We trained the two models using a dataset of
262 images of liver steatosis and 234 images of normal liver. We assessed the models using a dataset of 70 liver steatosis im-
ages and 63 normal liver images. Results. The proposed model that used Inception v3 obtained a 93.23% test accuracy with a
sensitivity of 89.9%% and a precision of 96.6%, and areas under each receiver operating characteristic curves (ROC AUC) of
0.93. The other proposed model that used VGG-16, obtained a 90.77% test accuracy with a sensitivity of 88.9% and a precision
of 92.85%, and areas under each receiver operating characteristic curves (ROC AUC) of 0.91. Conclusion. The deep learning
algorithms that we proposed to detect steatosis and classify the images in normal and fatty liver images, yields an excellent
test performance of over 90%. However, future larger studies are required in order to establish how these algorithms can be
implemented in a clinical setting.
Keywords: fatty liver disease; NAFLD; convolutional neural networks; deep learning

Introduction
With the rapid growth of the population and the in-
creasing rates of obesity and diabetes, fatty liver dis-
Received 22.08.2020  Accepted 01.12.2020
Med Ultrason
2021, Vol. 23, No 2, 135-139
Corresponding author: Adrian Săftoiu MD PhD MSc
Professor of Diagnostic and Therapeutic
Techniques in Gastroenterology
University of Medicine and Pharmacy
Craiova, Romania
Phone: +40 744 823355
Fax: +40 251 310287
E-mail: a drian.saftoiu@umfcv.ro
adriansaftoiu@gmail.com
ease (FLD) has become the most common liver disease
in Western countries, non-alcoholic fatty liver disease
(NAFLD) affecting around 25% of the global adult pop-
ulation [1]. Liver steatosis is defined as the accumula-
tion of more than 5% of fat in the hepatocytes [2]. In
the natural evolution of NAFLD, if no early strategies
are adopted, the disease can progress to steatohepatitis
(NASH) which is a risk factor for liver fibrosis, cirrhosis
and even hepatocellular carcinoma [3]. In order to con-
firm the diagnosis, to quantify liver steatosis and stage
NAFLD, liver biopsy is still considered the gold stand-
ard. However, it is an invasive procedure which carries a
high risk of serious complications and, therefore, it can-
not be routinely used for the evaluation and follow-up of
NAFLD [4].
136 Elena Codruța Constantinescu et al Transfer learning with pre-trained deep convolutional neural networks
Non-invasive imaging methods used for the diagno-
sis of FLD include ultrasonography, computed tomogra-
phy, magnetic resonance and magnetic resonance-based
fat quantification techniques. Conventional B-mode ul-
trasonography is mostly preferred for screening and for
the initial assessment of FLD due to the wide availabil-
ity and low costs [5]. Although a normal ultrasound im-
age cannot rule out liver steatosis, the overall sensitivity
and specificity of ultrasound are comparable to those of
histology, increasing with the degree of steatosis [5-7].
However, FLD can be diagnosed on ultrasound based
on some suggestive parameters: diffuse hyperechoic
structure (“bright liver”), deep beam attenuation, liver-
to-kidney contrast, bright vessel walls [8,9]. Taking into
consideration the public health burden given by NAFLD,
there is a great need for the development of a non-inva-
sive approach for the correct diagnosis and to determine
the progression of the liver disease in order to establish
an early treatment.
Over time, in order to improve the diagnostic accu-
racy of FLD and to make a quantitative assessment of
steatosis, several computer-aided methods were inves-
tigated, including hepato-renal index (HRI), gray level
co-occurrence matrix (GLCM), artificial intelligence
(AI) methods such as traditional machine learning (ML)
algorithms and deep learning (DL) algorithms. One of
the most popular types of DL algorithms and widely used
in medical imaging are convolutional neural networks
(CNN). CNN consist of a series of layers (convolutional,
pooling and fully connected) that can perform end-to-end
supervised learning in order to identify patterns, lines or
edges while some of the layers that are ”hidden” can
conduct unsupervised learning tasks. However, CNN ar-
chitecture consists of many convolutional layers, which
means that a great deal of data is needed for the train-
ing set in order to acquire competent accuracy [10]. To
overcome this issue, data augmentation techniques have
been proposed to reduce overfitting. In order to increase
the efficiency of CNN, transfer learning has become very
popular because it considerably reduces the training time
and requires less data to train, while increasing the classi-
fication performance. Up to date, the research in clinical
application of AI in liver imaging is small and limited,
and more studies are required to demonstrate the effi-
ciency of DL algorithms [11] but the importance of CNN
in the detection of liver masses has been demonstrated in
previous studies [12-14]. Moreover, several small studies
used DL with B-mode ultrasound images for the diagno-
sis and classification of FLD [15,16].
The aim of our study was to analyse the test perfor-
mance of recent DL algorithms for the ultrasound evalu-
ation of liver steatosis.
Materials and methods
Pre-trained convolutional neural network
implementation and parameter details
Very deep networks are prone to overfit and large
convolution operations are computationally expensive.
The Inception network, on the other hand, is complex
and uses many improvements to obtain better perfor-
mance, both in terms of speed and accuracy. VGG16,
another CNN architecture which is considered to be an
excellent convolution neural model, is a large network,
comprising approximately 138 million parameters [18].
In this study, we used fine-tuning which is a type
of transfer learning. We applied fine-tuning to the pre-
trained DL models Inception v3 and VGG-16. The top
layers of the pre-trained models are replaced by new fully
connected layers with random parameters, while the lay-
ers below are kept frozen. The new fully connected layers
can learn patterns from previous learned convolutional
layers, because a very small learning rate is utilised. By
applying fine-tuning, liver steatosis could be recognised
even if our networks, Inception v3 and VGG-16 were not
originally trained. This method achieved higher accuracy
than feature extraction [17,18]. After rescaling the im-
ages and using image augmentation, we flowed them in
batches of 16. In order to reduce overfitting and improve
the model’s ability to generalize, we used dropout, activ-
ity regularization and kernel regularization (see: https://
github.com/keras-team/keras).
The generic architecture of the proposed methods
based on transfer learning with fine-tuning can be ob-
served in figure 1.
Inception v3 model
We used Inception v3 model pre-trained on ImageNet
dataset to extract high-level features from B-mode ultra-
sound liver images [17]. Our proposed model included
all the layers in the Inception v3 network, except for the
last fully connected layer, making all the layers non-
trainable in order to reduce the overfitting. At the top of
Fig 1. The generic representation of the transfer learning meth-
od used to detect liver steatosis

the network we added two fully connected layers, one
with 1024 hidden units and ReLU activation and the oth-
er was the last sigmoid dense layer for classification. Our
optimizer was RMSprop with a learning rate of 0.0001
(see: https://github.com/keras-team/keras).
VGG-16 model
Another model that we used was the VGG-16 net-
work pre-trained on ImageNet [18]. The model we used
included all the layers in the network, except for the last
fully connected layer. At the top of the network we added
three fully connected layers with 512, 128, and 64 hidden
units and ReLU activation. The last layer is the sigmoid
dense layer used for classification. Our optimizer was
Adam with a learning rate of 0.0001 [19].
Patient selection and image dataset
The dataset contains images from 60 patients, 30 pa-
tients with hepatic steatosis and 30 healthy subjects, both
male and female, age ranging from 18 to 92. The partici-
pants were selected by an expert hepatologist from the
University of Medicine and Pharmacy, Research Center
of Gastroenterology and Hepatology, Craiova from the
outpatient clinic of a private healthcare network. Exclu-
sion criteria included excessive alcohol consumption or
>20 g ethanol per day, history or clinical and/or laborato-
ry evidence of liver disease, hepatotoxic medication use.
Written informed consent was obtained from all patients.
The study is non-interventional and all procedures are
used in the current daily practice.
The diagnosis of NAFLD was based on the imag-
ing studies (B-mode ultrasound and ARFI elastography)
along with the clinical exclusion of the patients who con-
sume > 20 g ethanol per day. The B-mode ultrasound
images were collected by the same physician and all
examinations were performed on Siemens Acuson NX3
ultrasound machine. The ultrasound diagnosis of steato-
sis was indicated by diffuse hyperechoic structure, deep
beam attenuation and increased liver echogenicity in
relation to the right kidney. Participants were evaluated
with pSWE, an alternative method for the non-invasive
evaluation of liver fibrosis that uses acoustic radiation
force impulse imaging (ARFI) to induce tissue disloca-
tion. The examinations were performed by another phy-
sician on a Siemens Acuson S2000 machine. All the par-
ticipants had ten valid measurements performed in the
right liver lobe with IQR/M ≤30% and a median value
ranging from to 1.8 to 3.7 kPa.
The database was created by cropping non-overlap-
ping patches (fig 2) and by specific transformation such
as resizing and shifting, obtaining 629 grayscale liver im-
ages with a resolution of 75x75. Out of the 629 images,
496 were used for training and 133 for testing. For the
training dataset, we used 262 images with liver steatosis
Med Ultrason 2021; 23(2): 135-139 137
Fig 2. The selection of non-overlapping patches from images
and 234 normal liver images. For the test dataset we used
70 liver steatosis images and 63 normal liver images.
Statistical analysis
The medical diagnosis metrics used to analyse the
performance of our CNN models were: sensitivity, preci-
sion, test accuracy and area under curve (AUC).
The other metrics used to evaluate the classifier qual-
ity were the receiver operating characteristic (ROC) met-
ric and Precision-Recall (PR). While precision measures
the relevancy of the result, recall measures how many
relevant results are returned.
Results
We used two pre-trained CNN models to detect the
characteristic features in each liver image. Our results
show a final diagnosis accuracy of 93.23% and the AUC
of 0.93 for the Inception v3 model. By comparison, when
using the VGG-16 model, the final diagnosis accuracy
was 90.77% and the area under curve of 0.91. All metrics
computed for Inception v3 vs. VGG-16 are summarized
in Table I and showed the improvements achieved by us-
ing the Inception v3 model.
The ROC curves for the two models (Inception v3 vs.
VGG-16) show the rate of false positive is near to zero
while the rate of true positive is between 0.9 and 1 (fig
3a). The precision-recall curves of the two models (In-
ception v3 vs. VGG-16) demonstrate a better precision
and recall of the Inception v3 model (fig 3b). High scores
Table I. Evaluation metrics for the classification of non-alco-
holic fatty liver disease using the two pre-trained CNN models.
Accuracy AUC Sensitivity Precision
Inception v3 93.23% 0.93 88.9% 96.6%
VGG-16 90.77% 0.91 88.9% 91.8%
138 Elena Codruța Constantinescu et al Transfer learning with pre-trained deep convolutional neural networks
Fig 3. Comparison between Inception v3 vs. VGG-16: the
ROC curves (a) and the Precision/Recall curves (b).
for both precision and recall, indicate that the classifier
was returning accurate results, along with returning a ma-
jority of all positive results, respectively.
Discussion
By analysing the results, our study demonstrated that
the use of DL methods with transfer learning and fine
tuning had significant effects on the detection of liver
steatosis by classifying images into normal and fatty liver
cases with high accuracy and precision.
In the last 2 years, several studies addressed the po-
tential of DL for the detection and evaluation of focal
liver lesions, the detection of fibrosis associated with vi-
ral hepatitis, the detection of NAFLD, to assess chronic
liver disease and to facilitate and predict treatment re-
sponse [20-23].
Yasaka et al [12] proposed a method based on CNN
for the differentiation of liver masses in CT, obtaining
the median accuracy of 84% for test data and the AUC
of 0.92. Ben-Cohen et al proposed a liver metastases
detection method in CT [13]. This study reported a true
positive rate of 94.6% with a 2.9 false positive result per
case. Trivizakis et al developed methods based on 3-D
CNN and 2-D CNN for cancer classification with appli-
cation to MRI liver tumor [14]. The classification per-
formance was 83% for 3-D CNN vs 69.6% and 65.2%
for 2-D CNNs. Moreover, a study [15] classified ultra-
sound liver images of patients who were to undergo bari-
atric surgery using HI, GLCM and CNN algorithm, with
liver biopsy as reference. The results showed that the
Inception-Resnet-v2 pre-trained model was efficient and
operator-independent, with higher accuracy (96.3%) than
HI and GLCM. Recently, Santhosh and Rajalakshmi de-
veloped a DL method validating the usefulness of CNNs
for accurate classification of fatty liver using ultrasound
images. They obtained a 91.37% accuracy for detecting
liver steatosis [16].
Also, our results support the robustness of CNNs with
transfer learning to detect liver steatosis with high ac-
curacy, precision and sensitivity even though we used a
small image dataset from 60 patients. We compared the
results obtained by the two powerful networks (Inception
v3 and VGG-16), both with high accuracy of 93.23% and
90.77% respectively. As can be observed by the obtained
results, the classification metrics of Inception v3 showed
a slightly improvement, with a higher precision in clas-
sifying the two types of images.
Considering the high classification accuracy ob-
tained, future potential advantages of an automatic diag-
nosis of FLD using AI algorithms could include (1) an
accurate and rapid diagnosis for an initial assessment of a
patient with risk factors, thus, making the screening and/
or monitoring efficient; (2) could provide assistance for
remote or isolated areas in the absence of any other diag-
nostic imaging resource as well as assisting sonographers
in decision making.
On the other hand, the proposed methods have several
limitations. Firstly, the dataset size is relatively small and
in order to improve the accuracy of DL network and to
validate the generality of the proposed methods, a larger
sample size is needed. This was also confirmed in other
studies that compared DL to other algorithms. Secondly,
the performance of our diagnosis methods was not com-
pared with experts. This will be part of a future study.
Third, to have more accurate data, US diagnostics should
be supported by liver biopsy as the gold standard, even
though studies show that the specificity and sensitivity
of ultrasound are similar to histology, 93.6% and 84.8%
respectively [4]. Moreover, the values we obtained us-
ing the pre-trained CNN were calculated as means of
images that were manually selected by a physician, and
this could have had a significant influence on the results.

A possible direction for future studies would be to opti-
mize the selection process for the region of interest from
B-mode ultrasound images.
In conclusion, taking into consideration the global
health burden of FLD, in particular NAFLD, there is a
pressing need to have an accurate, fast and operator-in-
dependent ultrasound diagnosis. The DL algorithms that
we proposed to detect steatosis and classify the images
in normal and fatty liver images, yields an excellent test
performance of over 90%. However, future larger studies
are required in order to establish how these algorithms
can be implemented in a clinical setting.
Acknowledgement: The research leading to these re-
sults received funding from the Competitiveness Opera-
tional Program 2014-2020 under the project P_37_357
“Improving the research and development capacity for
imaging and advanced technology for minimal invasive
medical procedures (iMTECH)” grant, Contract No.
65/08.09.2016, SMIS-Code: 103633. This work was also
partially supported by the grant POCU 380/6/13/123990,
co-financed by the European Social Fund within the
Sectorial Operational Program Human Capital 2014-
2020.
Conflict of interest: none
References
1. Younossi Z, Tacke F, Arrese M, et al. Global perspectives
on nonalcoholic fatty liver disease and nonalcoholic steato-
hepatitis. Hepatology 2019;69:2672-2682.
2. Brunt EM. Nonalcoholic steatohepatitis: definition and pa-
thology. Semin Liver Dis 2001;21:3–16.
3. Kikuchi L, Oliveira CP, Carrilho FJ. Nonalcoholic fatty
liver disease and hepatocellular carcinoma. Biomed Res Int
2014;2014:106247.
4. Sumida Y, Nakajima A, Itoh Y. Limitations of liver biopsy
and non-invasive diagnostic tests for the diagnosis of non-
alcoholic fatty liver disease/non-alcoholic steatohepatitis.
World J Gastroenterol 2014;20:475-485.
5. Hernaez R, Lazo M, Bonekamp S, et al. Diagnostic ac-
curacy and reliability of ultrasonography for the detection
of fatty liver: a meta-analysis. Hepatology 2011;54:1082–
1090.
6. Bohte AE, van Werven JR, Bipat S, Stoker J. The diagnostic
accuracy of US, CT, MRI and 1H-MRS for the evaluation
of hepatic steatosis compared with liver biopsy: a meta-
analysis. Eur Radiol 2011;21:87–97.
7. Wieckowska A, Feldstein AE.. Diagnosis of nonalcoholic
fatty liver disease: invasive versus noninvasive. Semin Liv-
er Dis 2008;28:386–395.
8. Badea R. Ficatul. In: Badea R, Dudea SM, Mircea PA,
Stamataian F. (Eds.). Tratat de ultrasonografie clinica.
Editura Medicala, Bucuresti, 2000:105-175.
Med Ultrason 2021; 23(2): 135-139 139
9. Dasarathy S, Dasarathy J, Khiyami A, Joseph R, Lopez R,
McCullough AJ. Validity of real time ultrasound in the di-
agnosis of hepatic steatosis: a prospective study. J Hepatol
2009;51:1061-1067.
10. Kim M, Yun J, Cho Y, et al. Deep learning in medical imag-
ing. Neurospine 2019;16:657-668.
11. Azer SA. Deep learning with convolutional neural net-
works for identification of liver masses and hepatocellular
carcinoma: A systematic review. World J Gastrointest On-
col 2019;11:1218-1230.
12. Yasaka K, Akai H, Abe O, Kiryu S. Deep Learning with
convolutional neural network for differentiation of liver
masses at dynamic contrast-enhanced CT: A Preliminary
Study. Radiology 2018;286:887-896.
13. Ben-Cohen A, Klang E, Kerpel A, Konen E, Amitai MM,
Greenspan H. Fully convolutional network and sparsity-
based dictionary learning for liver lesion detection in CT
examinations. Neurocomputing 2018;275:1585-1594.
14. Trivizakis E, Manikis GC, Nikiforaki K, et al. Extending
2-D convolutional neural networks to 3-D for advancing
deep learning cancer classification with application to MRI
liver tumor differentiation. IEEE J Biomed Health Inform
2019;23:923-930.
15. Byra M, Styczynski G, Szmigielski C, et al. Transfer learn-
ing with deep convolutional neural network for liver stea-
tosis assessment in ultrasound images. Int J Comput Assist
Radiol Surg 2018;13:1895-1903.
16. Reddy DS, Rajalakshmi P. A Novel Web Application
Framework for Ubiquitous Classification of Fatty Liver
Using Ultrasound Images.  IEEE 5th World Forum on In-
ternet of Things (WF-IoT). Limerick, Ireland 2019:502-
506.
17. Szegedy C, Vanhoucke V, Ioffe S, Shlens J, Wojna Z.
Rethinking the Inception Architecture for Computer Vi-
sion.  IEEE Conference on Computer Vision and Pattern
Recognition (CVPR), Las Vegas, NV 2016:2818-2826.
18. Simonyan K, Zisserman A. Very deep convolutional net-
works for large-scale image recognition, 2015. Available at:
https://arxiv.org/abs/1409.1556. Accessed 01.10.2019.
19. Kingma DP, Ba JL. Adam: A Method for Stochastic Optimi-
zation 2014. Available at: https://arxiv.org/pdf/1412.6980.
pdf. Accessed 01.09.2019.
20. Hassan TM, Elmogy M, Sallam ES. Diagnosis of focal liver
diseases based on deep learning technique for ultrasound
images. Arab J Sci Eng 2017;42:3127-3140.
21. Biswas M, Kuppili V, Edla DR, et al. Symtosis: A liver
ultrasound tissue characterization and risk stratification in
optimized deep learning paradigm. Comput Methods Pro-
grams Biomed 2018;155:165–177.
22. Gatos I, Tsantis S, Spiliopoulos S, et al. A Machine-learning
algorithm toward color analysis for chronic liver disease
classification, employing ultrasound shear wave elastogra-
phy. Ultrasound Med Biol 2017;43:1797–1810.
23. Ibragimov B, Toesca D, Chang D, Yuan Y, Koong A, Xing
L. Development of deep neural network for individualized
hepatobiliary toxicity prediction after liver SBRT.  Med
Phys 2018;45:4763–4774.
Original papers Med Ultrason 2021, Vol. 23, no. 2, 140-146
DOI: 10.11152/mu-2912

Contrast-enhanced ultrasound for the assessment of focal nodular

hyperplasia – results of a multicentre study
Roxana Şirli1, Ioan Sporea1, Alina Popescu1, Mirela Dănilă1, Daniela Larisa Săndulescu2,
Adrian Săftoiu2, Tudor Moga1, Zeno Spârchez3, Cristina Cijevschi4, Cătălina Mihai4,
Simona Ioanițescu5, Dana Nedelcu6, Nicoleta Iacob7, Grațian Miclăuș7, Ciprian Brisc8,
Radu Badea3

1Department of Gastroenterology and Hepatology, “Victor Babeş” University of Medicine and Pharmacy Timişoara,
2Research Center of Gastroenterology and Hepatology, University of Medicine and Pharmacy Craiova, 3Regional
Institute of Gastroenterology and Hepatology “Prof. Dr. Octavian Fodor”, ”Iuliu Hațieganu” University of Medicine
and Pharmacy Cluj Napoca, 4Department of Gastroenterology, “Gr. T. Popa” University of Medicine and Pharmacy
Iaşi, 5Center of Internal Medicine, Fundeni Clinical Institute, “Carol Davila” University of Medicine and Pharmacy,
Bucharest, 6Ponderas and Neolife Hospitals, Bucharest, 7Department of Anatomy and Embryology, “Victor Babeş”
University of Medicine and Pharmacy, Timişoara, 8Department of Gastroenterology, University of Oradea, Romania

Abstract
Aim: Contrast-enhanced ultrasound (CEUS) has become a relevant imaging method for the evaluation of focal liver le-
sions (FLL). The aim of this study was to evaluate the performance of CEUS for the assessment of focal nodular hyperplasia
(FNH) in a large study group. Material and methods: We performed a multicentre prospective observational study, which
included successive CEUS examinations from fourteen Romanian centres. CEUS examinations were performed in de novo
FLL, using low mechanical index ultrasound, following an intravenous bolus of 2.4 ml SonoVue. CEUS was considered
conclusive for FNH if a typical pattern was present following contrast (rapid “spoke-wheel” enhancement during the arterial
phase, hyperenhanced lesion during venous phase, hyper- or isoenhanced in the late phase). In all cases a reference method was
available (contrast enhanced CT or MRI or biopsy). The trial was registered in clinicaltrials.gov (Identifier NCT01329458).
Results: During the 6 years study, 2062 “de novo” FLL were evaluated by CEUS. From this cohort, 94/2062 (4.5%) had a
typical enhancing pattern for FNH as described in the EFSUMB guidelines. Contrast enhanced CT/MRI and biopsy diagnosed
additional 15 FNH. From the 94 cases diagnosed as FNH by CEUS, in nine the final diagnosis was different (five of them
adenomas). CEUS had 85% sensitivity, 99.5% specificity, 90.4% positive predictive value, 99.2% negative predictive value
and 98.8% diagnostic accuracy for the diagnosis of FNH. Conclusions: CEUS is a sensitive and very specific method for the
diagnosis of FNH.
Keywords: Contrast Enhanced Ultrasound; focal liver lesions; focal nodular hyperplasia; multicentre study

Introduction ed prevalence, more frequent in women (female: male

prevalence 4:1) [1]. It is not really a tumour but a re-
Focal nodular hyperplasia (FNH) is the second generative nodule as a reaction to a vascular congeni-
most frequent benign liver tumour with a 3% estimat- tal abnormality [2,3]. Most frequently it is discovered
by chance by conventional ultrasound in asymptomatic
Received 01.11.2020  Accepted 25.02.2021
Med Ultrason
women younger than 50, in 80-95% of cases as a solitary
2021, Vol. 23, No 2, 140-146 lesion, but in some cases multiple FNH can be found [4].
Corresponding author: Ciprian Brisc, Assoc Prof Considering their higher prevalence in women and the
Department of Gastroenterology, fact that they enlarge in women who use contraceptive
University of Oradea
65 Gheorghe Doja str, Oradea, România
drugs, FNH seems to be influenced by the hormone bal-
E-mail: brisciprian@gmail.com ance, even if hormones are not involved in their patho­-
Phone: +40-722591956 genesis.

From a pathological point of view, FNH include nor-
mal liver structures but with abnormal architecture. Typi-
cal for FNH is a fibrous scar (located in the centre, more
rarely in the periphery), which includes an artery larger
than usual, originating outside the nodule. This vascular
disorder generates a regenerative nodule, which lacks the
central terminal hepatic vein and has only capillarized
sinusoids derived from the feeding artery [2]. This type
of vascularization generates the typical aspect of FNH in
contrast imaging.
On contrast enhanced ultrasound (CEUS), FNHs
are hyper-vascular, appearing homogeneously hyper-
enhancing in the arterial phase, with very rapid and cen-
trifugal fill-in. This “spoke-wheel” pattern is essential for
differentiating FNHs from adenomas, from hyper-vascu-
lar malignant lesions, or from high-flow haemangiomas
[5,6]. An eccentric feeding vessel may be seen in approx-
imately 30% of cases. Usually, FNHs are hyperenhanc-
ing in the portal phase and hyper or isoenhancing in the
late phase [5,6]. The central scar can appear as hypoen-
hancing in the late phase [6,7]. At contrast enhanced-CT,
FNH is hyperdense with an isodense aspect during the
portal-venous phase. A hyperdense central scar is com-
monly seen in the late phase [8]. On unenhanced MRI,
FNHs appear as isointense on T1-weighted images and
iso- to slightly hyperintense on T2-weighted images [9].
The MRI aspect of FNH following liver-specific contrast
is of a hyperintense lesion in the initial phase with an
isointense aspect in the delayed phase, the central scar
being enhanced in the late phases [10].
FNHs are usually asymptomatic, being discovered
during routine conventional US examination. Finding a
focal liver lesion (FLL) can be highly stressful for the pa-
tient and thus all available methods should be performed
to elucidate its nature. FNH’s B-mode aspect is uncharac-
teristic; it can be isoechoic, slightly hyper- or hypoecho-
ic, with the central scar visible in 70-80% of cases [11].
Baseline colour Doppler ultrasound can reveal the central
vessel, originating from the centre to the periphery with a
low resistance index and a high flow pattern [7].
In the latter years, CEUS has become a reliable imag-
ing method for the evaluation of focal liver lesions. Inter-
national and national guidelines define precise criteria for
the diagnosis of FNH [6,12]. The aim of this study was to
evaluate the performance of CEUS for the assessment of
focal nodular hyperplasia in a large, prospective multicen-
tre study, to collect useful information for the practitioner.
Material and methods
We performed a multicentre prospective observation-
al study, during a 6-year period (April 2011-June 2017),
Med Ultrason 2021; 23(2): 140-146 141
which included successive CEUS examinations from
eight university centres (14 individual departments).
The trial was registered in clinicaltrials.gov (Identifier
NCT01329458).
Patients
We included consecutive patients (older than 18
years), with newly discovered FLL during B-mode irre-
spective of the FLL size. In all patients, B-mode ultra-
sound was performed followed by CEUS. Also, in all pa-
tients a second line imaging method (contrast enhanced
CT or MRI) or a biopsy were available, considered as
the reference method. Most MRI imaging studies were
performed using gadolinium based contrast media. Only
in inconclusive cases hepato-specific contrast media was
used. Demographic data, history of chronic hepatitis or
of previous cancers, as well as data regarding the indica-
tion for CEUS were recorded.
The exclusion criteria were: contraindications for
CEUS (acute myocardial infarction, class III/IV heart
failure, rhythm disorders, pregnant women); patients
who refused to participate; easy to diagnose FLL by con-
ventional US, such as simple cysts or hydatid cysts, mul-
tiple lesions suggestive for liver metastases, patients with
known FLL (hepatocellular carcinoma following percu-
taneous treatment, follow-up of patients with known me-
tastasis or other FLLs, etc.); patients in whom biopsy or
second line sectional imaging methods were unavailable.
A dedicated website was developed (http://study.um-
fcv.ro) for this study. Data were registered online for each
individual patient.
Informed consent for CEUS was obtained from every
patient. The study protocol was approved by the local
Ethical Committee of each centre and was in accordance
with the Helsinki Declaration of 1975.
Conventional and contrast-enhanced studies
In each patient, B-mode US and CEUS were per-
formed with the same high-end ultrasound machine (dif-
ferent for each centre) able to perform low-mechanical
index examinations. We documented the type of ultra-
sound machine, the name of the operator and the amount
of contrast agent used. CEUS was performed by experts
from each centre (levels II and III according to the EF-
SUMB classification), who were blinded to the CT/MRI
or histology results.
Convex probes using a low mechanic index (0.09-
0.11) were used. The US scan parameters (ie. focal zone,
time gain compensation) were not changed during the
CEUS study. One focus was placed below the lesion. We
documented for each FLL the number, size, placement
and US aspect on conventional B-mode ultrasound.
The contrast agent used was SonoVue® (Bracco
SpA, Milan, Italy), a perfluoro gas containing agent, as
142 Roxana Şirli et al CEUS for the assessment of focal nodular hyperplasia – results of a multicentre study

Fig 1. a) Slightly hypoechoic FLL in the right liver lobe. b) The FLL shows “spoke-wheel” pattern on Doppler US. c) CEUS, arterial
phase – the lesion is completely hyperenhanced 11 seconds after bolus injection – excentric feeding artery also visible. d) CEUS,
portal phase – the lesion is slightly hyperenhancing with visible central scar. e) CEUS, late phase – the lesion is iso/hyperenhancing
with well visible central scar (arrow).

per standard protocol [6,12]. Lesions’ enhancement pat-
terns were studied in 3 phases: arterial (10-30 seconds
following contrast bolus), portal (30-120 seconds) and
late phase (>120 seconds) [6,12]. In each patient, the
contrast study duration was at least 5 minutes after bo-
lus injection. To document the study, four video files no
longer than 30 seconds each were captured, containing
conventional B-mode examination, the arterial phase, the
portal phase and the late phase.
CEUS vascular patterns were defined by compar-
ing the FLL’s enhancement pattern to the surrounding
liver parenchyma. We classified CEUS vascular pat-
terns as: homogeneous hyperenhancement (the whole
FLL showed homogeneous enhancement); heterogene-
ous hyperenhancement (mixed irregular areas of contrast
enhancement were present); rim-like hyperenhancement
(a peripheral hyperenhancement limited to less than 25%
of the tumour); iso-enhancement (the FLL enhanced the
same as the adjacent parenchyma from the same depth);
hypoenhancement (the lesion enhanced less than the
adjacent parenchyma at the same depth); wash-out (hy-
poenhancement in the portal or late phases preceded by
hyper or iso-enhancement in the arterial phase).
A CEUS diagnosis of FNH was established after the
contrast study based on the patterns described in guide-
lines [6,12]: rapid “spoke-wheel” enhancement during
the arterial phase, hyperenhanced lesion during the ve-
nous phase, hyper or isoenhanced lesion in the late phase,
sometimes with the visualization of a central hypoechoic
scar (fig 1). A CEUS examination was considered con-
clusive if the FLL had a typical enhancement pattern ac-
cording to national and international guidelines [6,12]
and inconclusive if not. The CEUS diagnosis was com-
pared with the final diagnosis established based on all
available imaging and clinical data: contrast enhanced
CT, and/or MRI, and/or histology.
Statistical analysis
Statistical analysis was performed using the Graph-
Pad Prism program, version 7.02 (GraphPad Software,
La Jolla, USA). We assessed the accuracy of CEUS for
FNH characterization. The sensitivity (Se) was calcu-
lated as true positive cases divided by the total number
of cases in which the disease was present; the specificity
(Sp) was calculated as true negative cases divided by the
total number of cases in which the disease was absent;
the positive predictive value (PPV) was calculated as true
positive cases divided by all CEUS positive cases; the
negative predictive value (NPV) was calculated as true
negative cases divided by all CEUS negative cases and
accuracy was calculated as the sum of true positive and
true negative cases divided by the total number of cases.
We included in the statistical analysis all cases reported;
we considered the inconclusive CEUS cases as wrongly
diagnosed.
 Med Ultrason 2021; 23(2): 140-146 143
Results Table I. Patients’ characteristics
Number of patients 2062
During the 6 years study, 2062 “de novo” FLL were Mean age (years) 52.4±7.5
evaluated by CEUS. Patients’ characteristics are present- Gender, women 914 (44.3)
ed in Table I.
Underlying chronic liver disease 727 (35.3%)
From this cohort, 94/2062 (4.5%) had a typical en-
CEUS diagnosis
hancing pattern for FNH. Contrast enhanced CT/MRI Hepatocellular carcinomas 685 (33.2)
and biopsy diagnosed additional 15 FNH (12 in cases Hemangiomas 452 (21.9)
labelled as inconclusive on CEUS and one each labelled Metastases 418 (20.3)
by CEUS as haemangioma, adenoma and metastasis). Focal nodular hyperplasias 94 (4.5)
From the 94 cases diagnosed as FNH by CEUS, in nine Regenerative nodules 84 (4.1)
Focal fatty infiltrations 70 (3.4)
cases the final diagnosis was different (five of them ad- Cholangiocarcinomas 57 (2.7)
enomas, two of them hepatocellular carcinomas, one Abscesses 45 (2.2)
haemangioma and one focal fatty infiltration). Thus, the Complex Cysts 43 (2.1)
final diagnosis was FNH in 100 of the 2062 (4.8%) cases. Other benign lesions* 37 (1.8)
The characteristics of patients with FNH are presented in Adenomas 32 (1.6)
Fatty free areas 26 (1.3)
Table II. Other malignant lesions** 19 (0.9)
The indication for CEUS was in most FNH cases an The results are expressed as number, number (%) or mean±SD. *
incidentally found FLL 98% (98/100), in 1% (1/100) pseudolesion, scar area, angiomyolipoma, hamartoma; ** lympho-
case it was the evaluation of a patient with known onco- ma, hemangiosarcoma, hepatic epithelioid hemangioendothelioma.
logic disease and in 1% (1/100) case it was an inconclu-
sive CE-CT scan. At the moment of CEUS examination, Table II. The characteristics of patients with focal nodular
none of the patients diagnosed by CEUS with FNH were hyperplasia (FNH)
known with underlying liver disease. However, two of Number of patients with FNH 100
them were later diagnosed with underlying cirrhosis and Mean age 40.4±13.7 years
2 with chronic hepatitis. Gender, women 75 (75)
CE-MRI was the gold standard in 30% (30/100) cas- Underlying chronic liver disease 4 (4)
es. In 6% (6/100) patients histology was available – one Size:
of them after surgery performed for HCC. ≤ 35 mm 39 (39)
By comparing CEUS with the reference method (ei- ≥ 36 mm 61 (61)
ther CE-CT, CE-MRI or histology) CEUS had 85% Se, The results are expressed as number (%) or mean±SD
99.5% Sp, 90.4% PPV, 99.2% NPV and 98.8% diagnos-
tic accuracy for the diagnosis of FNH. hand, CEUS can be performed immediately after a con-
ventional B mode US examination and thus a diagnosis
Discussion can be established in minutes.
CEUS is a real-time imaging technique, which, simi-
Focal liver lesions (FLLs) are frequently discovered lar to contrast enhanced CT or MRI, shows tissue perfu-
during conventional B mode ultrasound (US) examina- sion, and is able to identify the type of FLL based on the
tion, either during a routine scan for various abdomi- enhancement pattern in arterial, portal and late vascular
nal symptoms or during the follow-up of chronic liver phases. The value of CEUS for the differential diagnosis
disease or of oncologic patients. In most cases, only B of FLLs was proved in a German national multicentre
mode US is not sufficient for a definite diagnosis. This study (DEGUM study) [13] and in a French national
is particularly true when FNH is involved, since its ap- multicentre study (STIC study) [14], each including more
pearance is totally uncharacteristic, sometimes isoechoic, than 1000 FLL and later in a Romanian multicentre study
or slightly hyper- or hypoechoic, the most suggestive for [15]. Meta-analyses demonstrated that CEUS accuracy
the diagnosis being the central scar, which is visible only is similar to contrast enhanced CT and MRI [16-18].
in 70-80% of cases [8,10,11]. Thus, a contrast imaging Other advantages of CEUS as compared with CT and
method or a histopathological exam are required to reach MRI are the absence of radiation exposure (as in CT ex-
a final diagnosis. The time needed for scheduling, and amination) and of side effects (no nephrotoxicity, no io-
the costs of contrast enhanced CT or contrast enhanced dine exposure), and the lower costs [19,20]. Limitations
MRI are not negligible, the same is true regarding guided of CEUS are linked to the limitations of US (poor acous-
biopsy, which is also an invasive procedure. On the other tic window, uncooperative patients) and to the lesion’s
144 Roxana Şirli et al CEUS for the assessment of focal nodular hyperplasia – results of a multicentre study
depth, since FLL located at more than 9 cm from the skin
are difficult to examine due to attenuation [6,12]. How-
ever, if we compare the methods, we discover from the
study of Bröker et al [21] that in a direct comparison be-
tween MRI with liver specific contrast and CEUS, MRI
performed better than CEUS for the diagnosis of FNH
and hepatocellular adenoma, CEUS having a sensitivity
and specificity of 85% and 87%, positive predictive value
(PPV) of 79% and negative predictive value (NPV) of
90%, while MRI reached 95% sensitivity, 95% specific-
ity, 92% PPV and 97% NPV.
The first guidelines regarding the use of CEUS were
issued in 2004 by the European Federation of Societies
for Ultrasound in Medicine and Biology (EFSUMB)
[22], which were revised in 2008 [23], in 2012 [6] and in
2020 [24], the last two developed in cooperation with the
World Federation for Ultrasound in Medicine and Biol-
ogy (WFUMB), thus with universal validity. Regarding
the dose of SonoVue, guidelines recommend the use of
2.4 ml (1/2 vial) for the characterization of FLL, even if
in the last years, with the development of more perfor-
mant US machines, doses as low as 1.2 ml are frequently
used [25]. A too high dose can lead to apparition of arti-
facts (shadowing, over-enhancement of small structures
and signal saturation, as well as “drowning” of small le-
sions in the signal of adjacent structures). A too low dose
can lead to an insufficient quantity of microbubbles in
the late phases, so that the wash-out cannot be observed.
However, if the dose is not optimal, the injection can be
repeated, with a higher or lower dose, as needed [25].
Furthermore, to better evaluate the wash-out, insonation
can be avoided in the arterial and portal phase of the rein-
jection, especially useful for diagnosis of hepatocellular
carcinoma [26].
According to published data, CEUS is an accurate
method to characterise FNH. In the DEGUM study the
accuracy for FNH was 95.5% [27], with 57.1% Se and
99.3% Sp. In the STIC study, CEUS had 82.5% Se and
94.3% Sp for the diagnosis of FNH [14]. The calculated
pooled sensitivity for FNH was 88% in the Friedrich-
Rust et al meta-analysis [18] and the Sp can go as high
as 100% [14,28]. In another study that evaluated the ac-
curacy of CEUS performed by two operators for the di-
agnosis of 85 biopsy proven FNH, CEUS had 80.9% Se,
95.7% Sp, 95.0% PPV, 83.3% NPV and 88.3% accuracy
for operator one and 78.7% Se, 93.6% Sp, 92.5% PPV,
81.5% NPV and 86.2% accuracy for operator two. This
study concluded that CEUS performed better than con-
trast enhanced CT for characterizing dynamic centrifugal
filling or the “spoke-wheel” sign in small lesions [29].
In our study, the performance of CEUS to diagnose
FNH was very good with excellent Sp (99.5%) and ac-
curacy (98.8%) and a slightly lower Se (85%). Regard-
ing the misdiagnosed cases, most confusions were made
with adenomas. Five cases diagnosed by CEUS as FNH
were in fact adenomas and one case diagnosed by CEUS
as adenoma was in fact FNH. The confusion is probably
understandable since both types of FLL show rapid hy-
perenhancement in the arterial phase, but in adenoma the
fill-in occurs initially at the periphery with very rapid
centripetal filling, as opposed to the centrifuge filling
seen in FNH.
Arterial phase can be better evaluated using paramet-
ric arrival time (PAT), a post processing modality of eval-
uation. In the portal phase, adenoma usually becomes
isoechoic or, more rarely, remains slightly hyperechoic.
Sometimes, in adenomas the washout occurs in the late
phase, thus being false positive for malignancy [6,12].
The imaging diagnosis of adenoma is very difficult and
often guided biopsy is needed. In the DEGUM study,
CEUS correctly diagnosed 57.9% of the adenomas [13].
Computer-assisted diagnosis can also be used to improve
the accuracy of differential diagnosis by CEUS between
FNH and adenoma [30].
In the Roche et al study [31] were included 43 FNH
and 20 adenomas, most lesions diagnosed based on the
histopathological exam. The conclusion of this study
was that CEUS had excellent Sp for diagnosing FNH
(100%), but the Se varies according to the lesion’s size:
93% for lesions ≤ 35 mm and 7.7% for lesions > 35 mm,
the overall Se being 67.4 %. The authors explained this
observation by the fact that larger FNH have an increased
vascular supply and several feeding arteries and thus,
the typical spoke-wheel pattern is not visible. The same
observation, that FNHs with signs of centrifugal filling
were smaller than those without the sign (3.1±1.5 cm vs.
5.2±3.2 cm, p=0.000) was made in the study by Wang
et al [29]. In our study, the five FNH misdiagnosed as
adenomas were 2.5 cm, 6 cm, 6 cm, 3 cm, 7.5 and 3.2 cm
in diameter, respectively.
Even if both adenoma and FNH are benign lesions, it
is important to be differentiated since in adenoma there is
a risk of rupture (and subsequent bleeding) or malignant
transformation.
MRI with liver specific-contrast seems to outper-
form CEUS in differentiating FNH form hepatocellular
adenoma. In a study by Grazioli et al, MRI had 96.9%
Se, 100% Sp, 100% PPV, 96.4% NPV, and 98.3% overall
accuracy [32].
Two FLL diagnosed by CEUS as FNH proved to be
hepatocellular carcinomas (HCCs). Both types of lesions
have arterial hyperenhancement but in HCC there is a
chaotic pattern [6,12]. Usually, HCCs show mild, late, or
very late washout, correlated with the differentiation of

the tumour [33,34]. Well-differentiated HCCs show very
late or no washout, thus the CEUS examination should
take at least 4-5 minutes [35]. Thus, in well-differenti-
ated HCCs, the CEUS aspect can be similar to that of
FNH but there is a very important factor that should be
taken into consideration: up to 90% of HCCs occur on
a background of chronic liver disease with severe fibro-
sis and cirrhosis [36]. US based elastography, available
in most US machines able to perform CEUS, should be
used to rule in or rule out the presence of cirrhosis in a
point of care US evaluation [37]. Thus, it is important
to keep in mind that a new lesion on a cirrhotic liver,
hyperenhancing in the arterial phase on CEUS, is prob-
ably HCC [35,38]. Computer assisted diagnosis [39] as
well as parametric imaging could be used to improve the
differential diagnosis by CEUS between FNH and HCC,
the mean transit time being significantly longer in FNH
than in HCC [40].
Both HCCs misdiagnosed as FNH were in patients
with incidentally discovered FLL, who were not known
with chronic liver disease, but in whom, starting from the
HCC, compensated liver cirrhosis was diagnosed.
A limitation of or study is that the pathological exam
was available only in a small number of cases (only 7
of the 100 FNH), but an advantage is the large number
included and the prospective design of the study.
In conclusion, CEUS is an accurate method to diag-
nose FNH, the main difficulties occurring in differentiat-
ing FNH from adenomas, especially in large lesions.
Acknowledgements
Part of the research published in this paper was made
with support from the grant awarded by the “Victor
Babeş” University of Medicine and Pharmacy Timişoara,
in PROGRAMUL III – C2 – PCFI - 2015/2016.
The authors wish to acknowledge the help provided
by the following colleagues: Dorina Pestroiu Calescu,
Lucian Ciobâca, Liana Gheorghe, Mihai Socaciu, Alina
Martie, Siegfried Christian Ivasc, Attila Tamas, Costin
Theodor Streba, Mihaela Iordache, Iulia Simionov, Mari-
ana Jinga, Adrian Anghel, Silviu Marcel Stanciu, Daniel
Stoicescu, Eugen Dumitru, Corina Pietrăreanu, Daniela
Bartoş, Roberta Mânzat Saplacan, Iuliana Pârvulescu,
Roxana Vădan, Gabriela Smira and Liliana Tuţă
References
1. Karhunen PJ. Benign hepatic tumours and tumour like con-
ditions in men. J Clin Pathol 1986;39:183-188.
2. Fukukura Y, Nakashima O, Kusaba A, Kage M, Kojiro M.
Angioarchitecture and blood circulation in focal nodular
hyperplasia of the liver. J Hepatol 1998;29:470-475.
Med Ultrason 2021; 23(2): 140-146 145
3. Gaiani S, Celli N, Piscaglia F, et al. Usefulness of contrast-
enhanced perfusional sonography in the assessment of
hepatocellular carcinoma hypervascular at spiral computed
tomography. J Hepatol 2004;41:421-426.
4. Wanless IR, Albrecht S, Bilbao J, et al. Multiple focal nodu-
lar hyperplasia of the liver associated with vascular malfor-
mations of various organs and neoplasia of the brain: a new
syndrome. Mod Pathol 1989;2:456-462.
5. Dietrich CF, Schuessler G, Trojan J, Fellbaum C, Ignee A.
Differentiation of focal nodular hyperplasia and hepatocel-
lular adenoma by contrast-enhanced ultrasound. Br J Ra-
diol 2005;78:704-707.
6. Claudon M, Dietrich CF, Choi BI, et al. Guidelines and
good clinical practice recommendations for Contrast En-
hanced Ultrasound (CEUS) in the liver - update 2012: A
WFUMB-EFSUMB initiative in cooperation with repre-
sentatives of AFSUMB, AIUM, ASUM, FLAUS and ICUS.
Ultrasound Med Biol 2013;39:187-210.
7. Piscaglia F, Lencioni R, Sagrini E, et al. Characterization of
focal liver lesions with contrast-enhanced ultrasound. Ul-
trasound Med Biol 2010;36:531-550.
8. Carlson SK, Johnson CD, Bender CE, Welch TJ. CT of fo-
cal nodular hyperplasia of the liver. AJR Am J Roentgenol
2000;174:705-712.
9. Ricci P, Laghi A, Cantisani V, et al. Contrast-enhanced so-
nography with SonoVue: enhancement patterns of benign
focal liver lesions and correlation with dynamic gado-
benate dimeglumine-enhanced MRI. AJR Am J Roentgenol
2005;184:821-827.
10. Irie H, Honda H, Kaneko K, et al. MR imaging of focal
nodular hyperplasia of the liver: value of contrast-enhanced
dynamic study. Radiat Med 1997;15:29-35.
11. D’Onofrio M, Crosara S, De Robertis R, Canestrini S, Mu-
celli RP. Contrast-Enhanced Ultrasound of Focal Liver Le-
sions. AJR Am J Roentgenol 2015;205:W56-W66.
12. Sporea I, Badea R, Brisc C, et al. Romanian National
Guidelines on Contrast Enhanced Ultrasound in clinical
practice. Med Ultrason 2017;19:401-415.
13. Strobel D, Seitz K, Blank W, et al. Contrast-enhanced ul-
trasound for the characterization of focal liver lesions--di-
agnostic accuracy in clinical practice (DEGUM multicenter
trial). Ultraschall Med 2008;29:499-505.
14. Tranquart F LGA, Correas JM, Ladam Marcus V, et al. Role
of contrast-enhanced ultrasound in the blinded assessment
of focal lesions in comparison with MDCT and CEMRI:
Results from a multicentre clinical trial. Eur J Cancer Suppl
2008;.6:9-15.
15. Sporea I, Badea R, Popescu A, et al. Contrast-enhanced ul-
trasound (CEUS) for the evaluation of focal liver lesions - a
prospective multicenter study of its usefulness in clinical
practice. Ultraschall Med 2014;35:259-266.
16. Xie L, Guang Y, Ding H, Cai A, Huang Y. Diagnostic value
of contrast-enhanced ultrasound, computed tomography
and magnetic resonance imaging for focal liver lesions: a
meta-analysis. Ultrasound Med Biol 2011;37:854-861.
17. Guang Y, Xie L, Ding H, Cai A, Huang Y. Diagnosis value
of focal liver lesions with SonoVue(R)-enhanced ultra-
146 Roxana Şirli et al CEUS for the assessment of focal nodular hyperplasia – results of a multicentre study
sound compared with contrast-enhanced computed tomog-
raphy and contrast-enhanced MRI: a meta-analysis. J Can-
cer Res Clin Oncol 2011;137:1595-1605.
18. Friedrich-Rust M, Klopffleisch T, Nierhoff J, et al. Con-
trast-Enhanced Ultrasound for the differentiation of benign
and malignant focal liver lesions: a meta-analysis. Liver Int
2013;33:739-755.
19. Giesel FL, Delorme S, Sibbel R, Kauczor HU, Krix M.
Contrast-enhanced ultrasound for the characterization of
incidental liver lesions - an economical evaluation in com-
parison with multi-phase computed tomography. Ultra-
schall Med 2009;30:259-268.
20. Sirli R, Sporea I, Martie A, Popescu A, Danila M. Contrast
enhanced ultrasound in focal liver lesions--a cost efficiency
study. Med Ultrason 2010;12:280-285.
21. Broker MEE, Taimr P, de Vries M, et al. Performance of
Contrast-Enhanced Sonography Versus MRI With a Liv-
er-Specific Contrast Agent for Diagnosis of Hepatocellu-
lar Adenoma and Focal Nodular Hyperplasia. AJR Am J
Roentgenol 2020;214:81-89.
22. Albrecht T, Blomley M, Bolondi L, et al. Guidelines for
the use of contrast agents in ultrasound. January 2004. Ul-
traschall Med 2004;25:249-256.
23. Claudon M, Cosgrove D, Albrecht T, et al. Guidelines and
good clinical practice recommendations for contrast en-
hanced ultrasound (CEUS) - update 2008. Ultraschall Med
2008;29:28-44.
24. Dietrich CF, Nolsoe CP, Barr RG, et al. Guidelines and
Good Clinical Practice Recommendations for Contrast En-
hanced Ultrasound (CEUS) in the Liver – Update 2020 –
WFUMB in Cooperation with EFSUMB, AFSUMB,
AIUM, and FLAUS. Ultraschall Med 2020;41:562-585.
25. Dietrich CF, Averkiou M, Nielsen MB, et al. How to per-
form Contrast-Enhanced Ultrasound (CEUS). Ultrasound
Int Open 2018;4:E2-E15.
26. Dietrich CF, Ignee A, Greis C, Cui XW, Schreiber-Dietrich
DG, Hocke M. Artifacts and pitfalls in contrast-enhanced
ultrasound of the liver. Ultraschall Med 2014;35:108-125.
27. Seitz K, Strobel D, Bernatik T, et al. Contrast-Enhanced
Ultrasound (CEUS) for the characterization of focal liver
lesions - prospective comparison in clinical practice: CEUS
vs. CT (DEGUM multicenter trial). Parts of this manuscript
were presented at the Ultrasound Dreilandertreffen 2008,
Davos. Ultraschall Med 2009;30:383-389.
28. Trillaud H, Bruel JM, Valette PJ, et al. Characterization of
focal liver lesions with SonoVue-enhanced sonography:
international multicenter-study in comparison to CT and
MRI. World J Gastroenterol 2009;15:3748-3756.
29. Wang W, Chen LD, Lu MD, et al. Contrast-enhanced ultra-
sound features of histologically proven focal nodular hy-
perplasia: diagnostic performance compared with contrast-
enhanced CT. Eur Radiol 2013;23:2546-2554.
30. Denis de Senneville B, Frulio N, Laumonier H, Salut C,
Lafitte L, Trillaud H. Liver contrast-enhanced sonography:
computer-assisted differentiation between focal nodular
hyperplasia and inflammatory hepatocellular adenoma by
reference to microbubble transport patterns. Eur Radiol
2020;30:2995-3003.
31. Roche V, Pigneur F, Tselikas L, et al. Differentiation of
focal nodular hyperplasia from hepatocellular adenomas
with low-mechanical-index contrast-enhanced sonography
(CEUS): effect of size on diagnostic confidence. Eur Radiol
2015;25:186-195.
32. Grazioli L, Morana G, Kirchin MA, Schneider G. Ac-
curate differentiation of focal nodular hyperplasia from
hepatic adenoma at gadobenate dimeglumine-enhanced
MR imaging: prospective study. Radiology 2005;236:166-
177.
33. Boozari B, Soudah B, Rifai K, et al. Grading of hypervas-
cular hepatocellular carcinoma using late phase of contrast
enhanced sonography - a prospective study. Dig Liver Dis
2011;43:484-490.
34. Liu GJ, Xu HX, Lu MD, et al. Correlation between en-
hancement pattern of hepatocellular carcinoma on real-
time contrast-enhanced ultrasound and tumour cellular
differentiation on histopathology. Br J Radiol 2007;80:321-
330.
35. Bolondi L, Gaiani S, Celli N, et al. Characterization of
small nodules in cirrhosis by assessment of vascularity: the
problem of hypovascular hepatocellular carcinoma. Hepa-
tology 2005;42:27-34.
36. Sherman M. Hepatocellular carcinoma: epidemiology,
surveillance, and diagnosis. Semin Liver Dis 2010;30:3-
16.
37. Ferraioli G, Wong VW, Castera L, et al. Liver Ultrasound
Elastography: An Update to the World Federation for Ul-
trasound in Medicine and Biology Guidelines and Recom-
mendations. Ultrasound Med Biol 2018;44:2419-2440.
38. Jang HJ, Kim TK, Wilson SR. Small nodules (1-2 cm) in
liver cirrhosis: characterization with contrast-enhanced ul-
trasound. Eur J Radiol 2009;72:418-424.
39. Huang Q, Pan F, Li W, et al. Differential Diagnosis of Atyp-
ical Hepatocellular Carcinoma in Contrast-Enhanced Ultra-
sound Using Spatio-Temporal Diagnostic Semantics. IEEE
J Biomed Health Inform 2020;24:2860-2869.
40. Zheng SG, Xu HX, Liu LN, et al. Parametric imaging with
contrast-enhanced ultrasound: usefulness for characteriza-
tion of dynamic effects of microvascularization for hepa-
tocellular carcinoma and focal nodular hyperplasia. Clin
Hemorheol Microcirc 2013;55:375-389.
Original papers Med Ultrason 2021, Vol. 23, no. 2, 147-152
DOI: 10.11152/mu-2792

Evaluation of bowel preparation before colonoscopy by

ultrasonographic monitoring of colonic fecal retention: a case series
Masaru Matsumoto1, Masayuki Fujioka2, Toshihiko Okada2,3, Yutaka Naka2,3, Ayumi
Amemiya4, Erina Matsushima4, Nao Tamai1,6, Yuka Miura1, Gojiro Nakagami5,6, Hiromi
Sanada5,6

1Department of Imaging Nursing Science, Graduate School of Medicine, The University of Tokyo, Tokyo, 2Katsuragi
Hospital, Osaka, 32nd department of Internal medicine, Osaka Medical College, Osaka, 4Department of Nursing
Physiology, Graduate School of Nursing, Chiba University, Chiba, 5Department of Gerontological Nursing / Wound
Care Management, Graduate School of Medicine, The University of Tokyo, Tokyo, 6Global Nursing Research Center,
Graduate School of Medicine, The University of Tokyo, Tokyo, Japan.

Abstract
Aims: While bowel preparation for colonoscopy is the key to successful examination, taking laxatives and showing stools
to others causes both physical and mental distress to the patient. Thus, an alternative method to evaluation bowel preparation
is necessary. In the current study, we studied the colonic fecal retention by ultrasonography (US) and examined the US finding
which reflected completion of BP. Material and methods: The subjects were outpatients who underwent colonoscopy. This
report summarizes the ultrasonographic images of patients who underwent multiple US examinations for all five sites of the
colon just before, during, and immediately after bowel preparation. According to the standard protocol, the patients took 2
L of polyethylene glycol-ascorbic acid as a laxative, which was discontinued when the nurse visually judged the stool to be
clear. Results: Seven patients in their 50s–80s, none of whom were unable to complete a colonoscopy due to residual feces
were included in study. Following bowel preparation, the US images showed anechoic areas with haustration in four or all five
areas of the colon. Three of the seven patients received low-dose laxatives (1.1–1.2 L); all three had watery stools in three or
more colon areas and none of them were constipated at the time of taking 1 L of laxatives. Conclusions: Completion of bowel
preparation can be assessed by the observation of anechoic areas with haustration in multiple colonic sites by ultrasonography.
Keywords: ultrasonography; colonoscopy; fecal retention; laxative; bowel preparation

Introduction Previous studies have indicated that up to one-quarter

Colonoscopy is the gold standard procedure for the
diagnosis and treatment of colon diseases [1]. Bowel
preparation (BP) is critical for high quality colonoscopy,
as it is closely related to the quality of the procedure [2].
Received 26.08.2020  Accepted 12.12.2020
Med Ultrason
2021, Vol. 23, No 2, 147-152
Corresponding author: Hiromi Sanada, PhD, RN, WOCN, FAAN
Global Nursing Research Center, Graduate
School of Medicine, The University of Tokyo
7-3-1 Hongo, Bunkyo-ku, Tokyo,
113-0033, Japan
Phone: (+81) 3-5841-3419
Fax: (+81) 3-5841-3442
E-mail: hsanada-tky@umin.ac.jp
of colonoscopies may be conducted with inadequate BP
[3,4], and suboptimal BP is thought to occur in up to 20%
of all colonoscopy procedures [5,6]. Insufficient BP in-
creases the requirement for repeat examinations, which
ultimately increases healthcare costs and patient burden
[7,8]. Therefore, it is necessary to be able to evaluate BP
appropriately, perform bowel cleansing correctly and
properly complete BP.
BP is also associated with patient distress. First, a
nurse usually determines the degree of bowel cleansing
from the color and properties of the stool discharged by a
patient [9]. In other words, the patient’s excretion is seen
by others, which may cause mental distress to patients.
Therefore, it is important to determine another meth-
od to directly visualize the condition of the colon and
148 Masaru Matsumoto et al Evaluation of bowel preparation before colonoscopy by US monitoring of colonic fecal retention
manage bowel cleansing. Second, patients who undergo
colonoscopy usually have to take a considerable amount
of laxatives, often around 2L [10], and taking more lax-
atives than is necessary may cause physical distress. The
laxative requirement and BP before colonoscopy should
depend on the individual’s colonic condition, so as to
minimize physical distress. Indeed, it has been reported
that the quality of BP is different between patients with
and without constipation [11]. Therefore, it is import-
ant to develop a visualization tool that can directly and
non-invasively monitor the state of the patient’s colon
and manage BP.
Previous studies have reported that US can visualize
fecal retention in the colon and some studies have shown
that stool retention can be assessed by the presence of
a crescent-shaped hyperechoic area [12-16]. The aim of
our study was to evaluate by US the colonic fecal reten-
tion in patients prepared for colonoscopy and to establish
which US findings reflect the completion of BP.
Material and methods
This study, conducted in May 2019, was approved by
the Ethical Committee of the University of Tokyo (No.
2020017NI). The institutional review board approved
this case series and waived the need for informed consent
from patients; all participants were given the opportunity
to opt-out by the poster and on the web site regarding the
use of data.
Participants and setting
The study subjects were outpatients who underwent
colonoscopy in a Japanese acute care hospital. In this
hospital, patients with constipation or those who did not
want to take large doses of laxatives were often checked
for fecal retention using US during BP. A maximum of
three US examinations were conducted (fig 1) and the BP
was considered to be finished when the stool was judged
to be transparent by a nurse using photographic examples
of rectal effluent [9].
Bowel preparation
BP was standardized as follows: The day before the
examination, intakes were restricted to a light, low-fiber
breakfast and after midday, only clear liquids were al-
lowed. In addition, the evening before the examination,
patients were required to take three Sennoside tablets.
On the day of the examination, patients were hospital-
ized in the outpatient unit for the second phase of colon
preparation with 2L polyethylene glycol-ascorbic acid
(MoviPrep®, EA Pharma Co.,Ltd., Tokyo, Japan). Nurs-
es confirmed the stool color and properties in the toilet
when the patient determined that the stool became a clear
liquid.
Fig 1. Ultrasonography during bowel preparation. A maximum
of three ultrasonographies (US) were conducted. Bowel prepa-
ration was considered to be finished when the nurse confirmed
a clear stool in the toilet; the total oral dose at that time was
recorded.
Ultrasound
US of the rectum, ascending, transverse, descending
and sigmoid colon was performed using ultrasound sys-
tems (Aplio 500 and Aplio a450systems, Canon Medi-
cal System Corporation, Otawara, Tochigi, Japan) with
a curved-array probe (1.5–6.0 MHz). The resulting im-
ages were supplemented by transverse and longitudinal
sonographic scans. All sonographic procedures were
performed by a certified sonographer specializing in the
abdominal region, with 20 years of experience. The so-
nographer was blinded to the results of the assessment
for stool color and properties.
With reference to previous studies [12,17], we de-
fined the US levels of fecal retention as follows: a weak/
strong fecal retention finding was indicated by a margin-
ally highly echoic colonic lumen and posterior echo be-
hind the colon on transverse images and a flattened outer
boundary wall and highly echoic colon wall on longitudi-
nal images; a watery stool retention was indicated by an
anechoic area on transverse images and an anechoic area
with haustration on longitudinal images [18]. The US
finding of multiple reflections was determined to be gas
retention. The typical ultrasonographic findings related
to fecal retention are shown in figure 2 and 3.
Colonoscopy
Several trained gastroenterologists performed the co-
lonoscopies. A colonoscopy was defined as incomplete
when there was no visualization of anatomic features,
such as the ileocecal valve, appendiceal orifice, ileoco-
lonic anastomosis, or terminal ileum, as previously de-
scribed [19].
Results
A total of seven patients (five men, 71.4%) were in-
cluded in study and their characteristics are summarized
in Table I. The patients’ age at presentation ranged from
50 to 83 years, with a median age of 69 years. Six patients
(85.7%) took Sennoside (12 mg; 3 tablets after dinner

Fig 2. Ultrasonographic findings related to fecal retention: a) colonic fecal retention in a 69-year-old male patient. A longitudinal
ultrasonographic image showing hyperechoic areas and acoustic shadow with haustrations on the wall of the descending colon (ar-
rows); b) colonic gas in a 59-year-old male patient. A longitudinal ultrasonographic image showing multiple reflections on the wall of
the descending colon (arrow); c) watery stool in a 83-year-old male patient. A longitudinal ultrasonographic image showing anechoic
areas, indicating water retention in the descending colon (arrows).
and 3 tablets before sleep) prescribed by their doctors.
Three patients (42.9%) had daily bowel movements, with
normal stool properties and amounts. On the morning of
the colonoscopy, six of the seven patients (85.7%) had
a bowel movement. Time from the start of 2 L polyeth-
ylene glycol-ascorbic acid to taking 1 L ranged 67 to 80
minutes, with a median time of 73 minutes. Time from
the start of 2 L polyethylene glycol-ascorbic acid to the
completion of bowl preparation ranged 91 to 155 min-
utes, with a median time of 121 minutes.
The change in ultrasonographic findings, total dose
of laxative and degree of bowel cleanliness are shown in
Table II. None of the patients were determined to have
Fig 3. Time course of ultrasonographic images at colon in a
69-year-old male patient (ID1). Longitudinal ultrasonograph-
ic images showing anechoic areas, indicating water retention
in the descending colon (arrows). The stool was judged to be
transparent by the nurse, and the patient finished bowel prepa-
ration after taking 1.1 L of laxative.
Med Ultrason 2021; 23(2): 147-152 149
an incomplete colonoscopy. The ultrasound images at the
completion of BP showed anechoic areas with haustra-
tion in four, or all five areas of the colon. Three patients
(Cases 1–3) received low-dose laxatives (1.1 to 1.2 L);
the US images of these three patients showed anecho-
ic findings indicating watery stool retention in three or
more colon areas. None of the three patients were consti-
pated at the time of consuming 1 L of laxatives.
Discussions
This study firstly showed colonic US imaging at BP
before colonoscopy. The US results demonstrated that
four or all five areas of the colon showed anechoic areas
with haustration and the lack of hyperechoic area in the
stool indicated that BP had been completed. Patients who
received 1 L of laxative, had watery stools in three or more
of the colon sites and no fecal retention in any of these
sites, received a smaller dose of laxatives. Therefore, it is
possible to determine by US that BP has been completed.
The US images at the completion of BP showed an-
echoic areas with haustration in four or all five areas of
the colon, which indicated that BP was complete. As a
practical point for the future, endoscopy nurses should
be able to observe the colon with a hand-held US device
to assess BP completion without causing the patient the
mental distress of having to have their stool observed.
In the same way, that the effectiveness of hand-held
US observation and treatment of constipation has been
shown [20,21], the effectiveness of US observation by
endoscopy nurses needs to be studied in the near future.
There will also be a need to develop a US education pro-
gram for nurses to observe colonic fecal retention [22].
In addition, non-invasive assessment of BP with US may
help to determine the timing of BP completion for each
patient and minimize the physical distress caused by lax-
atives. Indeed, constipated patients have been reported
 150
Table I. Patients characteristics, bowel preparation and defecation status.
ID Age Sex BMI Primary disease Laxative until Defecation status Bowel preparation
(kg/m2) the day before
Usual Morning of colonoscopy day TTT 1L (min) TTC (min)
1 69 M 19.3 HT, hyperlipidemia Sennoside 12 mg* Bowel movement: 3 times/week 1 soft stool 67 121
Stool: properties: normal
Stool amount: normal
Masaru Matsumoto et al

2 83 M 31.0 HT, diabetes, constipation Sennoside 12 mg* Bowel movement: 2-3 times/week 4 soft or watery stools 75 155
Stool: properties: normal or hard
Stool amount: normal
3 51 M 26.6 Anal polyp, hernia Sennoside 12 mg* Bowel movement: 7 times/week 3 soft stools 73 91
Stool: properties: normal
Stool amount: normal
4 50 F 41.7 HT, diabetes, reflux es- Sennoside 12 mg* Bowel movement: 7 times/week 1 soft stool 80 121
ophagitis Stool: properties: normal
Stool amount: normal
5 71 F 21.7 Bronchial asthma, bowel None Bowel movement: 7 times/week 3 soft stools 72 155
obstruction Stool: properties: soft
Stool amount: normal
6 59 M 27.2 Colon polyps, esophagitis Sennoside 12 mg* Bowel movement: 7 times/week 3 soft stools 68 124
Stool: properties: normal
Stool amount: normal
7 69 M 26.6 Chronic gastritis, Sennoside 12 mg* Bowel movement: 2 times/week None no data 105
constipation Stool: properties: normal
Stool amount: normal

BMI: body mass index, TTT: time to take, TTC: time to completion, tb: tablets, HT: hypertension; *: 3 tb after dinner and 3 tb before sleep
Evaluation of bowel preparation before colonoscopy by US monitoring of colonic fecal retention
 Med Ultrason 2021; 23(2): 147-152 151
Table II. Change of colonic contents based on ultrasonographic findings, total dose of laxative and degree of bowel cleanliness.
ID Site Ultrasonographic findings Total dose of
Baseline When taking 1 L of laxative At the end of bowel preparation laxative
1 AC + watery stool watery stool 1.1L
TC + watery stool watery stool
DC - watery stool watery stool
SC watery stool watery stool watery stool
R watery stool watery stool watery stool
2 AC + watery stool watery stool 1.2 L
TC + watery stool watery stool
DC - watery stool watery stool
SC - watery stool watery stool
R - - watery stool
3 AC - watery stool watery stool 1.2 L
TC - watery stool watery stool
DC - watery stool watery stool
SC gas - watery stool
R + - -
4 AC + + watery stool 1.4L
TC - watery stool watery stool
DC - watery stool watery stool
SC - watery stool watery stool
R - - watery stool
5 AC watery stool watery stool watery stool 1.5 L
TC watery stool watery stool watery stool
DC + watery stool watery stool
SC gas watery stool watery stool
R - + watery stool
6 AC watery stool watery stool watery stool 1.6L
TC - watery stool watery stool
DC gas - watery stool
SC + gas watery stool
R - - watery stool
7 AC + no data watery stool 2.0 L
TC + no data watery stool
DC + no data watery stool
SC + no data watery stool
R - no data -
AC: Ascending colon, TC: Transverse colon, DC: Descending colon, SC: Sigmoid colon, R: Rectum, +: Hyperechoic finding indicating
fecal retention, -: No specific finding

to have less effective BP than non-constipated patients
[11,18] and in these cases, observation-based, individu-
alized support for BP is required before colonoscopy. In
the cases IDs 1, 2 and 3, all received low doses of laxa-
tives (1.1–1.2 L); these three patients had watery stools
in three or more colonic areas and no fecal retention was
observed in any of these cases after 1 L of laxative. ID7
usually defecated twice a week, which was the least fre-
quent of the participants in this study. Although data are
not available for 1 L of laxatives, patients with constipa-
tion may need to take more laxatives than those without
symptoms. It will be necessary to investigate whether
evaluation of BP using US can reduce the amount of lax-
atives taken by patients.
The present study has some limitations. Because this
study was based on the assumption that polyethylene
glycol was taken internally, we did not consider the use
of sodium phosphate or sodium picosulfate with magne-
sium citrate. Furthermore, the quality of BP was not as-
sessed from endoscopic images [23]. Future studies will
need to investigate the relationship between US images
and endoscopic images.
152 Masaru Matsumoto et al Evaluation of bowel preparation before colonoscopy by US monitoring of colonic fecal retention
Conclusions. Our results demonstrate that comple-
tion of BP can be assessed by observing the anechoic
area with haustration observed in multiple colonic sites
by US.
Acknowledgments: The authors are deeply grateful
to the study participants, all of whom contributed greatly
to this study. This research was supported by joint re-
search funding with the Department of Imaging Nursing
Science.
Conflict of interest: Masaru Matsumoto, Nao Tamai
and Yuka Miura belong to a social collaboration depart-
ment that receives funding from Fujifilm Corporation.
References
1. Mihalko SL. Implementation of colonoscopy for mass
screening for colon cancer and colonic polyps: efficiency
with high quality of care. Gastroenterol Clin North Am
2008;37:117-128.
2. Coriat R, Polin V, Oudjit A, et al. Gastric emptying evalua-
tion by ultrasound prior colonoscopy: an easy tool following
bowel preparation. World J Gastroenterol 2014;20:13591-
13598.
3. Harewood GC, Sharma VK, de Garmo P. Impact of colo-
noscopy preparation quality on detection of suspected co-
lonic neoplasia. Gastrointest Endosc 2003;58:76-79.
4. Lebwohl B, Kastrinos F, Glick M, Rosenbaum AJ, Wang T,
Neugut AI. The impact of suboptimal bowel preparation on
adenoma miss rates and the factors associated with early re-
peat colonoscopy. Gastrointest Endosc 2011;73:1207-1214.
5. Cheng CL, Liu NJ, Tang JH, et al. Predictors of Suboptimal
Bowel Preparation Using 3-l of Polyethylene Glycol for
an Outpatient Colonoscopy: A Prospective Observational
Study. Dig Dis Sci 2017;62:345-351.
6. Radaelli F, Paggi S, Repici A, et al. Barriers against split-
dose bowel preparation for colonoscopy. Gut 2017;66:1428-
1433.
7. Rex DK, Imperiale TF, Latinovich DR, Bratcher LL. Im-
pact of bowel preparation on efficiency and cost of colonos-
copy. Am J Gastroenterol 2002;97:1696-1700.
8. Clark BT, Rustagi T, Laine L. What level of bowel prep
quality requires early repeat colonoscopy: systematic
review and meta-analysis of the impact of preparation
quality on adenoma detection rate. Am J Gastroenterol
2014;109:1714-1723.
9. So H, Boo SJ, Seo H, et al. Patient descriptions of rectal ef-
fluents may help to predict the quality of bowel preparation
with photographic examples. Intest Res 2015;13:153-159.
10. Khan MA, Patel KB, Nooruddin M, et al. Polyethylene Gly-
col-3350 (Miralax®)+1.9-L sports drink (Gatorade®)+2
tablets of bisacodyl results in inferior bowel preparation for
colonoscopy compared with Polyethylene Glycol-Ascorbic
Acid (MoviPrep®). Turk J Gastroenterol 2018;29:67-74.
11. Kunz L, Gillespie D. A Comparison of Bowel Preparations
for Colonoscopy in Constipated Adults. Gastroenterol Nurs
2017;40:364-372.
12. Matsumoto M, Tanaka S, Yabunaka K, et al. Ultrasono-
graphic evaluation of changes over time in one defecation
cycle in adults with functional constipation: A report of two
cases. Drug Discover Ther 2018;12:304-308.
13. Yabunaka K, Matsumoto M, Yoshida M, et al. Assessment
of rectal feces storage condition by a point-of-care pocket-
size ultrasound device for healthy adult subjects: A prelimi-
nary study. Drug Discov Ther 2018;12:42-46.
14. Yabunaka K, Matsuo J, Hara A, et al. Sonographic visuali-
zation of fecal loading in adults comparison with computed
tomography. J Diagn Med Sonog 2015;31:86-92.
15. Yabunaka K, Nakagami G, Komagata K, Sanada H. Ultra-
sonographic follow-up of functional chronic constipation in
adults: A report of two cases. SAGE Open Med Case Rep
2017;5:2050313x17694234.
16. Matsumoto M, Tsutaoka T, Nakagami G, et al. Deep learn-
ing-based classification of rectal fecal retention and analy-
sis of fecal properties using ultrasound images in older
adult patients. Jpn J Nurs Sci 2020;17:e12340.
17. Kawamura T, Nishiguchi T. Congenital Chloride Diarrhea
(CCD): A Case Report of CCD Suspected by Prenatal Ul-
trasonography and Magnetic Resonance Imaging (MRI).
Am J Case Rep 2017;18:707-713.
18. Fan XP, Zhu Q, Zhou YJ, Ma T, Xia CX, Huang HL. Com-
parative Study of Three Regimens of Bowel Preparation
Before Transabdominal Ultrasonography of the Colon. Ul-
trasound Med Biol 2016;42:2140-2145.
19. Neerincx M, Terhaar sive Droste JS, Mulder CJ, et al. Co-
lonic work-up after incomplete colonoscopy: significant
new findings during follow-up. Endoscopy 2010;42:730-
735.
20. Matsumoto M, Yabunaka K, Yoshida M, et al. Improve-
ment of Constipation Symptoms in an Older Adult Patient
by Defecation Care Based on Using a Handheld Ultrasound
Device in Home Care Settings: A Case Report. J Wound
Ostomy Continence Nurs 2020;47:75-78.
21. Matsumoto M, Yoshida M, Yabunaka K, et al. Safety and
efficacy of a defecation care algorithm based on ultrasono-
graphic bowel observation in Japanese home-care settings:
a single-case, multiple-baseline study. Geriatr Gerontol Int
2020;20:187-194.
22. Matsumoto M, Yoshida M, Miura Y, et al. Feasibility of the
constipation point-of-care ultrasound educational program
in observing fecal retention in the colorectum: a descriptive
study. Jpn J Nurs Sci 2020. doi:10.1111/jjns.12385.
23. Kastenberg D, Bertiger G, Brogadir S. Bowel prepara-
tion quality scales for colonoscopy. World J Gastroenterol
2018;24:2833-2843.
Original papers Med Ultrason 2021, Vol. 23, no. 2, 153-160
DOI: 10.11152/mu-3005

The value of abdominal ultrasonography compared to colonoscopy

and faecal calprotectin in following up paediatric patients with
ulcerative colitis
Ioana Fodor1, Oana Serban2, Daniela E Serban3, Dorin Farcau1, Otilia Fufezan4, Carmen
Asavoaie4, Sorin Claudiu Man1, Dan Lucian Dumitrascu2

13rdPediatric Department, “Iuliu Hatieganu” University of Medicine and Pharmacy, 22nd Internal Medicine Department,
“Iuliu Hatieganu” University of Medicine and Pharmacy, 32nd Pediatric Department, “Iuliu Hatieganu” University of
Medicine and Pharmacy, 4Radiology Department, Clinical Emergency Hospital for Children, Cluj-Napoca, Romania

Abstract
Aim: To evaluate the value of abdominal ultrasonography (US) in the follow-up of paediatric patients with ulcerative
colitis (UC) compared to faecal calprotectin (FC) and colonoscopy. Material and method: In this retrospective study we
enrolled 30 paediatric patients previously diagnosed with UC, examined by abdominal US and colonoscopy within the same
week. FC was also determined during the same week. Disease activity was established using the paediatric ulcerative colitis
activity index (PUCAI). The global endoscopic activity was evaluated using the Mayo endoscopic subscore. Results: Endos-
copy revealed pathological findings of active disease in 27 out of 30 patients; 3 patients were in endoscopic remission. Only
18 of them had clinical active disease (PUCAI >10), [sensitivity (Se) 66.7% and specificity (Sp) 33% of PUCAI in detecting
endoscopic active disease). Twenty-three (76.7%) patients had FC >250 mcg/g, but in 2 of these cases the colonoscopy was
normal (Se 77.8% and Sp 33.3% in detecting active disease). At US examination, pathological findings (increased bowel wall
thickness, hypervascularity, lymphadenopathies, and/or mesenteric inflammatory fat) were found in 27 patients (90%), all with
endoscopic active disease (agreement US - colonoscopy, at patient level, k=1.0, p<0.001, Se 100% and Sp 100%). At seg-
ment level (totally 180 bowel segments examined by US), the overall agreement between US and colonoscopy was k=0.767,
p<0.001, Se 86.5%, Sp 90.1%. Of the 27 patients with US pathological findings in any of colonic segments, 23 had FC >250
mcg/g (85.1%). The inter-observer agreement for the US measurements had an overall ICC of 0.926 with p<0.001. Conclu-
sion: Abdominal US findings demonstrate a good to excellent concordance with endoscopic examination and are correlated
with elevated FC levels. Therefore, US appears as an accurate technique in assessing activity in patients with UC and might
replace colonoscopic evaluation for the follow-up.
Keywords: ultrasonography; colonoscopy; faecal calprotectin; IBD; ulcerative colitis

Introduction increasing prevalence [1-3] and a major long-term impli-

cation in the patient’s growth, pubertal development and
Paediatric inflammatory bowel disease (IBD), with the quality of life [4].
the two main entities, Crohn’s disease (CD) and ulcera- Children with UC commonly present with abdominal
tive colitis (UC), is a chronic, relapsing inflammatory pain, diarrhoea and rectal bleeding [5,6], but extraintes-
condition of the gastrointestinal tract, with a worldwide tinal manifestations (6 to 17% of the patients [7,8]) in-
Received 24.11.2020  Accepted 30.12.2020
cluding arthritis, uveitis or liver and colorectal carcinoma
Med Ultrason [9,10] can also be detected.
2021, Vol. 23, No 2, 153-160 The European Society of Paediatric Gastroenterol-
Corresponding author: Ioana Fodor ogy, Hepatology and Nutrition (ESPGHAN) revised
3rd Pediatric Department
2-4 Campeni street,
Porto criteria [11] for UC diagnosis include typical clini-
400217, Cluj-Napoca, Romania cal manifestations and the confirmation of chronic active
E-mail: fodorioana@ymail.com inflammation by colonoscopy (classically, starting from
154 Ioana Fodor et al Following up paediatric patients with ulcerative colitis: abdominal US vs. colonoscopy and faecal calprotectin
rectum with ascending progression in a continuous man-
ner with no small bowel involvement) with multiple biop-
sies. Typical endoscopic findings are represented by ery-
thematous, bleeding, friable, erosive, ulcerative mucosa
and loss of vascularity [11,12]. The microscopic features
of UC are represented by active (neutrophils infiltration)
and chronic inflammation (modified architecture of the
crypts and lymphoplasmacytic infiltrates) [11,13,14].
The disease activity is currently assessed using the
paediatric ulcerative colitis activity index (PUCAI), co-
lonoscopy with multiple biopsies and faecal calprotectin
(FC). Colonoscopy with biopsies is the gold standard
for IBD diagnosis [15]. The recent ESPGHAN and Eu-
ropean Crohn’s and Colitis Organisation (ECCO) guide-
lines recommend colonoscopic evaluation not only for
the diagnosis but also for follow-up and assessment of
therapeutic response (when major changes in treatment
are necessary, cancer surveillance, when FC is elevated,
but the symptoms are not clearly disease-related, and in
patients with sustained clinical remission with high level
of FC) [9]. The main limitations of endoscopy are the
need of general anaesthesia and the possible complica-
tions, such as gastrointestinal bleeding and perforation
[16,17]. Colonoscopy is not easily accepted by children
and their caregivers, especially the repeated procedures,
performed just to confirm the mucosal healing.
FC represents a non-invasive marker of inflammation
[18]. No ideal cut-off value of FC has been yet estab-
lished [19,20] but according to ESPGHAN revised Porto
criteria in children and adolescents [11], a value above
250 µg/g reflects more accurately the mucosal inflam-
mation, with sensitivity (Se), specificity (Sp), positive
predictive value (PPV) and negative predictive value
(NPV) of 71%, 100%, 100% and 47.1%, respectively
[21].
Among the imaging techniques, Magnetic Resonance
Enterography (MRE) has high accuracy in identifying
the bowel wall modifications and the disease complica-
tions [11,22]. However, high-quality images in children
are more difficult to be obtained when compared to adults
[23], due to the movement artefacts, poor acceptance of
the oral contrast and the need of general anaesthesia [24].
The low accessibility, long examination time, need for
specific radiological competence and high cost are other
limits of this method [25,26].
Abdominal ultrasound (US) is a non-invasive, low
cost, accessible, reproductible and easily accepted by
children imaging technique, used more and more fre-
quently in the recent years in assessing bowel inflamma-
tion [27,28]. It is largely used in clinical practice for ab-
dominal complaints or check-up examinations. US was
found to have 79.7% sensitivity and 96.7% specificity
[29] in assessing the extension and activity of IBD [30],
with a good concordance with the MRE findings [23,31].
In a recent systematic review in pediatric IBD, van Was-
senaer et al [30] concluded that the increased bowel wall
thickness (BWT), bowel wall vascularization, presence
of enlarged lymph nodes, alteration of wall stratification,
creeping fat, abscesses or other complications and the ab-
sence of colonic haustration were the major US criteria
for IBD. The majority of the included studies regarded
both UC and CD patients (with a smaller number of UC),
or only CD patients. Therefore, data about the usefulness
of US in paediatric UC are limited [32]. Moreover, moni-
toring IBD patients using US is not yet part of the routine
standardized clinical practice, despite the good results of
published studies, in both children and adults [33-37].
Given these reasons, our aim was to analyse the value
of US in following up UC paediatric patients, compared
to FC and colonoscopy.
Material and methods
This was a retrospective study, approved by the local
Ethics Committee. Written informed consent was signed
by the patient’s legal tutors before investigations, accord-
ing to the hospital protocol.
Study Population
From the hospital database, we selected the paediatric
patients previously diagnosed with UC, according to the
revised Porto criteria [11]. Patients were examined dur-
ing admission by abdominal US and colonoscopy, within
the same week, between January 2018 and January 2020.
We excluded from the study the patients that did not pro-
vide a faecal sample. Patients without a legal tutor on
admission were also excluded. Disease location and phe-
notype were determined using the Paris Classification of
IBD [38] and data about the medications were collected.
Fecal calprotectin
All patients provided a faecal sample, stored at 4 °C
until processed [19], for the FC determination using
ELISA.
Abdominal US
The protocol of the hospital regarding the abdomi-
nal US in IBD patients includes a standard preparation,
patients being advised to fast for at least 8 hours. All the
US examinations were performed by one of the two pae-
diatric radiologists with more than 10 years of experience
in bowel US, using XarioTM 200 (Cannon Medical Sys-
tems Corporation, Otawara, Japan) or Aplio 500 machine
(Canon  Medical Systems Corporation, Otawara, Japan)
with linear (7-14 MHz) and convex (4-6 MHz) transduc-
ers, without oral contrast medium administration. The
investigation protocol included systematic evaluation
 Med Ultrason 2021; 23(2): 153-160 155

Fig 1. Normal aspect of the bowel segments: a) longitudinal scan of the terminal ileum; b) transverse scan of the ascending colon;
c) longitudinal scan of the transverse colon; d) longitudinal scan of the descending colon; e) transverse aspect of the sigmoid;
f) longitudinal scan of the rectum (measurements in all cases between callipers).

of the abdomen starting with the intra-abdominal solid
organs using a low frequency convex probe, with the pa-
tient lying supine [27]. For a better examination of the
large bowel, the high frequency linear probe [39] with
graded compression [40] was used, starting at the cae-
cum in the right iliac fossa, after the localization of the
terminal ileum and progressing distally for the evaluation
of ascending, transverse, descending and sigmoid colon
[27]. Representative images from each bowel segment,
normal or pathological (fig 1 and 2), were stored in each
case in the Picture Archiving Communication System
(PACS), as the standard protocol of our hospital requires.
The images stored in PACS were analysed individu-
ally by the two radiologists, blinded to the result of the
colonoscopy, value of FC and PUCAI. In all bowel seg-
ments the following parameters were evaluated: BWT
(normal value <3 mm for all segments), bowel vascular-
ity (using power Doppler), presence of lymphadenopa-
thies and mesenteric inflammatory fat, following the
published protocols [27,41]. In case of disagreement
between the examiners, the images were reviewed and
conclusion was reached by consensus. The disease ac-
tivity for each bowel segment on US was classified as
present/absent based on the radiologist’s comprehensive
examination.
Colonoscopy
All endoscopic evaluations were performed using
paediatrics OLYMPUS CV-190 (Hamburg, Germany)
endoscopes. All patients underwent bowel preparation
with osmotic laxative, the afternoon before the colonos-
copy [42]. On the day of examination, only clear liquids
were permitted. The colonoscopy procedure and the as-
sessment of inflammation, disease localization (rectum,
sigmoid colon, descending colon, transverse, ascending
colon, caecum and terminal ileum), severity and com-
plications were documented using a standardized work-
sheet by a paediatric gastroenterologist with more than
10 years of experience. The global endoscopic activity
was evaluated using the Mayo endoscopic subscore [43]
(0 – normal/inactive, 1 – mild, 2 – moderate and 3 – se-
vere disease) and the result was recorded immediately
after the procedure.

Fig 2. Pathological aspect of the colon: a) transverse scan of the ascending colon, color Doppler ultrasound – the walls are thickened
and hypervascularized; b) transverse scan of the proximal descending colon with increase bowel wall thickness; c) longitudinal scan
of the distal descending colon with increase bowel wall thickness; d) mesenteric inflammatory lymphadenopathies. Of note the in-
creased thickness and echogenicity of the submucosal layer.
156 Ioana Fodor et al Following up paediatric patients with ulcerative colitis: abdominal US vs. colonoscopy and faecal calprotectin

Statistical analysis Table I. Patients’ baseline characteristics

The Shapiro-Wilks test was used to assess the distri- Patient baseline characteristics (N=30)
bution of continuous variable. Descriptive analysis was Age (years) 12.3 (8.4-16.45)
performed and the results were presented as number (per- Sex (males) 22 (73)
cent) for categorical variables and median (interquartile Disease localization at diagnosis/study inclusion
range) for continuous variables – all being non-normally
Proctitis 10 (33.3)/9 (30)
distributed. The differences of medians between more
Left-sided colitis 15 (50) /13 (43.3)
than two independent samples were assessed using the
Independent Samples – Kruskal-Wallis test. The agree- Pancolitis 5 (16.6)/8 (26.6)
ment between the diagnosis methods of disease activity Disease duration (years) 2.68 (0.5-3.7)
was measured by calculating the Cohen’s kappa coeffi- Mayo endoscopic subscore
cients (k). Cohen’s k values were interpreted as follows: 0 3 (10)
< 0 no agreement, 0–0.20 poor, 0.21–0.40 fair, 0.41–0.60 1 10 (33.3)
moderate, 0.61–0.80 good, and 0.81–1 very good agree- 2 15 (50)
ment. The sensitivity (Se) and specificity (Sp) for each 3 2 (6.6)
method were also calculated, considering endoscopy as Clinical activity (PUCAI)
gold standard for disease activity. The intraclass correla- Inactive (<10) 10 (33.3)
tion coefficient (ICC) was calculated to assess the overall Mild (10-34) 14 (46.6)
inter-observer agreement for the US measurements. The
Moderate (35-64) 6 (20)
ICC value was interpreted as poor when less than 0.50,
Severe (>65) 0 (0)
moderate between 0.50 and 0.75, good between 0.75 and
Medication*
0.90, and excellent when greater than 0.90. A p-value
less than 0.05 was considered statistically significant. 5-ASA 27 (90)
IMB SPSS Statistics v.23 and Microsoft Office 365 Excel Steroids 11 (36.6)
were used to perform the statistical analysis. Azathioprine 1 (3.3)
FC (mcg/g) 300.0 (246.5-700.0)
Results Results are expressed as number (%) or as median (interquartile
range); Mayo endoscopic subscore: 0 – normal/inactive, 1 – mild,
2 – moderate and 3 – severe disease; PUCAI: pediatric ulcerative
A total number of 30 children previously diagnosed colitis activity index; FC: fecal calprotectin; * more than one type
with UC were selected from the hospital database. Pa- of medication possible in any patient
tients’ baseline characteristics are shown in Table I.
Of the 30 patients included in study, 27 (90%) had follows: in one rectum, 4 sigmoid colon, 4 descending
active lesions at colonoscopy, but only 18 of the 27 had colon, 2 transverse colon and one ascending colon seg-
clinical active disease with PUCAI >10, (Se 66.7%, Sp ments, all with colonoscopy inflammatory signs. In 2 de-
33% of PUCAI in detecting active disease). FC >250 scending colon, 2 transverse colon and 2 ascending colon
mcg/g was found in 23 (76.7%) patients, but in 2 of these segments US detected increased BWT despite the normal
cases the colonoscopy was normal (Se 77.8%, Sp 33.3% colonoscopic aspect.
of FC in identifying the active disease). The differences between the medians of the BWT of
At US, pathological findings were found in 27 pa- each bowel segment according to the Mayo endoscop-
tients, all with endoscopic active disease (excellent ic subscore are listed in Table III. Lymphadenopathies,
agreement between US and colonoscopy, at patient level, bowel hypervascularisation and mesenteric inflammato-
k=1.0, p<0.001, Se=100%, Sp=100%). When the results ry fat were detected in all patients with Mayo endoscopic
were analysed on a segment level (totally 180 bowel seg- subscore 3, whereas only 7 patients (46.6%) with Mayo
ments evaluated by US), the overall agreement between subscore 2 presented hypervascularity and adenopathies,
US and colonoscopy was good (k=0.767, p<0.001, Se with no mesenteric inflammatory fat at US.
86.5%, Sp 90.1%) (details in Table II). Of the 27 patients
with US pathological findings in any of colonic seg- Discussion
ments, 23 had FC >250 mcg/g (85.1%).
The overall inter-observer agreement for the US In this retrospective study, we assessed the value of
measurements was excellent (ICC=0.926, p<0.001). US in evaluating UC paediatric patients during follow-
Analysing the discordances between US and colo- up, compared to FC and colonoscopy. We found that US
noscopy, we found that US detected normal findings, as was able to identify all patients with endoscopic active
 Med Ultrason 2021; 23(2): 153-160 157
Table II. Ultrasound and colonoscopy agreement for each bowel segment
Segment Ultrasound Colonoscopy k P Se (%) Sp (%)
Rectum 25 (83.3) 26 (86.7) 0.870 <0.001 96.2 100
Sigmoid colon 22 (73.3) 26 (86.7) 0.595 <0.001 84.6 100
Descending colon 15 (50.0) 17 (56.7) 0.600 0.001 76.5 84.6
Transverse colon 10 (33.3) 10 (33.3) 0.700 <0.001 80.0 90.0
Ascending colon 9 (30.0) 8 (26.7) 0.754 <0.001 87.5 90.9
Results are expressed as number (%); Se: sensibility; Sp: specificity; k: < 0 no agreement, 0–0.20 poor, 0.21–0.40 fair, 0.41–0.60 moderate,
0.61–0.80 good, and 0.81–1 very good agreement; p< 0.05: statistically significant.

Table III. Mayo endoscopic subscore and median of the bowel wall thickness for each bowel segment
Median BWT at US p
Mayo Score 0 1 2 3
Bowel segment
Rectum 2.3 (2.3-2.45) 3.45 (3.0-4.0) 4.0 (3.1-4.75) 3.65 (3.3-4.0) 0.043
Sigmoid colon 2.2 (2.2-2.35) 3.3 (2.5-3.8) 4.0 (3.3-4.75) 3.95 (3.4-4.5) 0.031
Descending colon 2.2 (2.1-2.2) 2.65 (2.5-3.8) 3.3 (2.5-4.25) 3.6 (3.5-3.7) 0.067
Transverse colon 2.4 (2.4-2.55) 2.6 (2.4-3.1) 2.5 (2.35-3.15) 3.15 (2.1-4.2) 0.978
Ascending colon 2.6 (2.3-2.7) 2.45 (2.3-3.5) 2.3 (2.2-2.9) 2.85 (2.2-3.5) 0.693
Results are expressed as median (range); BWT: bowel wall thickness; US: ultrasound; p< 0.05: statistically significant.

disease, at patient level, but there were some differences
between the identification of the affected bowel segments
by the two aforementioned methods. The discordances
between US and colonoscopy may occur, as described
in the study of Christensen et al [44], taking into account
that the mucosal healing is not in all cases uniform, prox-
imal to distal, but also in a patchy manner. Moreover, US
has the potential to examine all the layers of the bowel
wall and the extraintestinal features, whereas colonosco-
py is limited to the visualisation of the intestinal mucosa.
The US is recognised as first line imaging technique
in assessing patients with IBD [45,46]. Generally, B-
mode US and Doppler techniques are used and, in the
last years, elastography; the oral contrast is useful only
for small bowel investigation and intravenous contrast
has not been approved for paediatric population. The
performance of US in IBD paediatric patients was evalu-
ated in many studies, by comparing this method to co-
lonoscopy, histology or MRE. Barber el al [23] found
in CD paediatric patients a good concordance between
US and MRE findings, regarding the disease activity
and localisation. These authors demonstrated also high
specificity (the highest for sigmoid colon) of both imag-
ing techniques in identifying the active disease and the
histology confirmed disease at bowel segment level. A
significant learning curve was demonstrated for MRE,
with improvement of the results after 10 examinations,
but not for US. This suggests that, for experienced ra-
diologists, both methods can be used for follow-up.
The same good agreement between US and MRE was
obtained by Dillman et al [47] in children with CD. In
this retrospective study, the interobserver agreement was
best for the maximum BWT (95%CI 0.67 [0.64-0.70])
and relatively weak for the length of involved segment
(95%CI 0.41 [0.35-0.40]). In our UC patients, we found
an excellent agreement between radiologists, concerning
the measurement of the BWT. The discrepancy with the
aforementioned study could be related mainly to the un-
derlying disease. While in CD the bowel is inflamed in a
discontinuous and inhomogeneous manner (requiring the
choice of the same segment for measurement in order to
have a good interobserver agreement), in UC the disease
is continuous and involves the entire circumference of
the colon.
IBD cases may present different evolutive trends
[48]; therefore, these patients must be monitored for
long periods. The disease can progress to more extensive
forms, more frequently in paediatric patients compared
to adults (29.2 % vs 20.2%) [48]. In our study, 5 children
were diagnosed with pancolitis at onset and, during the
follow-up, 3 other patients were found to have pancolitis
at colonoscopy. US correctly identified all these three
cases. Due to the good agreement between US and co-
lonoscopy findings on patient and on segment level, we
concluded that US can be used not only for assessing the
favourable evolution of the disease but also to identify
flares and to establish the new extension of UC. This is
of utmost importance, as in these cases the colonoscopy
could be replaced by US, much easier accepted by chil-
dren and/or parents.
158 Ioana Fodor et al Following up paediatric patients with ulcerative colitis: abdominal US vs. colonoscopy and faecal calprotectin
Mucosal healing or endoscopic remission is the ideal
target for nowadays disease management strategy [49].
Clinical remission is considered when PUCAI <10 [50],
but approximately 20% of children in Turner et al [9] and
over 50% of children in Sarbagili-Shabat et al studies
[51] had endoscopic inflammation despite clinical inac-
tive disease. This category of patients, being asympto-
matic and probably with normal FC, cannot undergo co-
lonoscopy examination at every follow up visit. In these
cases, US can bring important information about the con-
dition of the colonic wall.
Kellar et al [52] developed a US activity score - the
simple paediatric ultrasound score (SPAUSS) - in order
to determine the most reliable parameter (mesenteric
inflammatory fat, mesenteric lymph nodes, hyperemia/
Doppler color flow and BWT) to predict bowel inflam-
mation. Out of these 4 parameters, BWT and mesenteric
inflammatory fat had significant prediction value for the
severity of the disease. In our study, all the children with
Mayo subscore of 3 presented mesenteric inflammatory
fat and BWT >3 mm in all analysed segments.
No definition/recommendation about US findings re-
garding the remission in UC was published. However,
as US can detect almost all the bowel segments, assess
the inflammation, measure the BWT and has good cor-
relation with colonoscopy findings, PUCAI and FC, the
technique has to be considered for future analysis of IBD
in children. US should be regarded as a valuable imaging
technique for UC evaluation.
The main limitations of our study are the retrospective
design, heterogeneity and the low number of patients. In
addition, we did not compare the US findings, especially
the power Doppler signal with the histology results. The
lack of comparison of the US aspect with other imag-
ing techniques, especially, MRE, is another limit of our
study. It would have been of interest to compare US data
at diagnosis and at the inclusion in the study, but due to
lack of information regarding the US data at the diagno-
sis in some patients, we could not realize this analysis.
Due to the retrospective design of our study, we did not
analyse whether the US findings should prompt the clini-
cian to reconsider the treatment options.
Conclusion
In conclusion, due to the good to excellent agreement
between abdominal US and colonoscopy findings on pa-
tient and bowel segment levels, US should be considered
as a useful imaging technique to assess and follow-up
UC patients. The good correlation between elevated FC
level and pathological US findings confirms the value of
US in detecting bowel inflammatory process. US might
replace the colonoscopy evaluation on follow-up, espe-
cially when PUCAI and FC suggest active disease.
Conflict of interest: none
References
1. Windsor JW, Kaplan GG. Evolving Epidemiology of
IBD. Curr Gastroenterol Rep 2019;21:40.
2. Ng SC, Shi HY, Hamidi N, et al. Worldwide incidence and
prevalence of inflammatory bowel disease in the 21st cen-
tury: a systematic review of population-based studies. Lan-
cet 2018;390:2769-2778.
3. Sýkora J, Pomahačová R, Kreslová M, Cvalínová D, Štych
P, Schwarz J. Current global trends in the incidence of pedi-
atric-onset inflammatory bowel disease. World J Gastroen-
terol 2018;24:2741-2763.
4. Cabrera JM, Sato TT. Medical and surgical management
of pediatric ulcerative colitis.  Clin Colon Rectal Surg
2018;31:71-79.
5. Yu YR, Rodriguez JR. Clinical presentation of Crohn’s,
ulcerative colitis, and indeterminate colitis: Symptoms, ex-
traintestinal manifestations, and disease phenotypes. Semin
Pediatr Surg 2017;26:349-355.
6. North American Society for Pediatric Gastroenterology,
Hepatology, and Nutrition; Colitis Foundation of America,
Bousvaros A, et al. Differentiating ulcerative colitis from
Crohn disease in children and young adults: report of a
working group of the North American Society for Pediat-
ric Gastroenterology, Hepatology, and Nutrition and the
Crohn’s and Colitis Foundation of America. J Pediatr Gas-
troenterol Nutr 2007;44:653-674.
7. Aloi M, Cucchiara S. Extradigestive manifestations of IBD
in pediatrics. Eur Rev Med Pharmacol Sci 2009;13 Suppl
1:23-32.
8. Dotson JL, Hyams JS, Markowitz J, et al. Extraintestinal
manifestations of pediatric inflammatory bowel disease and
their relation to disease type and severity. J Pediatr Gastro-
enterol Nutr 2010;51:140-145.
9. Turner D, Ruemmele FM, Orlanski-Meyer E, et al. Man-
agement of paediatric ulcerative colitis, part 1: ambulatory
care-an evidence-based guideline from European Crohn’s
and Colitis Organization and European Society of Paediat-
ric Gastroenterology, Hepatology and Nutrition.  J Pediatr
Gastroenterol Nutr 2018;67:257-291.
10. Greuter T, Bertoldo F, Rechner R, et al. Extraintestinal
manifestations of pediatric inflammatory bowel disease:
prevalence, presentation, and anti-TNF Treatment. J Pediatr
Gastroenterol Nutr 2017;65:200-206.
11. Levine A, Koletzko S, Turner D, et al. ESPGHAN revised
porto criteria for the diagnosis of inflammatory bowel dis-
ease in children and adolescents.  J Pediatr Gastroenterol
Nutr 2014;58:795-806.
12. Spiceland CM, Lodhia N. Endoscopy in inflammatory
bowel disease: Role in diagnosis, management, and treat-
ment. World J Gastroenterol 2018;24:4014-4020.
13. Van Limbergen J, Russell RK, Drummond HE, et al.
Definition of phenotypic characteristics of childhood-

onset inflammatory bowel disease.  Gastroenterology
2008;135:1114-1122.
14. DeRoche TC, Xiao SY, Liu X. Histological evaluation in
ulcerative colitis. Gastroenterol Rep (Oxf) 2014;2:178-192.
15. Oliva S, Thomson M, de Ridder L, et al. Endoscopy in
pediatric inflammatory bowel disease: a position paper on
behalf of the Porto IBD Group of the European Society for
Pediatric Gastroenterology, Hepatology and Nutrition.  J
Pediatr Gastroenterol Nutr 2018;67:414-430.
16. Thakkar K, El-Serag HB, Mattek N, Gilger M. Complica-
tions of pediatric colonoscopy: a five-year multicenter ex-
perience. Clin Gastroenterol Hepatol 2008;6:515–520.
17. Navaneethan U, Kochhar G, Phull H, et al. Severe disease
on endoscopy and steroid use increase the risk for bowel
perforation during colonoscopy in inflammatory bowel dis-
ease patients. J Crohns Colitis 2012;6:470-475.
18. Pathirana WGW, Chubb SP, Gillett MJ, Vasikaran SD. Fae-
cal calprotectin. Clin Biochem Rev 2018;39:77-90.
19. Mumolo MG, Bertani L, Ceccarelli L, et al. From bench to
bedside: Fecal calprotectin in inflammatory bowel diseases
clinical setting. World J Gastroenterol 2018;24:3681-3694.
20. D’Amico F, Bonovas S, Danese S, Peyrin-Biroulet L. Re-
view article: faecal calprotectin and histologic remission in
ulcerative colitis.  Aliment Pharmacol Ther 2020;51:689-
698.
21. D’Haens G, Ferrante M, Vermeire S, et al. Fecal calprotec-
tin is a surrogate marker for endoscopic lesions in inflam-
matory bowel disease. Inflamm Bowel Dis 2012;18:2218-
2224.
22. Yoon HM, Suh CH, Kim JR, et al. Diagnostic performance
of magnetic resonance enterography for detection of active
inflammation in children and adolescents with inflammato-
ry bowel disease: a systematic review and diagnostic meta-
analysis. JAMA Pediatr 2017;171:1208-1216.
23. Barber JL, Maclachlan J, Planche K, et al. There is good
agreement between MR enterography and bowel ultrasound
with regards to disease location and activity in paediatric
inflammatory bowel disease. Clin Radiol 2017;72:590-597.
24. Mollard BJ, Smith EA, Lai ME, Phan T, Christensen RE,
Dillman JR. MR enterography under the age of 10 years: a sin-
gle institutional experience. Pediatr Radiol 2016;46:43-49.
25. Manetta R, Capretti I, Belleggia N, et al. Magnetic reso-
nance enterography (MRE) and ultrasonography (US) in
the study of the small bowel in Crohn’s disease: state of the
art and review of the literature. Acta Biomed 2019;90:38-
50.
26. Spinelli A, Allocca M, Jovani M, Danese S. Review article:
optimal preparation for surgery in Crohn’s disease. Aliment
Pharmacol Ther 2014;40:1009-1022.
27. Elliott CL, Maclachlan J, Beal I. Paediatric bowel ul-
trasound in inflammatory bowel disease.  Eur J Radiol
2018;108:21-27.
28. Kucharzik T, Kannengiesser K, Petersen F. The use of ul-
trasound in inflammatory bowel disease. Ann Gastroenterol
2017;30:135-144.
29. Calabrese E, Maaser C, Zorzi F, et al. Bowel ultrasonogra-
phy in the management of Crohn’s Disease. A Review with
Med Ultrason 2021; 23(2): 153-160 159
Recommendations of an International Panel of Experts. In-
flamm Bowel Dis 2016;22:1168-1183.
30. van Wassenaer EA, de Voogd FAE, van Rijn RR, et al. Di-
agnostic accuracy of transabdominal ultrasound in detect-
ing intestinal inflammation in paediatric IBD patients-a sys-
tematic review. J Crohns Colitis 2019;13:1501-1509.
31. Barber JL, Zambrano-Perez A, Olsen ØE, et al. Detect-
ing inflammation in inflammatory bowel disease - how
does ultrasound compare to magnetic resonance enterog-
raphy using standardised scoring systems?  Pediatr Radiol
2018;48:843-851.
32. Civitelli F, Di Nardo G, Oliva S, et al. Ultrasonography
of the colon in pediatric ulcerative colitis: a prospective,
blind, comparative study with colonoscopy.  J Pediatr
2014;165:78-84.
33. Maaser C, Petersen F, Helwig U, et al. Intestinal ultrasound
for monitoring therapeutic response in patients with ul-
cerative colitis: results from the TRUST&UC study. Gut
2020;69:1629-1636.
34. Bryant RV, Friedman AB, Wright EK, et al. Gastrointestinal
ultrasound in inflammatory bowel disease: an underused re-
source with potential paradigm-changing application.  Gut
2018;67:973-985.
35. Rajagopalan A, Sathananthan D, An YK, et al. Gastrointes-
tinal ultrasound in inflammatory bowel disease care: Patient
perceptions and impact on disease-related knowledge. JGH
Open 2019;4:267-272.
36. Kucharzik T, Wilkens R, Maconi G, et al. DOP10 Intestinal
ultrasound response and transmural healing after 48 weeks
of treatment with Ustekinumab in Crohn’s disease: STAR-
DUST trial substudy. UEGW virtual 2020 LB12.
37. Kucharzik T, Helwig U, Seibold F, et al. Improvement
of intestinal ultrasound parameters within 12 weeks after
treatment induction in CD and UC patients - first interim
analysis of the trust beyond study. UEGW virtual 2020
P0430.
38. Levine A, Griffiths A, Markowitz J, et al. Pediatric modi-
fication of the Montreal classification for inflammatory
bowel disease: the Paris classification. Inflamm Bowel Dis
2011;17:1314-1321.
39. Cogley JR, O’Connor SC, Houshyar R, Al Dulaimy K.
Emergent pediatric US: what every radiologist should
know. Radiographics 2012;32:651-665.
40. Puylaert JB. Mesenteric adenitis and acute terminal ile-
itis: US evaluation using graded compression.  Radiology
1986;161:691-695.
41. Spalinger J, Patriquin H, Miron MC, et al. Doppler US in pa-
tients with crohn disease: vessel density in the diseased bow-
el reflects disease activity.  Radiology 2000;217:787-791.
42. Adamiak T, Altaf M, Jensen MK, et al. One-day bowel
preparation with polyethylene glycol 3350: an effective
regimen for colonoscopy in children. Gastrointest Endosc
2010;71:573-577.
43. Schroeder KW, Tremaine WJ, Ilstrup DM. Coated oral
5-aminosalicylic acid therapy for mildly to moderately ac-
tive ulcerative colitis. A randomized study. N Engl J Med
1987;317:1625-1629.
160 Ioana Fodor et al Following up paediatric patients with ulcerative colitis: abdominal US vs. colonoscopy and faecal calprotectin
44. Christensen B, Hanauer SB, Gibson PR, Turner JR, Hart
J, Rubin DT. Segmental histologic normalisation occurs in
ulcerative colitis but does not improve clinical outcomes. J
Crohns Colitis 2020;14:1345-1353.
45. Chiorean L, Schreiber-Dietrich D, Braden B, et al. Ultra-
sonographic imaging of inflammatory bowel disease in pedi-
atric patients.  World J Gastroenterol 2015;21:5231-5241.
46. Arndt H, Hauenstein C, Weber MA, Däbritz J, Bierwirth
C. Imaging of chronic inflammatory bowel diseases in
childhood and adolescence: Repetitorium. Radiologe
2020;60:1085-1096.
47. Dillman JR, Smith EA, Sanchez R, et al. Prospective co-
hort study of ultrasound-ultrasound and ultrasound-MR en-
terography agreement in the evaluation of pediatric small
bowel Crohn disease. Pediatr Radiol 2016;46:490-497.
48. Roda G, Narula N, Pinotti R, et al. Systematic review with
meta-analysis: proximal disease extension in limited ulcer-
ative colitis. Aliment Pharmacol Ther 2017;45:1481-1492.
49. Kim S, Park S, Kang Y, Koh H. Combining faecal calpro-
tectin and sigmoidoscopy can predict mucosal healing in
paediatric ulcerative colitis.  Eur J Gastroenterol Hepatol
2020;32:17-21.
50. Turner D, Otley AR, Mack D, et al. Development, valida-
tion, and evaluation of a pediatric ulcerative colitis activity
index: a prospective multicenter study.  Gastroenterology
2007;133:423-432.
51. Sarbagili-Shabat C, Weiner D, Wardi J, Abramas L, Yaakov
M, Levine A. Moderate to severe endoscopic inflammation
is frequent after clinical remission in pediatric ulcerative
colitis. J Pediatr Gastroenterol Nutr 2020. doi:10.1097/
MPG.0000000000003018.
52. Kellar A, Wilson S, Kaplan G, DeBruyn J, Tanyingoh D,
Novak KL. The Simple Pediatric Activity Ultrasound
Score (SPAUSS) for the accurate detection of pediatric
inflammatory bowel disease.  J Pediatr Gastroenterol Nutr
2019;69:e1-e6.
Original papers Med Ultrason 2021, Vol. 23, no. 2, 161-167
DOI: 10.11152/mu-2827

Microvascular Doppler ultrasound in children with acute

pyelonephritis
Gayoung Choi1, Bo-Kyung Je1, Doran Hong1, Jaehyung Cha2

1Department of Radiology, Korea University College of Medicine, Ansan Hospital, Gyeonggi, 2Medical Science
Research Center, Korea University College of Medicine, Seoul, Republic of Korea

Abstract
Aims: To compare the diagnostic performance of microvascular Doppler ultrasonography (MVUS) to B-mode and con-
ventional colour Doppler US (CDUS) for detecting acute pyelonephritis (APN) lesions in children. Material and methods:
An IRB-approved retrospective study was performed. From July 2018 to January 2019, 41 APN lesions in 28 children (15
boys, 13 girls; age range, 1-196 months; mean age, 53 months) who underwent 99mTc‒dimercaptosuccinic acid renal scintig-
raphy (DMSA) or contrast-enhanced computed tomography (CECT) and US including B-mode, CDUS, and MVUS were
enrolled in this study. Three paediatric radiologists independently reviewed the B-mode, CDUS and MVUS images for the
DMSA or CECT-proven APN lesions and evaluated the lesion visibility, lesion distinguishability and diagnostic confidence
between the MVUS and CDUS images. Results: A total 41 of APN lesions were verified by DMSA (41 lesions) or CECT (3
lesions) during the same hospitalization period with renal US. Among 41 APN lesions, 52.8% was visible on B-mode, 85.4%
on CDUS, and 94.3% on MVUS (p<0.001). Comparing the extent and margins of the lesions, MVUS had better results than
CDUS in 41.5% of the lesions, CDUS had better results in 6.5% and they were equal in 52% (p<0.001). The diagnostic con-
fidence of the APN lesions was higher for MVUS than CDUS in 36.6%, higher for CDUS than MVUS in 4.9%, and equal in
the remaining 58.5% (p<0.05). The interobserver agreement was fair to moderate. Conclusions: MVUS showed improved
detectability of hypoperfused areas in paediatric APN and provided higher diagnostic confidence.
Keywords: pyelonephritis; paediatrics; ultrasonography; colour Doppler; microvascular imaging

Introduction aging studies are required [1,2]. For diagnosing APN

in paediatric patients, ultrasonography (US), magnetic
Acute pyelonephritis (APN) in children requires an resonance imaging (MRI), renal cortical scintigraphy
immediate diagnosis for the appropriate treatment, in with 99mTc‒dimercaptosuccinic acid (DMSA) scans and
order to prevent complications, including irreversible contrast-enhanced computed tomography (CECT) are
scarring, renal hypertension and chronic renal failure. available [3,4].
The diagnosis of APN in children, based on clinical and Hypoperfusion is one of the major processes occur-
laboratory findings, is often challenging. Therefore, im- ring during the development of APN, owing to cortical
arteriolar vasoconstriction with inflammatory cells clog-
Received 15.09.2020  Accepted 25.01.2021 ging the peritubular capillary and swelling of the renal
Med Ultrason interstitium after bacterial infection [5]. Depending on
2021, Vol. 23, No 2, 161-167
Corresponding author: Bo-Kyung Je, MD, PhD
the superb detectability of the hypoperfused areas in the
Department of Radiology, DMSA scans and CECT, these two modalities showed
Korea University Hospital the highest sensitivity and specificity in diagnosing APN
123 Jeokgeumro, Danwongu, Ansan, [1]. However, they have disadvantages for use in paedi-
Gyeonggi 15355 Republic of Korea
Phone: 82-31-4125229
atric patients, such as radiation exposure and invasive-
Fax: 82-31-4125224 ness that requires an intravenous injection. MRI is also
E-mail: radje@korea.ac.kr sensitive and specific. However, it involves paediatric
162 Gayoung Choi et al Microvascular Doppler ultrasound in children with acute pyelonephritis
patient sedation, long examination times and high costs.
In comparison, US has many advantages as the primary
diagnostic modality for paediatric APN, such as an easy
application, good cost-effectiveness and safety not re-
quiring radiation exposure [6]. In spite of these advan-
tages of US, the diagnostic sensitivity of US has been
reported as 20 to 69%, much lower compared with the
other techniques [1,7,8].
However, US technology has evolved and new emerg-
ing techniques have improved the diagnostic performance
of US. The sensitivity and specificity of conventional col-
our Doppler US (CDUS) for detecting hypoperfused ar-
eas have also been improved. Recently, a new innovative
Doppler technique was introduced; the technique can de-
pict tissue microvessels with slow blood flow that are re-
moved along with clutter artifacts by conventional colour
Doppler filters [9,10]. This microvascular Doppler US
(MVUS) is a promising method to demonstrate tissue mi-
crovasculature and distinguish vascular lesions [10,11].
To our knowledge, there has been no published study
on the diagnostic performance of MVUS in detecting
APN lesions compared to CDUS. In this study, we ret-
rospectively reviewed the US features of 41 APN lesions
that were confirmed by DMSA scans or CECT. The pur-
poses of this study were to evaluate the diagnostic capa-
bilities of B-mode, CDUS and MVUS images for pae-
diatric APN and compare the diagnostic performance of
MVUS and CDUS techniques.
Materials and methods
Subjects
The Institutional Review Board approved this retro-
spective study and waived the requirement for informed
consent. From July 2018 to January 2019, 424 children
underwent renal US at our institution. With the exclu-
sion of renal US requested for non-infectious reasons and
follow-up renal US for previous urinary tract infections
(UTI), a total of 123 renal US examinations were per-
formed for the suspicion of a first febrile UTI episode.
Among them, the patients who underwent DMSA scans
or CECT during the same hospitalization period with re-
nal US were selected. Then, the patients with renal per-
fusion defects on DMSA scans or CECT were included.
Finally, 41 APN lesions in 28 children (15 boys, 13 girls;
age range, 1-196 months; mean age, 53 months) were en-
rolled (fig 1).
Imaging studies
A paediatric radiologist with 15 years of experience
performed the renal US examinations with a commer-
cially available US scanner (RS 85, Samsung Medison,
Seoul, Korea), equipped with linear (4-15 MHz) and
Fig 1. Enrolment of subjects. UTI: urinary tract infection;
DMSA scan: 99mTc-dimercaptosuccinic acid scan; CECT: con-
trast-enhanced CT including the kidney
convex transducers (1-7 MHz). The MVUS examination
(called MV-FlowTM) was performed with these transduc-
ers. All renal US examinations included B-mode, CDUS
and MVUS images in both the longitudinal and trans-
verse axes of each kidney, obtained in the same area. The
three scanning modes were readily switchable by press-
ing buttons on the scanner control panel. The depth and
focal zones were identical and other scanning parameters
were applied as provided. No sedatives or other invasive
interventions were given, but some children with irrita-
bility were helped by extra-assistance including video
watching during the US examinations.
A DMSA scan was performed 2 hours after intra-
venous injection of 99mTc-labeled DMSA in the routine
protocol (1-2 mCi according to body weight) using a
tomographic gamma camera (Infinia, GE Healthcare,
Milwaukee, WI, USA), which included anterior, poste-
rior, and both posterior oblique images. All patients were
asleep with oral chloral hydrate (PocralR syrup; Hanlim,
Korea) during the examination.
A CECT scan was obtained 70 seconds after the injec-
tion of intravenous iodine contrast media with 128-slice
spectral CT (IQon Spectral CT, Philips Healthcare,
Cleveland, OH, USA). The typical kVp of CECT was
determined according to the body weight of each patient:
80kVp for children under 10kg, 100kVp for 10-30kg,
120kvP for over 30kg. Three patients underwent CECT
and none required sedation.
Evaluation of images
DMSA scans and CECT were used as the verifying
modalities for the APN lesions. A certified nuclear medi-
cine specialist and a paediatric radiologist interpreted the

DMSA scans and CECT, respectively, and recorded the
lesion site (right or left) and location (upper pole, mid-
portion, or lower pole). If the patients underwent voiding
cystourethrography (VCU) during the same hospitaliza-
tion period, the results were also recorded.
Three certified radiologists (the first reader with 5
years of experience, the second reader with 10 years,
and the third reader with 15 years) reviewed the B-mode,
CDUS and MVUS images of the renal US examinations
obtained in the same area of each kidney. The results
of DMSA scans and CECT were given to three readers
along with renal US images; however, the patients’ in-
formation was blinded. The first and second readers had
no experience with MVUS. They evaluated the following
features independently without consensus: a) Lesion vis-
ibility on B-mode, CDUS, and MVUS: three radiologists
decided whether an APN lesion was visible or not on
each US image in three modes. They defined the US fea-
tures of APN as follows: (1) a change in renal parenchy-
mal echogenicity, (2) swelling of the renal parenchyma,
(3) loss of renal corticomedullary differentiation, and (4)
an area of hypoperfusion; b) Lesion distinguishability on
the Doppler study: the radiologists compared CDUS and
MVUS to evaluate which image better defined the extent
and margin of the APN lesions; c) Diagnostic confidence
in the Doppler study: the radiologists compared CDUS
and MVUS to evaluate which image provided more di-
agnostic confidence for APN.
Statistical analyses
Lesion visibility: the results of three readers’ interpre-
tations were summated and the average lesion visibility
was calculated. Then, the Cochran’s Q test was used to
analyse the statistical significance between the results of
the three US modes.
Lesion distinguishability and diagnostic confidence
in the Doppler study: the results of the three readers’
interpretations were summated. Then, an R program
coded-proportion test was used to analyse the statistical
significance of the results.
Interobserver agreement: As three radiologists evalu-
ated the US images independently, the interobserver
agreement between them was calculated by Fleiss Kap-
pa, and the strength was interpreted as follows: poor
agreement for kappa values below zero, slight for kappa
values of 0.01 ‒ 0.20, fair for kappa values of 0.21 ‒ 0.40,
moderate for kappa values of 0.41 ‒ 0.60, substantial for
kappa values of 0.61 ‒ 0.81, and almost perfect agree-
ment for kappa values of 0.81 ‒ 1.00.
The statistical analyses were performed with the R
program (version 3.6.1; package RVAideMemoire for
Cochran Q test; package irr for Fleiss Kappa). Statistical
significance was defined as a p value of <0.05.
Med Ultrason 2021; 23(2): 161-167 163
Results
All patients were admitted for fever. Among 41 APN
lesions, all lesions were verified by DMSA scans and
three lesions were verified by CECT as well. US exami-
nations were performed on the admission day for five le-
sions, the day after admission for 28 lesions, two days
after admission for six lesions and four and five days later
in the remaining two lesions. Most of the DMSA scans
and CECTs were performed after US, except in one pa-
tient where the DSMA scan and CECT were performed
on the day of admission and the US examination was
done five days later. The average interval was 1.4 days
between US and DMSA in 41 lesions, and -1.7 days be-
tween US and CECT in 3 lesions.
The site and location of the 41 APN lesions were as
follows: the upper pole was most affected with 30 lesions
(73.2%, 14 in the right kidney, 16 in the left kidney), fol-
lowed by 9 lesions in the lower pole (21.9%, 2 in the
right, 7 in the left) and 2 lesions in the mid-portion (4.9%,
1 in each). The left kidney was more affected than the
right kidney (58.5% vs. 41.5%).
VCU was performed in 37 (90.2%) out of 41 APN
lesions, which showed ipsilateral vesicoureteral reflux in
10 lesions.
Lesion visibility
Among 41 APN lesions, 52.8% were visible on B-
mode, 85.4% on CDUS and 94.3% on MVUS. Cochran’s
Q test analysed the statistical significance between the
results in the three US modes and the lesion visibility
was superior in MVUS (p< 0.001) (Table I, fig 2, fig 3).
Lesion distinguishability
MVUS was better at defining the extent and margins
of the APN lesions than CDUS in 41.5% of the APN le-
sions. CDUS was better than MVUS in 6.5% of the le-
sions. The two modes showed equal distinguishability in
Table I. Result of Cochran’s Q test for lesion visibility
Comparison B-mode B-mode CDUS
vs. CDUS vs. MVUS vs. MVUS
χ2 34.8 47.3 3.37
df 1 1 1
adjusted p value <0.001 <0.001 <0.001
CDUS: colour Doppler US; MVUS: microvascular Doppler US
Table II. Result of proportion test for lesion distinguishability
Comparison CDUS > MVUS CDUS = MVUS
MVUS > CDUS <0.001 0.35
MVUS = CDUS <0.001 ―
CDUS: colour Doppler US, MVUS: microvascular Doppler US.
Inequality signs mean superiority of lesion distinguishability.
164 Gayoung Choi et al Microvascular Doppler ultrasound in children with acute pyelonephritis
Fig 2. A 12-year-old girl with febrile UTI. (A) DMSA scan
confirmed APN in the upper (arrow) and lower poles of the left
kidney. B-mode (B), colour Doppler (C), and microvascular
Doppler (D) for the upper pole lesion. All three readers inter-
preted an APN lesion seen on C and D, but not B. All of them
reported that the lesion distinguishability and diagnostic confi-
dence between C and D were equal.
Fig 3. A 5-month-old girl with febrile UTI. (A) DMSA scan
confirmed APN in the upper pole of the right kidney (arrow).
B-mode (B), colour Doppler (C), and microvascular Doppler
(D) for the upper pole lesion. All three readers interpreted an
APN lesion seen on the B, C, and D. One of them reported that
D showed better lesion distinguishability and more diagnostic
confidence than C, but two readers reported that the lesion dis-
tinguishability and diagnostic confidence were equal in both C
and D.
52% of the lesions. The R program coded-proportion test
concluded that MVUS was superior to CDUS in distin-
guishing APN lesions (p<0.001) (Table II, fig 4).
Diagnostic confidence
The diagnostic confidence was higher for MVUS
than CDUS in 36.6% of APN lesions, higher for CDUS
than MVUS in 4.9%, and equal in the remaining 58.5%.
Although the diagnostic confidence between CDUS
and MVUS was equal in 56.1% of the 41 APN lesions,
MVUS was preferred in 36.6%, whereas CDUS was
preferred in only 4.9% of the APN lesions, meaning that
the three readers preferred MVUS over 7 times more
when the two modes of Doppler images were provided
simultaneously. The R program coded-proportion test
concluded that MVUS provided more confidence than
CDUS for the diagnosis of paediatric APN (p<0.001 for
MVUS>CDUS, p=0.02 for MVUS=CDUS) (Table III,
fig 5).
Interobserver agreements
The results of the Fleiss Kappa analysis of the three
readers’ agreement are as follows: moderate agreement
for lesion visibility in B-mode (0.576) and CDUS (0.414),
slight agreement in MVUS (0.091), fair agreement for le-
sion distinguishability (0.324), and slight agreement for
confidence of diagnosis (0.118).
Fig 4. A 5-month-old boy with febrile UTI. (A) DMSA scan
confirmed APN in the upper pole of the right kidney (arrow). B-
mode (B), colour Doppler (C), and microvascular Doppler (D)
for the upper pole lesion. Two readers interpreted an APN le-
sion seen on C and D, and one reader interpreted an APN lesion
seen on B, C, and D. All of them reported better lesion distin-
guishability on D than C. Two of them reported that D provided
more diagnostic confidence than C, and one of them reported
that the diagnostic confidence as equal between C and D.

Table III. Result of proportion test for diagnostic confidence
Comparison CDUS > MVUS CDUS = MVUS
MVUS > CDUS <0.001 0.02
MVUS = CDUS <0.001 ―
CDUS: colour Doppler US, MVUS: microvascular Doppler US.
Inequality signs mean superiority of diagnostic confidence.
Fig 5. A 17-month-old boy with febrile UTI. (A) DMSA scan
confirmed APN in the mid-portion of the right kidney (arrow).
B-mode (B), colour Doppler (C), and microvascular Doppler
(D) for the mid-portion lesion. All three readers interpreted an
APN lesion seen on B, C, and D. Two of them reported that
D showed better lesion distinguishability than C, and one of
them reported as equal. Two of them reported that D provided
more diagnostic confidence than C, and one of them reported
as equal.
Discussion
The US findings of APN are changes in renal paren-
chymal echogenicity, swelling of the renal parenchyma,
loss of renal corticomedullary differentiation, urothelial
thickening of the renal pelvis and ureter and an area of
decreased perfusion, which may involve the kidneys
focally or diffusely [8,12]. As US can also provide a
structural overview of the renal anatomy and rule out hy-
dronephrosis, abscess or calculus, it is worthwhile as a
primary imaging modality for APN in children [13,14].
However, these US findings were not always well-char-
acterized, and a number of papers have reported better
US techniques for diagnosing paediatric APN including
B-mode, colour Doppler, power Doppler and contrast-
enhanced US [2,15‒18]. Power Doppler US (PDUS) has
been tried for the past 14 years as a method for assessing
more renal vascular structures and achieving better sen-
Med Ultrason 2021; 23(2): 161-167 165
sitivity for detecting renal blood flow than CDUS. There-
fore, PDUS showed superior detectability of APN repre-
sented as hypoperfused areas in the kidneys. Recently,
a third-generation Doppler US technique, MVUS, has
emerged to depict spatially small and temporally slow
blood flow using low pulse repetition frequency with an
advanced clutter filter that can differentiate microvascu-
lar flow from clutter artifacts [10,19]. Despite the relative
novelty of microvascular imaging, Some studies already
have reported its strength and clinical benefits in various
organs including liver, breast, thyroid gland, ovary, tes-
tis, brain, kidney, GI tract, lymph node, extremities and
carotid plaques [10,11,20,21].
In kidneys, MVUS has shown the superb ability to
visualize more detailed branching patterns of interlo-
bar, arcuate and interlobular arteries to the renal cortex,
compared to CDUS or PDUS [10,22,23]. Additionally,
contrast-enhanced US has emerged as a promising alter-
native to DMSA with high sensitivity and specificity for
diagnosing APN, yet it still requires intravenous admin-
istration of US contrast media and not enough studies re-
ported for clinical application in children [24].
In this study, we evaluated APN lesion visibility using
MVUS technique called MV-FlowTM and compared it to
B-mode and CDUS. We found that MVUS showed bet-
ter visibility of APN lesions than B-mode and CDUS and
provided better depiction and better distinguishability of
minute hypoperfused areas compared to CDUS. Since
the main pathophysiology of APN is hypoperfusion at
the level of microvessels including cortical arteriolar
vasoconstriction and peritubular capillary occlusion by
inflammatory cells, imaging of perfusion by minute ves-
sels using advanced techniques such as MVUS are help-
ful in the diagnosis of APN. Thus, this study revealed
that the feasibility of MVUS to detect APN lesions was
superior to that of CDUS. Further studies with MVUS
could extend its clinical value for many vascular indica-
tions, including renal infarction, renal cortical necrosis,
renal tumours, chronic renal disease, acute tubular necro-
sis, renal transplantation, as well as acute pyelonephritis.
US examination is recognized as an operator-de-
pendent technique. The artifacts or factors degrading
the quality of paediatric US, such as motion, crying and
bowel gas could be improved according to the radiolo-
gists’ experience and know‒how. To minimize operator-
dependent issues in this study, the most experienced ra-
diologist performed all US scans including the B-mode,
CDUS and MVUS scans. Despite our expectations, the
Fleiss Kappa values were all less than 0.6, which meant
the interobserver agreement was not perfect. For exam-
ple, the lesion visibility in MVUS showed the highest re-
sult (94.3%). However, the kappa value among the three
166 Gayoung Choi et al Microvascular Doppler ultrasound in children with acute pyelonephritis
readers was the lowest (0.091). We suggest that these
conflicting results were due to the lack of experience
with MVUS in two readers and the independent interpre-
tation of the three readers without consensus. The highest
kappa value was seen when the readers interpreted the B-
mode images, which were the basic and most familiar US
images to all three radiologists. We could expect that the
agreement would have been higher if more of the readers
were familiar with MVUS images.
There were several limitations to our study. First, the
data were retrospectively obtained. Therefore, we could
only enrol a small number of patients. Second, the sec-
ond-generation Doppler US technique, PDUS, was not
obtained in this study. The advantage of MVUS would
be reinforced if a comparison between PDUS and MVUS
was presented. Third, we did not enrol the patients af-
fected by UTI without APN lesion. Fourth, all US ex-
aminations in our study were performed by one paediat-
ric radiologist. We expect that further studies performed
by more radiologists aware of MVUS may show better
agreement between the observers.
In conclusion, MVUS showed improved detectabil-
ity of hypoperfused areas in paediatric APN compared to
CDUS and increased US diagnostic performance of APN
without radiation exposure. The more sensitive diagnosis
of childhood APN is feasible with the new Doppler US
technique, MVUS. Confident, early diagnosis of APN
using MVUS may help reduce additional radiation expo-
sure and also enable proper treatment to prevent chronic
sequelae of the kidneys.
Acknowledgments: This study was supported by the
Korea University (K1810931).
Conflict of interest: none
References
1. Majd M, Nussbaum Blask AR, Markle BM, et al. Acute py-
elonephritis: comparison of diagnosis with 99mTc‒DMSA,
SPECT, spiral CT, MR imaging, and power Doppler US in
an experimental pig model. Radiology 2001;218:101-108.
2. Stogianni A, Nikolopoulos P, Oikonomou I, et al. Child-
hood acute pyelonephritis: comparison of power Doppler
sonography and Tc‒DMSA scintigraphy. Pediatr Radiol
2007;37:685-690.
3. Vourganti S, Agarwal PK, Bodner DR, Dogra VS. Ultra-
sonographic evaluation of renal infections. Radiol Clin
North Am 2006;44:763-775.
4. Das CJ, Ahmad Z, Sharma S, Gupta AK. Multimodal-
ity imaging of renal inflammatory lesions. World J Radiol
2014;6:865-873.
5. Ha SK, Seo JK, Kim SJ, et al. Acute pyelonephritis focus-
ing on perfusion defects on contrast enhanced computer-
ized tomography(CT) scans and its clinical outcome. Ko-
rean J Intern Med 1997;12:122-127.
6. Hwang M, Piskunowicz M, Darge K. Advanced Ul-
trasound Techniques for Pediatric Imaging. Pediatrics
2019;143:e20182609.
7. Lavocat MP, Granjon D, Allard D, Gay C, Freycon MT,
Dubois F. Imaging of pyelonephritis. Pediatr Radiol
1997;27:159-165.
8. Craig WD, Wagner BJ, Travis MD. Pyelonephritis: radio-
logic‒pathologic review. Radiographics 2008;28:255-277.
9. Park AY, Seo BK. Up-to-date Doppler techniques for breast
tumor vascularity: superb microvascular imaging and con-
trast‒enhanced ultrasound. Ultrasonography 2018;37:98-
106.
10. Yoo J, Je BK, Choo JY. Ultrasonographic Demonstration
of the Tissue Microvasculature in Children: Microvascular
Ultrasonography Versus Conventional Color Doppler Ul-
trasonography. Korean J Radiol 2020;21:146-158.
11. 11. Jiang ZZ, Huang YH, Shen HL, Liu XT. Clinical appli-
cations of superb microvascular imaging in the liver, breast,
thyroid, skeletal muscle, and carotid plaques. J Ultrasound
Med 2019;38:2811-2820.
12. Chung EM, Soderlund KA, Fagen KE. Imaging of the Pedi-
atric Urinary System. Radiol Clin North Am 2017;55:337-
357.
13. Subcommittee on Urinary Tract Infection. Steering Com-
mittee on Quality Improvement and Management. Rob-
erts KB. Urinary tract infection: clinical practice guide-
line for the diagnosis and management of the initial UTI
in febrile infants and children 2 to 24 months. Pediatrics
2011;128:595-610.
14. Bae HJ, Park YH, Cho JH, Jang KM. Comparison of
99mTc‒DMSA Renal Scan and Power Doppler Ultra-
sonography for the Detection of Acute Pyelonephritis and
Vesicoureteral Reflux. Child Kidney Dis 2018;22:47-51.
15. Dacher JN, Pfister C, Monroc M, Eurin D, LeDosseur P.
Power Doppler sonographic pattern of acute pyelonephri-
tis in children: comparison with CT. AJR Am J Roentgenol
1996;166:1451-1455.
16. Sakarya ME, Arslan H, Erkoc R, Bozkurt M, Atilla MK.
The role of power Doppler ultrasonography in the diagnosis
of acute pyelonephritis. Br J Urol 1998;81:360-363.
17. el Hajjar M, Launay S, Hossein‒Foucher C, Foulard M,
Robert Y. Power Doppler sonography and acute pyelone-
phritis in children: comparison with Tc‒DMSA scintigra-
phy. Arch Pediatr 2002;9:21-25.
18. Bykov S, Chervinsky L, Smolkin V, Halevi R, Garty I. Pow-
er Doppler sonography versus Tc‒99m DMSA scintigraphy
for diagnosing acute pyelonephritis in children: are these
two methods comparable? Clin Nucl Med 2003;28:198-203.
19. Jung HK, Park AY, Ko KH, Koh J. Comparison of the diag-
nostic performance of power Doppler ultrasound and a new
microvascular Doppler ultrasound technique (AngioPlus)
for differentiating benign and malignant breast masses. J
Ultrasound Med 2018;37:2689-2698.
20. Ohno Y, Fujimoto T, Shibata Y. A new era in diagnostic
ultrasound, superb microvascular imaging: preliminary

results in pediatric hepato-gastrointestinal disorders. Eur J
Pediatr Surg 2017;27:20-25.
21. Ayaz E, Aslan A, İnan İ, Yıkılmaz A. Evaluation of Ovar-
ian Vascularity in Children by Using the “Superb Micro-
vascular Imaging” Ultrasound Technique in Comparison
With Conventional Doppler Ultrasound Techniques. J Ul-
trasound Med 2019;38:2751-2760.
22. Quaia E, Correas JM, Mehta M, Murchison JT, Gennari
AG, van Beek EJR. Gray scale ultrasound, color Doppler
Med Ultrason 2021; 23(2): 161-167 167
ultrasound, and contrast-enhanced ultrasound in renal pa-
renchymal diseases. Ultrasound Q 2018;34:250-267.
23. Correas JM, Anglicheau D, Joly D, Gennisson JL, Tanter
M, Helenon O. Ultrasound-based imaging methods of the
kidney-recent developments. Kiney Int 2016;90:1199-
1210.
24. Jung HJ, Choi MH, Pai KS, Kim HG. Diagnostic perfor-
mance of contrast-enhanced ultrasound for acute pyelone-
phritis in children. Sci Rep 2020;10:10715.
Original papers Med Ultrason 2021, Vol. 23, no. 2, 168-175
DOI: 10.11152/mu-2732

Comparison of IOTA three-step strategy and logistic regression

model LR2 for discriminating between benign and malignant
adnexal masses
Juan José Hidalgo1,2, Antoni Llueca3, Irene Zolfaroli2, Nadia Veiga4, Ester Ortiz4, Juan Luis
Alcázar5

1Department of Obstetrics and Gynaecology, Hospital Comarcal Universitario de Vinaros, Castellón, 2Department of
Obstetrics and Gynaecology, Hospital Clínico Universitario de Valencia, 3Department of Medicine. University Jaume
I, Castellón, 4Department of Obstetrics and Gynaecology, Hospital Peset, Valencia, 5Department of Obstetrics and
Gynaecology, Clínica Universidad de Navarra, University of Navarra, Pamplona, Spain

Abstract
Aims: To compare the diagnostic performance of two ultrasound-based diagnostic systems for the classification of benign
or malignant adnexal masses, the three-step strategy and the predictive logistic regression model LR2, both proposed by the
International Ovarian Tumour Analysis (IOTA) Group. Material and methods: Prospective observational study at a single
centre that included patients diagnosed with a persistent adnexal mass by transvaginal ultrasound over a period of two years.
They were evaluated by a non-expert sonographer by applying the three-step diagnostic strategy and the LR2 predictive model
to classify the masses as benign or malignant. Patients were treated surgically or followed up for at least one year, taking as
the standard reference for benignity or malignancy the histological diagnosis of the lesion or ultrasound changes suggestive of
malignancy during the follow-up period. Sensitivity, specificity, positive and negative likelihood ratios and overall accuracy
of both systems was calculated and compared. Results: One hundred patients were included, with a mean age of 50.6 years
(range 18-87). Surgery was performed on 62 (62%) patients and 38 (38%) were managed expectantly. Eighty-three (83%)
lesions were benign and 17 (17%) were malignant. The IOTA three-step strategy presented sensitivity of 94.1% (95%CI,
86.7-98.3%) and specificity 97.6% (95%CI, 94.8-99%). The LR2 logistic regression model showed sensitivity 94.1% (95%CI,
73-98.9%) and specificity 81.9% (95%CI 72.3-88.7%). Comparison of the two systems showed a statistically significant dif-
ference in specificity in favour of the three-step strategy. Conclusions: The IOTA three-step strategy, in addition to being sim-
ple to use in clinical practice, has a high diagnostic accuracy for the classification of benignity and malignancy of the adnexal
masses, overtaking that of other predictive models such as the LR2 logistic regression model.
Keywords: adnexal masses; diagnosis; ultrasound; three-step strategy; LR2

Introduction Lesions with suspected malignancy must be referred to

specialized gynaecological oncology centres because the
An accurate initial diagnosis of benign or malignant correct initial surgical staging and cytoreduction surgery
ovarian masses is essential to guide the treatment [1]. are some of the most important prognostic factors [2-6].
Subjective ultrasound assessment by experts has
Received 16.07.2020  Accepted 11.11.2020 proven to be the most accurate method for character-
Med Ultrason
2021, Vol. 23, No 2, 168-175
izing adnexal masses [7-9], with sensitivity of 88-98%
Corresponding author: Antoni Llueca, MD and specificity of 89-96% to establish their malignancy
Department of Medicine. probability [10,11]. Nevertheless, it presents some limi-
University Jaume I (UJI) tations related to the subjectivity, experience and skill of
Av. de Vicent Sos Baynat s/n, 12071,
Castellón de la Plana, Spain
the sonographer, showing limited diagnostic accuracy in
E-mail: antonillueca@gmail.com less experienced professionals [12]. This is the reason
Phone: +34 964387440 why objective predictive models have been proposed to

help non-expert examiners to achieve similar results to
experts [13]. To date, none of them has been generally
accepted in clinical practice nor have they been shown
to improve the results obtained by the subjective assess-
ment of experts [14,15].
One of these models, named clinically the oriented
three-step strategy, was proposed by the International
Ovarian Tumour Analysis (IOTA) Group in 2012, show-
ing better results than previous models and a closer ap-
proach to clinical reality [16]. Five studies have been
reported performing an external validation of this ap-
proach, showing a sensitivity and specificity of 87.5-
95.2% and 87.6-100%, respectively [17-21]. With this
strategy, adnexal lesions are evaluated in three consecu-
tive steps. The first step evaluates six simple variables or
descriptors (four for benignity and two for malignancy)
of an immediate application. If none are applicable or
both benign and malignant can be applied, the lesion is
considered non-qualifiable and must be evaluated by a
second step according to the Simple Rules. These first
two steps can be performed by non-expert sonographers
with basic training. If the diagnosis cannot be made by
the second step, a third step should be applied consisting
of subjective evaluation by an expert sonographer [16].
Another predictive system is the logistic regression
model LR2, one of the most evaluated and most widely
used adnexal lesion diagnostic systems in clinical prac-
tice, proposed by IOTA in 2005 [22]. This model esti-
mates the individual probability of risk of malignancy of
an adnexal mass applying to a logistic regression formula
the result of six predictor variables. A probability result
greater than 0.1 (10%) implies that the lesion should be
considered malignant. Some studies have validated this
model, showing that sensitivity for the diagnosis of ma-
lignancy is above 90% [13,22,23]. Furthermore, a sys-
tematic review and meta-analysis in 2014 found this to
be one of the most accurate predictive models among the
19 evaluated [15].
To our knowledge, there are no published studies to
date comparing the clinically oriented three-step strategy
with other predictive models. In our study, our objective
is to compare the diagnostic accuracy of the three-step
strategy with that of the LR2 model in the same group
of patients.
Materials and methods
Design of the study
An observational, prospective study was performed
in a single centre (University District Hospital of Vina-
ros, Castellón, Spain) from September 2015 to August
2017.
Med Ultrason 2021; 23(2): 168-175 169
Patients
Patients were eligible if they were over the age of 18
and were diagnosed with at least one adnexal mass by
transvaginal or transrectal ultrasound. To remain in the
study, they should have undergone surgery within three
months after diagnosis or complete clinical and ultra-
sound follow-up for at least 12 months.
In the case of bilateral masses, the most complex
or the largest (in case of morphological similarity) was
selected. Patients who were pregnant during the study
period were excluded. Likewise, patients who could not
complete the follow-up visits or had a history or pres-
ence of a gynaecological neoplastic process were also
excluded.
The Medical Ethics Committee of the centre approved
the study. All patients received written and verbal infor-
mation and signed their informed consent to participate.
The patients included in the study constitute a subgroup
within a more extensive study for the external validation
of the three-step ultrasound strategy [21].
Ultrasound evaluation
The patients in the study were initially evaluated by
a non-expert sonographer (J.J.H.), with a Level-2 train-
ing according to the European Federation of Societies for
Ultrasound in Medicine and Biology (EFSUMB) for gy-
naecological ultrasound [24].
The ultrasound study was performed using transvagi-
nal or transrectal B-mode ultrasound and 2D Power Dop-
pler according to the IOTA terms, definitions and meas-
urements [25]. Large lesions that could not be entirely
evaluated by vaginal ultrasound were also evaluated ab-
dominally. At least ten ultrasound images and two video
fragments were electronically stored for each mass. The
scans were performed with a Voluson S6® ultrasound
system with RIC 5-9 MHz (GE Health care Ultrasound,
Milwaukee, WI, USA).
Three-step strategy
At the initial assessment, the first step of the three-
step IOTA strategy, consisting of the evaluation of six
simple descriptors was applied (Table I) [16]. If the le-
sion could not be classified with this first step (none of
the six descriptors was applicable or benignity and ma-
lignancy descriptors were present), it was evaluated with
the Simple Rules, the second step of the system (Table I)
[26].
Finally, if the mass could not be classified as benign
or malignant with the Simple Rules (none of the features
was present or they were both malignant and benign),
the diagnosis was inconclusive, and their images were
referred for subjective assessment by an expert sonogra-
pher (J.L.A.), with training equivalent to Level-3 of the
EFSUMB [24].
170 Juan José Hidalgo et al
Table I. Simple descriptors (Step 1) and Simple Rules (Step 2) for classifying adnexal masses with IOTA three-step strategy
Simple descriptors [16]
Predictors of benignity
• Unilocular tumour with ground glass echogenicity
in a premenopausal woman
• Unilocular tumour with mixed echogenicity
and acoustic shadows in a premenopausal woman
• Unilocular anechoic tumour with regular walls
and maximum diameter of lesion < 10 cm
• Remaining unilocular tumour with regular walls
Predictors of Malignancy
• Tumour with ascites and at least moderate
color Doppler blood flow in a postmenopausal
woman
• Age > 50 years and CA 125 > 100 U/mL
Comparison of IOTA three-step strategy and logistic regression model LR2
Simple rules [26]
For predicting a benign tumour (B-rules)
B1 Unilocular
B2 Solid components where the largest has a largest diameter < 7 mm
B3 Acoustic shadows
B4 Smooth multilocular tumour with largest diameter <100 mm
B5 No blood flow (color score 1)
For predicting a Malignant tumour (M-rules)
M1 Irregular solid tumour
M2 Ascites
M3 At least four papillary projections
M4 Irregular multilocular solid tumour with largest diameter ≥100 mm
M5 Very strong blood flow (color score 4)

LR2 logistic regression model
Once the lesion was classified using the three-step
strategy, it was evaluated using the LR2 model for cal-
culating the malignancy probability assessing six vari-
ables: (1) age, (2) ascites, (3) papillary blood flow, (4)
maximum solid component diameter, (5) irregular inter-
nal cystic walls and (6) acoustic shadows. The result was
applied to the logistic regression formula y = 1/(1 + e−z),
where z = −5.3718 + 0.0354 (1) + 1.6159 (2) + 1.1768
(3) + 0.0697 (4) + 0.9586 (5) - 2.9486 (6), using IOTA
Models® application software, available with IOS® op-
erating system. A probability above 10% was considered
a high risk of malignancy [22].
Therapeutic attitude
The therapeutic attitude consisted of surgical treat-
ment or expectant clinical and ultrasound follow-up.
Patients who showed high risk of malignancy were ad-
dressed to the referral oncology gynaecology unit. Pa-
tients with masses considered benign symptomatic were
referred to their general gynaecologist. In both cases, the
therapeutic attitude was determined by their referring
gynaecologist. Asymptomatic patients with a low-risk
lesion were offered surgical treatment by a general gy-
naecologist or expectant management with clinical and
ultrasound follow-up at 3, 6 and 12 months during the
first year, and annually thereafter.
The referring gynecologist decided whether to deter-
mine serum tumor markers during the follow-up of the
patients in whom a conservative attitude was chosen. In
these cases, the value of these markers was not known by
the investigating sonographer at the time of successive
follow-up ultrasound evaluations.
Reference standard
Lesions managed expectantly were considered be-
nign if they did not show ultrasound changes suggesting
malignancy for at least one year. If the lesion disappeared
spontaneously before completing one year of follow-up,
the patient was excluded from the study. If any ultra-
sound scan revealed morphological changes in the lesion,
the probability of malignancy was recalculated with the
three-step strategy and the LR2 system. In the removed
lesions, the standard reference was the histological diag-
nosis. Tumors that were histologically diagnosed as bor-
derline were considered malignant for statistical analysis.
Statistical analysis
For the qualitative variables, the frequency distribu-
tion value for each of the categories was obtained, pre-
senting the data in the form of absolute counts and per-
centages. Quantitative variables were studied following
Kolmogorov-Smirnov analysis to determine if the distri-
bution of values was normal. Quantitative variables with
normal distribution were expressed as a central tendency
measure with the standard deviation and range. When
they did not follow a normal distribution, they were pre-
sented as median and range.
Sensitivity, specificity, and positive (LR+) and nega-
tive (LR-) likelihood ratio were calculated with a 95%
confidence interval of the three-step model in order to
discriminate between benign and malignant adnexal le-
sions. The sensitivity and specificity obtained with the
application of the three-step model were compared with
the McNemar test with those obtained with the logistic
regression model LR2. A p<0.05 was considered statisti-
cally significant for all comparisons.
Statistical analysis was performed using the IBM
SPSS version 20® for Windows (SPSS Inc., Chicago, IL).
Results
One hundred and two patients with an adnexal mass
were studied. Two (1.9%) patients were excluded, one
for a follow-up shorter than 12 months and the other for
 Med Ultrason 2021; 23(2): 168-175 171
being pregnant during the study. Therefore, 100 patients
were included in the final analysis. The average age was
50.6 years (SD 17.4; range 18-87 years). Fifty-eight
(58%) patients were premenopausal.
Sixty-two (62%) patients were surgically treated and
the rest had follow-up for at least 12 months. Eighty-three
(83%) lesions were benign. Out of the total surgically re-
moved adnexal masses, 45 (72.6%) were benign and 17
(27.4%) malignant. The mean follow-up time was 14.2
months (SD 2.8; range 12-22 months). All patients in the
expectant management group showed lesions that were
considered benign at the end of the follow-up, showing
no ultrasound or clinical changes suggestive of malig-
nancy. Figure 1 shows a flow chart with the patient se-
lection, their management and the final diagnostic result.
Fig 1. Study flowchart showing the selection and management
The mean age of patients with malignant masses was of patients with an adnexal mass
51.6 (SD 17.7; range 22-85), whereas the mean age for
women with benign lesions was 49.2 (SD 15.9; range II summarizes the histological diagnoses of all adnexal
18-87). Benign lesions were diagnosed in 48 (57.8%) lesions removed with their diagnostic step for the three-
premenopausal and 35 (42.2%) postmenopausal patients. step strategy and the average LR2 outcome for each his-
Malignant lesions occurred in 8 (47%) premenopausal tological group.
and 9 (53%) postmenopausal women. Figure 2 shows a flowchart with the classification
The most common histological diagnoses among of benignity or malignancy of adnexal masses with the
the benign lesions removed were: 15 (33.3%) endome- three-step strategy, indicating the diagnostic results of
triomas and 12 (26.7%) serous cystadenomas. The most each of them according to the reference standard. With
common malignant tumours were: 6 (35.3%) serous the first two steps, 91 (91%) lesions could be classified
carcinomas and 3 (17.6%) mucinous carcinomas. Table by non-expert sonographers (37 (37%) with simple de-

Table II. Histology of the adnexal masses removed surgically (n=62) with number of cases correctly and erroneously diagnosed in
each step of the three-step strategy and the average LR2 outcome for each histological group.
Histology n (%) SD SR Expert LR2 (%)
Benign masses (n=45)
Endometrioma 15 (33.3) 8 5 (+1 FP)§ 1 5.1
Serous cystadenoma 12 (26.7) 4 7 1 17.4
Teratoma 7 (15.6) 3 4 0 1.3
Mucinous cystadenoma 4 (8.9) 1 (+1 FP)§ 1 1 2
Fibroma 2 (4.4) 0 0 2 24.4
Cystadenofibroma 2 (4.4) 0 1 1 8.8
Fibrothecoma 1 (2.2) 0 1 0 8.4
Hydrosalpinx 1 (2.2) 0 0 1 35.5
Tubo-ovarian abscess 1 (2.2) 0 1 0 17.8
Malignant masses (n=17)
Serous cystadenocarcinoma 6 (35.3) 1 5 0 60.8
Mucinous adenocarcinoma 3 (17.6) 1 2 0 80.6
Serous borderline tumor 2 (11.8) 0 1 (+1 FN)¥ 0 19.9
Endometrioid carcinoma 1 (5.9) 1 0 0 88.9
Clear cell carcinoma 1 (5.9) 0 0 1 62.8
Metastasis 1 (5.9) 0 1 0 65.4
Carcinosarcoma 1 (5.9) 1 0 0 96.3
Gastrointestinal stromal tumor 1 (5.9) 0 1 0 69.1
Presacral hemangiopericytoma 1 (5.9) 0 1 0 72.8
SD: Simple descriptor (First step); SR: Simple Rules (Second step); FP: False Positive; FN: False Negative; §: One further case was diag-
nosed as malignant (false positive); ¥: One further case was diagnosed as benign (false negative)
172 Juan José Hidalgo et al Comparison of IOTA three-step strategy and logistic regression model LR2
Fig 2. Study flowchart showing application in patients with
an adnexal mass (n=100) of IOTA three-step strategy. RS: ref-
erence standard; Results for each step according to reference
standard: histology in surgery group and change in sonographic
appearance in expectant management group.
scriptors and 54 (54%) using Simple Rules). An expert
sonographer classified the remaining 9 (9%) lesions.
With the first step (simple descriptors) no malignant le-
sion was classified as benign, and only one benign was
diagnosed as malignant (false positive). With the second
step (Simple Rules), a malignant lesion was classified as
benign (false negative) and a benign one as malignant
(false positive). With the third step (expert ultrasound
evaluation) all lesions were correctly classified. Figure 3
show the characteristics of the lesions that were misclas-
sified by the three-step strategy.
The final diagnostic performance of this strategy for
the whole study population was: sensitivity 94.1% (95%
CI, 86.7-98.3%); specificity 97.6% (95% CI, 94.8-99%);
LR+ 39.2 (95% CI, 16.6-100.2); and LR-0.06 (95% CI,
0.02-0.15). The diagnostic accuracy was 97%.
With the application of LR2 taking a result above
10% as a high probability of malignancy, 31 (31%) le-
sions were diagnosed as malignant and 69 (69%) as be-
nign. The mean probability of malignancy for all lesions
was 18% (SD 27.3; range 0.2-96.3). Of the 31 lesions
classified as malignant, 16 (51.6%) were definitely ma-
lignant and 15 (48.4%) benign (false positives) according
to the reference standard. Of the 69 lesions classified as
benign, 68 (98.5%) were benign and 1 (1.5%) was malig-
Fig 3. Ultrasound images (Gray scale and color Doppler) of
the misdiagnosed lesions with the IOTA three-step strategy
and characteristics of the patients with these adnexal masses:
a) False negative. Age 38. Diagnosed benign with Simple Rules
(rule B5), risk of malignancy with LR2: 4.4%. Histological
diagnosis: Serous borderline tumor; b) False positive. Age 87.
Diagnosed malignant with simple descriptors (age >50 year +
CA125=243), risk of malignancy with LR2: 78.8%. Histologi-
cal diagnosis: Mucinous cystadenoma; c) False positive. Age
33. Diagnosed malignant with Simple Rules (rule M1), risk of
malignancy with LR2: 40.4%. Histological diagnosis: Endome-
trioma
nant (false negative). The lesion that was false-negative
with LR2 gave the same result when applying the three-
step strategy.
The diagnostic performance of the LR2 model was:
sensitivity 94.1% (95% CI, 73-98.9%); specificity 81.9%
(95% CI, 72.3-88.7%); LR+ 5.2 (95% CI, 3.2-8.3); and
LR-0.07 (95% CI, 0.01-0.48), being the diagnostic accu-
racy 84%.
Comparison of the two diagnostic models using the
McNemar test was statistically significant (p<0.05),
which means that, given that the sensitivity for both was
identical, the three-step model is more specific in the
classification of lesions than LR2.
There was an agreement in the diagnosis of benignan-
cy or malignancy between the two systems in 87 (87%)
patients. The 13 discordant cases (13%) corresponded to
lesions diagnosed as benign with the three-step strategy
and malignant with LR2. In all cases, the final diagno-
sis according to the reference standard was benign, so all
these cases corresponded to false positives from the LR2
model. Regarding the timing of the three-step strategy in
 Med Ultrason 2021; 23(2): 168-175 173
Table III. Diagnostic parameters of lesions with a discordant classification between the three-step strategy and LR2
Age Three-step classification LR2 classification Histological diagnosis Ultrasound follow-up
(Diagnostic step) (Malignancy probability) (Diagnosis)
1 42 Benign (SR) Malignant (17,8%) Tuboovarian abscess -
2 44 Benign (SR) Malignant (10,2%) - Yes (Benign)
3 46 Benign (EA) Malignant (48%) - Yes (Benign)
4 59 Benign (SR) Malignant (14,6%) Serous cystadenoma -
5 59 Benign (SR) Malignant (12,2%) - Yes (Benign)
6 73 Benign (SR) Malignant (16,5%) - Yes (Benign)
7 42 Benign (SR) Malignant (11%) - Yes (Benign)
8 49 Benign (SR) Malignant (14,5%) Cystadenofibroma -
9 58 Benign (EA) Malignant (40,3%) Fibroma -
10 69 Benign (SR) Malignant (24,3%) Serous cystadenoma -
11 47 Benign (EA) Malignant (61,2%) Serous cystadenoma -
12 55 Benign (SR) Malignant (26%) - Yes (Benign)
13 42 Benign (EA) Malignant (35,5%) Hydrosalpinx -
SR: Simple Rules (Second step); EA: Expert Assessment (Third step)

which each of the discordant cases was classified, nine
were applied the Simple Rules and four the subjective as-
sessment by the expert. Table III shows the characteris-
tics of these discordant cases between the two diagnostic
systems.
Discussion
Our study aims to compare the diagnostic accuracy
of two ultrasound-based predictive systems, the three-
step strategy and the logistic regression model LR2, both
proposed by IOTA, in a group of patients with adnexal
masses for classification as benign or malignant of those
lesions. Our results show that the three-step strategy pre-
sents better diagnostic performance than the LR2 model
and, therefore, the former should preferably be used in
the appraisal of the adnexal masses by non-expert sonog-
raphers.
Both systems demonstrated similar sensitivity
(94.1%) for the prediction of malignancy, which implies
that their diagnostic capacity for patients with ovarian
cancer is high. Therefore, both models would have a low
rate of false negatives, meaning that few patients with a
malignant pathology would be diagnosed as benign, an
essential parameter in the diagnosis of ovarian cancer.
However, we observed a statistically significant differ-
ence in specificity in favour of the three-step strategy
(97.6% vs 81.9%), which would mean that fewer patients
with benign lesions would be misdiagnosed with malig-
nancy, avoiding unnecessary surgical interventions and
their potential complications, as well as improving the
economic costs and psychological repercussions in these
patients.
In this sense, the results of LR+ and LR- also showed
a higher performance of the three-step strategy compared
to LR2 by presenting results of 39.2 and 0.06 respec-
tively (vs 5.2 and 0.07 of LR2), which would indicate
a better practical utility of the first system, allowing the
confirmation of the presence of malignancy in the adnex-
al lesions with greater certainty, since a diagnostic test
with LR+>10 and LR-<0.1 is highly relevant and useful
[27].
So far, five external validation studies of the IOTA
three-step strategy have been published. However, none
of them has been compared with other predictive models
in the same group of patients [17-21]. We chose the LR2
logistic regression system as a comparison model since it
is one of the most evaluated, best-performing and most
used adnexal lesion diagnostic systems in clinical prac-
tice [13,15,22,23]. This is why other diagnostic models
have used this system as a reference in comparative stud-
ies, with the results being similar [28] or favourable to
LR2 [23,29-31].
This would show that the three-step strategy has
proven more diagnostic accuracy than one of the refer-
ence systems in the classification of adnexal lesions, that
is, the LR2.
The main support of our study relies on its perfor-
mance since it has been developed under conditions
similar to the usual clinical practice. We included both
surgically treated and expectantly managed lesions with
clinical and ultrasound follow-up, a therapeutic attitude
prevailing in many patients with adnexal masses. In addi-
tion, patients were evaluated by non-expert sonographers
and in a non-referral centre for ultrasound diagnosis of
ovarian cancer, also a common clinical practice for most
174 Juan José Hidalgo et al Comparison of IOTA three-step strategy and logistic regression model LR2
patients at their initial evaluation. We are aware that the
main limitation of our study was the number of patients
included, as a larger sample size would have strength-
ened our results.
Nevertheless, our results for the three-step strategy
can be compared to those published about other diagnos-
tic models. For example, the two most commonly used
systems such as the Risk of Malignancy Index and the
LR2 have shown in their validation studies sensitivity
and specificity of 76-87% and 57-97% [32-35] and 92-
94% and 75-85%, respectively [13,22,23,30]. Therefore,
these results and those obtained in comparison with LR2
in our study indicate a lower diagnostic performance of
the rest of the predictive models compared to the three-
step strategy. This fact could be validated with direct
comparison analysis with a larger sample.
We are aware that the main limitation of our study is
the limited number of patients included. We consider that
our results can be valid with the sample of 100 patients
analysed, although it would be worthwhile corroborating
them in a larger sample. Another limitation of our study
may have been that it was carried out in a single hospital,
in our case a primary-level hospital without a special-
ized ovarian cancer unit or expert sonographers, so the
inconclusive cases were evaluated by an expert sonogra-
pher by means of the electronic submission of ultrasound
images, being more appropriate for the expert to evaluate
the patients personally. Therefore, the best approach for
future studies would be to include a greater number of
centers at different levels of care, which would also allow
the sample of patients included to be expanded.
In conclusion, our results show that the three-step
strategy has high diagnostic accuracy for the classifica-
tion of benign or malignant adnexal lesions, surpassing
the LR2 logistic regression model. These findings sug-
gest that this strategy should be chosen in initial assess-
ments of ovarian pathology by non-expert sonographers,
as it also employs easily identifiable ultrasound param-
eters by most gynaecologists.
Conflict of interest: none
References
1. Cho KR, Shih IM. Ovarian cancer. Annu Rev Pathol
2009;4:287-313.
2. Giede KC, Kieser K, Dodge J, Rosen B. Who should op-
erate on patients with ovarian cancer? An evidence-based
review. Gynecol Oncol 2005;99:447-461.
3. Paulsen T, Kjaerheim K, Kaern J, Tretli S, Tropé C. Im-
proved short-term survival for advanced ovarian, tubal, and
peritoneal cancer patients operated at teaching hospitals. Int
J Gynecol Cancer 2006;16 Suppl 1:11-17.
4. Earle CC, Schrag D, Neville BA, et al. Effect of surgeon
specialty on processes of care and outcomes for ovarian
cancer patients. J Natl Cancer Inst 2006;98:172-180.
5. Engelen MJ, Kos HE, Willemse PH, et al. Surgery by con-
sultant gynecologic oncologists improves survival in pa-
tients with ovarian carcinoma. Cancer 2006;106:589-598.
6. Vernooij F, Heintz P, Witteveen E, van der Graaf Y. The
outcomes of ovarian cancer treatment are better when pro-
vided by gynecologic oncologists and in specialized hos-
pitals: A systematic review. Gynecol Oncol 2007;105:801-
812.
7. Timmerman D. The use of mathematical models to evaluate
pelvic masses; can they beat an expert operator? Best Pract
Res Clin Obstet Gynaecol 2004;18:91-104.
8. Valentin L, Jurkovic D, Van Calster B, et al. Adding a sin-
gle CA 125 measurement to ultrasound imaging performed
by an experienced examiner does not improve preopera-
tive discrimination between benign and malignant adnexal
masses. Ultrasound Obstet Gynecol 2009;34:345-354.
9. Timmerman D, Schwärzler P, Collins WP, et al. Subjective
assessment of adnexal masses with the use of ultrasonogra-
phy: an analysis of interobserver variability and experience.
Ultrasound Obstet Gynecol 1999;13:11-16.
10. Sokalska A, Timmerman D, Testa AC, Van Holsbeke C,
Lissoni AA, Leone FPG. Diagnostic accuracy of trans-
vaginal ultrasound examination for assigning a specific
diagnosis to adnexal masses. Ultrasound Obstet Gynecol
2009;34:462-470.
11. Valentin L, Hagen B, Tingulstad S, Eik-Nes S. Compari-
son of ‘pattern recognition’ and logistic regression models
for discrimination between benign and Malignantnt pelvic
masses: a prospective cross-validation. Ultrasound Obstet
Gynecol 2001;18:357-365.
12. Lee TS, Kim JW, Park NH, Song YS, Kang SB, Lee HP.
Assessing clinical performance of gynaecology residents:
sonographic evaluation of adnexal masses based on mor-
phological scoring systems. Ultrasound Obstet Gynecol
2005;26:776-779.
13. Nunes N, Yazbek J, Ambler G, Hoo W, Naftalin J, Jurkovic
D. Prospective evaluation of the IOTA Logistic Regression
Model (LR2) for the diagnosis of ovarian cancer. Ultra-
sound Obstet Gynecol 2012;40:355-359.
14. Geomini P, Kruitwagen R, Bremer GL, Cnossen J, Mol
BW. The accuracy of risk scores in predicting ovar-
ian Malignantncy: a systematic review. Obstet Gynecol
2009;113:384-394.
15. Kaijser J, Sayasneh A, Van Hoorde K, et al. Presurgical di-
agnosis using mathematical models and scoring systems: a
systematic review and meta-analysis. Hum Reprod Update
2014;20:449-462.
16. Ameye L, Timmerman D, Valentin L, et al. Clinically ori-
ented three-step strategy for assessment of adnexal pathol-
ogy. Ultrasound Obstet Gynecol 2012;40:582-591.
17. Sayasneh A, Kaijser J, Preisler J, et al. A multicenter pro-
spective external validation of the diagnostic performance
of IOTA simple descriptors and rules to characterize ovar-
ian masses. Gynecol Oncol 2013;130:140-146.

18. Testa A, Kaijser J, Wynats L, et al. Strategies to diagnose
ovarian cancer: new evidence from phase 3 of the multicen-
tre international IOTA study. Br J Cancer 2014;111:680-688.
19. Peces Rama A, Llanos Llanos MC, Sánchez Ferrer ML,
Alcázar Zambrano JL, Martínez Mendoza A, Nieto Díaz
A. Simple descriptors and simple rules of the Interna-
tional Ovarian Tumor Analysis (IOTA) Group: a prospec-
tive study of combined use for the description of adnexal
masses. Eur J Obstet Gynecol Reprod Biol 2015;195:7-11.
20. Alcázar JL, Pascual MA, Graupera B, et al. External vali-
dation of IOTA simple descriptors and simple rules for
classifying adnexal masses. Ultrasound Obstet Gynecol
2016;48:397-402.
21. Hidalgo JJ, Ros F, Aubá M, et al. Prospective external
validation of IOTA three-step strategy to characterize and
classify adnexal masses and retrospective assessment of an
alternative two-step strategy using simple rules risk. Ultra-
sound Obstet Gynecol 2019;53:693-700.
22. Timmerman D, Testa AC, Bourne T, et al; International
Ovarian Tumor Analysis Group. Logistic regression model
to distinguish between the benign and malignant adnexal
mass before surgery: a multicenter study by the Inter-
national Ovarian Tumor Analysis Group. J Clin Oncol
2005;23:8794-8801.
23. Van Holsbeke C, Van Calster B, Bourne T, et al. Exter-
nal validation of diagnostic models to estimate the risk
of Malignantncy in adnexal masses. Clin Cancer Res
2012;18:815-825.
24. Education and Practical Standards Committee, European
Federation of Societies for Ultrasound in Medicine and Bi-
ology. Minimum training recommendations for the practice
of medical ultrasound. Ultraschall Med 2006;27:79-105.
25. Timmerman D, Valentin L, Bourne T, et al;, International
Ovarian Tumor Analysis (IOTA) Group. Terms, definitions
and measurements to describe the ultrasonographic features
of adnexal tumours: a consensus opinion from the interna-
tional ovarian tumour analysis (IOTA) group. Ultrasound
Obstet Gynecol 2000;16:500-505.
26. Timmerman D, Testa AC, Bourne T, et al. Simple ultra-
sound based-rules for the diagnosis of ovarian cáncer. Ul-
trasound Obstet Gynecol 2008;31:681-690.
Med Ultrason 2021; 23(2): 168-175 175
27. Deeks JJ, Altman DG. Diagnostics tests 4: likelihood ratios.
BMJ 2004;17;329:168-169.
28. Meys EMJ, Jeelof LS, Achten NMJ, et al. Estimating risk
of Malignantncy in adnexal masses: external validation of
the ADNEX model and comparison with other frequent-
ly used ultrasound models. Ultrasound Obstet Gynecol
2017;49:784-792.
29. Kaijser J, Van Gorp T, Van Hoorde K, et al. A com-
parison between an ultrasound-based prediction mod-
el (LR2) and the Risk of Ovarian Malignancy Algo-
rithm (ROMA) to assess the risk of malignancy in
women with an adnexal mass. Gynecol Oncol 2013;129:377-
383.
30. Sayasneh A, Wynants L, Preisler J, et al. Multicentre ex-
ternal validation of IOTA prediction models and RMI by
operators with varied training. Br J Cancer 2003;108:2448-
2454.
31. Terrin N, Schmid CH, Griffith JL, D’Agostino RB, Selker
HP. External validity of predictive models: a comparison
of logistic regression, classification trees, and neural net-
works. J Clin Epidemiol 2003;56:721-729.
32. Moore RG, Jabre-Raughley M, Brown AK, et al. Compari-
son of a novel multiple marker assay versus the Risk of
Malignancy Index for the prediction of epithelial ovarian
cancer in patients with a pelvic mass. Am J Obstet Gynecol
2010;203:228.e1-e6.
33. Van Gorp T, Veldman J, Van Calster B, et al. Subjective
assessment by ultrasound is superior to the risk of Ma-
lignantncy index (RMI) or the risk of ovarian Malignant-
ncy algorithm (ROMA) in discriminating benign from
malignant adnexal masses. Eur J Cancer 2012;48:1649-
1656.
34. Jacobs I, Oram D, Fairbanks J, Turner J, Frost C, Grudz-
inskas JG. A risk of Malignantncy index incorporating CA
125, ultrasound and menopausal status for the accurate pre-
operative diagnosis of ovarian cancer. Br J Obstet Gynaecol
1990;97:922-929.
35. Bailey J, Tailor A, Naik R, et al. Risk of Malignantncy in-
dex for referral of ovarian cancer cases to a tertiary cen-
tre: does it identify the correct cases? Int J Gynecol Cancer
2006;16 Suppl 1:30-34.
Original papers Med Ultrason 2021, Vol. 23, no. 2, 176-180
DOI: 10.11152/mu-2789

The presence of effusions between the volar plate of the proximal

interphalangeal joint and the flexor digitorum tendon is a common
phenomenon: a single-center, cross sectional study
Leixi Xue1, Yi Zhang1, Dong Yan1, Jinxiang Fu2, Zhichun Liu1

1Department of Rheumatology and Immunology, 2Department of Hematology, the Second Affiliated Hospital of

Soochow University, Suzhou, China

Abstract
Aim: In clinical practice, an anechoic signal was often exhibited between the volar plate (VP) of the proximal interphalan-
geal joint (PIPJ) (PIPJVP) and the flexor digitorum tendon (FDT) on ultrasound, which suggests the presence of effusions
(PIPJVP-FDT effusions). The purpose of this study was to investigate the prevalence of PIPJVP-FDT effusions and to explore
the possible mechanism preliminarily. Material and methods: A single-center, cross sectional study in hand osteoarthritis
(HOA) patients, rheumatoid arthritis (RA) patients and healthy controls was conducted. Ultrasound examination was per-
formed by the same real-time scanner with 18-MHz linear array transducer. Bilateral interphalangeal joints (IPJs) of the
thumb, 2ed, 3rd, 4th and 5th PIPJs were examined. The PIPJVP-FDT effusions was defined as an anechoic signal between the
PIPJVP and FDT in two perpendicular ultrasound planes. Results: In total, 200 patients with HOA, 78 patients with RA and
101 healthy controls were eligible for the study. 37.6% of healthy controls and 35.0% of HOA patients showed PIPJVP-FDT
effusions, while only 11.5% of RA patients had PIPJVP-FDT effusions (p<0.001). The 2ed, 3rdand 4th PIPJs showed more
PIPJVP-FDT effusions, while the IPJs of the thumbs and 5th PIPJs showed less PIPJVP-FDT effusions (p<0.05). Furthermore,
the prevalence of PIPJVP-FDT effusions in different age groups were similar in HOA patients and healthy controls. Conclu-
sion: To the best of our knowledge, this paper is the first to demonstrate that the presence of PIPJVP-FDT effusions is a very
common phenomenon in HOA patients and healthy individuals, and may be unrelated to inflammation, degeneration and age.
Keywords: effusion; ultrasound; volar plate; proximal interphalangeal joint

Introduction tions, the VP provides a smooth sliding surface for the

The volar plate (VP) of the proximal interphalangeal
joint (PIPJ) (PIPJVP) which is composed of a membra-
nous, flexible, proximal part and a cartilaginous, thicker,
meniscoid distal part is an important part of the joint [1].
It reinforces the volar aspect of the capsule and protects
the PIPJ from hyperextension, lateral displacement, and
torsional forces [2]. Under normal physiological condi-
Received 21.08.2020  Accepted 10.12.2020
Med Ultrason
2021, Vol. 23, No 2, 176-180
Corresponding author: Zhichun Liu
Department of Rheumatology and Immunology,
the Second Affiliated Hospital of Soochow
University, 1055 Sanxiang Road,
215006 Suzhou, China
E-mail: liuzhichun5190@163.com
flexor tendon (FDT) [1], and both of them participate in
the grasping activity of the hand [2].
In recent years, ultrasound has become a popular im-
aging tool for evaluating the lesions of hand joints [3,4].
The different structures of PIPJ have different reflecting
capabilities that determine their appearance on the ultra-
sound image. Each structure has its own characteristic
and can be clearly observed by a high-frequency ultra-
sound transducer. The VP on ultrasound is often an isoec-
hogenic and homogeneous structure underlying the FDT
[1,5], while longitudinal imaging of FDT shows a typical
network of linear hyperechoic fibrillar patterns on ultra-
sound when the insonation angle is 90º [6]. In general,
there should be no abnormal echo signal between PIPJVP
and FDT on ultrasound. However, an anechoic signal is
often exhibited between PIPJVP and FDT on ultrasound

in clinical practice, which may suggest the presence of
effusions (PIPJVP-FDT effusions) [7]. The prevalence
and mechanism of PIPJVP-FDT effusions have not yet
been reported. Therefore, the purpose of this study was to
investigate the prevalence of the PIPJVP-FDT effusions
by ultrasound in different groups, including patients with
hand osteoarthritis (HOA), patients with rheumatoid ar-
thritis (RA) and healthy controls and to further explore
the possible mechanism of its occurrence preliminarily.
Material and methods
Patients enrollment
A single-center, cross sectional study in HOA pa-
tients, RA patients and healthy controls was conducted at
the outpatient clinic of the Department of Rheumatology
and Immunology, the Second Affiliated Hospital of Soo-
chow University from June 2017 to February 2018. The
diagnosis of HOA patients in this study met the following
two points: pain, swelling and (or) morning stiffness in
PIPJs, and cortical protrusion (i.e. step-up) observed in
two perpendicular ultrasound planes at the PIPJ margin.
However, HOA patients were excluded from this study
when they were combined with other joint diseases or
connective tissue diseases, such as RA, psoriatic arthri-
tis, gout, undifferentiated arthritis, systemic lupus erythe-
matosus, unexplained tendinitis and tenosynovitis of the
hand. The RA patients with PIPJs involvement fulfilled
the classification criteria of 2010 American College of
Rheumatology/European League Against Rheumatism
[8]. RA patients complicated with other joint diseases
or connective tissue diseases, such as HOA, psoriatic
arthritis, systemic lupus erythematosus, were excluded.
It should be emphasized that health volunteers were
excluded if they met any of the following: recent joint
pain, swelling pain, or morning stiffness in PIPJs; corti-
cal protrusion observed in two perpendicular ultrasound
planes in PIPJs; cortical defect observed in two perpen-
dicular ultrasound planes in PIPJs; synovial hypertrophy
or synovial fluid within articular cavity or flexor tendon
sheath of hand on ultrasound. This study was approved
by the Human Ethics Review Committee of the Second
Affiliated Hospital of Soochow University and written
informed consent was obtained from each patient and
healthy control, according to the World Medical Associa-
tion Declaration of Helsinki, revised in 2000, Edinburgh.
Ultrasound examination
Ultrasound examination was performed by the same
real-time scanner (MyLab30, Esaote, Italy) with a 8-MHz
linear array transducer by a specialized in musculoskel-
etal ultrasonography who was blinded to other clinical
information. All the participants were seated with their
Med Ultrason 2021; 23(2): 176-180 177
Fig 1. A) Normal longitudinal sonogram of the proximal inter-
phalangeal joint; B) and C) Both longitudinal and transverse
sonogram showing an anechoic signal between the volar plate
of the proximal interphalangeal joint and flexor digitorum ten-
don in the same individual, which suggests the presence of ef-
fusions. E, effusion; FDT, flexor digitorum tendon; VP, volar
plate.
hands relaxed in a neutral position on a table. Bilateral
interphalangeal joints (IPJs) of the thumb, 2ed, 3rd, 4th and
5th PIPJs were examined. Each joint was scanned in lon-
gitudinal plane and transverse plane. The PIPJVP-FDT
effusions was defined as an anechoic signal between the
PIPJVP and FDT in two perpendicular ultrasound planes
(fig 1).
Statistical analysis
All variables were analyzed based on IBM Corp.
Released 2012. IBM SPSS Statistics for Windows, Ver-
sion 21.0. Armonk, NY: IBM Corp. Continuous variables
were presented as means and standard deviation (normal
distribution) or median and range (non-normal distribu-
tion). Categorical variables were reported as absolute fre-
quency and percentage from the sub/group. The normal-
ity was tested by the Shapior-Wilk test. Since the number
of joints with PIPJVP-FDT effusions in the patients and
controls were non-normal distributed, their differences
among HOA, RA and control groups were tested by the
Kruskal-Wallis test and the subsequent pairwise compar-
ison. The results of the pairwise comparison were adjust-
ed by the Benjamini-Hochberg method. The correlation
of the PIPJVP-FDT effusions number between the right
and left hands in each patient was analyzed by Spearman
rank correlation analysis and the difference was tested by
the Wilcoxon signed-rank test. Chi-square test was used
to compare the prevalences of PIPJVP-FDT effusions
in different joints and different groups. The significance
level was set at p<0.05.
Results
In total, 200 patients with HOA, 78 patients with
RA and 101 healthy controls were eligible for the study.
Although the average age of HOA and RA group was
178 Leixi Xue et al Effusions between the volar plate of the proximal interphalangeal joint and the flexor digitorum tendon
Table I. Characteristics of baseline demographic and PIPJVP-
FDT effusions
HOA patients RA patients Controls
(n=200) (n=78) (n=101)
Female sex, n 162 (81.0) 57 (73.1) 79 (78.2)
(%)
Age (years), 56.8 ± 12.2† 50.9 ± 43.6 ± 12.8
mean ± SD 16.3†‡
Patients with 70 (35.0) 9 (11.5)s§ 38 (37.6)
PIPJVP-FDT
effusions, n (%)
Total number of 165 25 139
PIPJVP-FDT
effusions, n
Fig 2. The distribution of effusions number in RA patients, † p<0.05 compared with controls; s p<0.001 compared with con-
HOA patients and healthy controls. HOA, hand osteoarthritis; trols; ‡ p<0.05 compared with OA patients; § p<0.001 compared
RA, rheumatoid arthritis. with OA patients; HOA, hand osteoarthritis; RA, rheumatoid ar-
thritis; PIPJVP-FDT effusions, effusions between volar plate of the
older than the healthy group (p<0.05), there were no proximal interphalangeal joint and flexor digitorum tendon.
sex differences. We found that 37.6% (38 out of 101)
of healthy controls and 35.0% (70 out of 200) of HOA Table II. The number of PIPJVP-FDT effusions in the right and
patients showed PIPJVP-FDT effusions, while only left hands
11.5% (9 out of 78) of RA patients had PIPJVP-FDT HOA Control
effusions (p<0.001). The total number of PIPJVP-FDT Right hand PIPJVP-FDT effusions, 83 (50.3) 72 (51.8)
effusions was 139 in 38 healthy controls and 165 in 70 n (%)
HOA patients, but only 25 in 9 RA patients (Table I). Left hand PIPJVP-FDT effusions, 82 (49.7) 67 (48.2)
The differences of the PIPJVP-FDT effusions number n (%)
among HOA, RA and control groups were analyzed. The Total number, n 165 139
results of pairwise comparison showed that compared
HOA, hand osteoarthritis; RA, rheumatoid arthritis; PIPJVP-FDT
with that in RA patients, the median of PIPJVP-FDT ef- effusions, effusions between volar plate of the proximal inter-
fusions number was larger in HOA patients (p=0.001) phalangeal joint and flexor digitorum tendon.
and healthy controls (p<0.001) (fig 2), but there were no
differences between HOA patients and healthy controls Table III. The number of PIPJVP-FDT effusions in proximal
(p>0.05). interphalangeal joints
As the total number of PIPJVP-FDT effusions in RA HOA Control
patients was too few, we analyzed only the characteristics
of PIPJVP-FDT effusions in HOA patients and healthy PIPJVP-FDT effusions in IPJs, 2（1.2） 2 (1.4)‡
controls. The correlation between right and left hand PIP- n (%)
JVP-FDT effusions was studied and the Spearman corre- PIPJVP-FDT effusions in 2ed PIPJs, 40 (24.2)†‡ 35 (25.2)†
lation coefficient was 0.690 (p<0.01). The difference of n (%)
the PIPJVP-FDT effusions number between right and left PIPJVP-FDT effusions in 3rd PIPJs, 69 (41.8)†‡ 45 (32.4)†‡
hands was tested, and the results showed that there were n (%)
no significant statistic differences (p>0.05) (Table II). PIPJVP-FDT effusions in 4th PIPJs, 44 (26.7)†‡ 40 (28.8)†‡
The 2ed, 3rdand 4th PIPJs showed more PIPJVP-FDT effu- n (%)
sions, while the IPJs of the thumbs and 5th PIPJs showed PIPJVP-FDT effusions in 5th PIPJs, 10 (6.1) 17 (12.2)†
less PIPJVP-FDT effusions (Table III). n (%)
We further analyzed the effect of age on the presence Total number, n 165 139
of PIPJVP-FDT effusions in HOA patients and healthy
controls. An age value approximating the average age † p<0.05 compared with IPJs in the same group; ‡ p<0.05 com-
was served as the grouping point. The prevalence of pared with 5th PIPJs in the same group. HOA, hand osteoarthri-
tis; IPJs, interphalangeal joints; PIPJs, proximal interphalangeal
PIPJVP effusions in HOA patients with age < 55 and ≥
joints; RA, rheumatoid arthritis; PIPJVP-FDT effusions, effusions
55 years were 36.6% and 33.6% (p>0.05), respectively. between volar plate of the proximal interphalangeal joint and flexor
In healthy controls, there were also no differences in the digitorum tendon.

prevalence of PIPJVP effusions between age < 45years
and ≥ 45 years (34.7% versus 40.4%, p>0.05).
Discussion
In our study, we investigated the characteristics of
PIPJVP-FDT effusions in HOA patients, RA patients
and healthy controls by ultrasound. The prevalence of
PIPJVP-FDT effusions in healthy controls was compa-
rable with that in HOA patients and significantly higher
than that in RA patients. Moreover, compared with that
in RA patients, the median of the PIPJVP-FDT effusion
number was larger in HOA patients and healthy controls,
but there were no differences between HOA patients and
healthy controls (p>0.05). Furthermore, age made no dif-
ferences in the prevalence of PIPJVP-FDT effusions in
HOA patients and healthy controls. Therefore, the pres-
ence of PIPJVP-FDT effusions is a very common phe-
nomenon in HOA patients and healthy individuals, and
may be unrelated to inflammation, degeneration and
age.
Our study also showed that the prevalence of PIP-
JVP-FDT effusions in 2ed, 3rd and 4th PIPJs were higher
than that in IPJs and 5th PIPJs. The IPJs have a narrower
range of movement, the 5th PIPJs have the lower mo-
tion intensity and the middle three fingers can produce
a more powerful mechanical force in grasping, holding
and twisting [9,10]. The above analysis may explain
why the three middle PIPJs are more likely to produce
PIPJVP-FDT effusions. Therefore, the occurrence of
PIPJVP-FDT effusions may be related to the intensity of
joint movement. In addition, the anatomical structure of
the thumb is distinctly different from the other fingers, so
the different anatomy may also contribute to the different
prevalence of effusions.
The intensity of joint movement may also interpret
the differences in PIPJVP-FDT effusions among differ-
ent groups. RA patients often suffer from severe hand
joint pain, dysfunction and even deformity, thus sig-
nificantly restricting the movement of hand joints [11];
therefore, RA patients had the lower prevalence and me-
dian number of PIPJVP-FDT effusions. In HOA patients,
the movement intensity of hand joints decreased only
slightly due to pain and morning stiffness, so there were
no significant differences in the prevalence and median
number of PIPJVP-FDT effusions between HOA patients
and healthy controls.
Although we did not study the role of the dominant
hand in the pathogenesis of PIPJVP-FDT effusions, con-
sidering that more people are right-handed, there should
be more PIPJVP-FDT effusions in the right hand than in
the left hand. Surprisingly, the present study showed that
Med Ultrason 2021; 23(2): 176-180 179
the number of PIPJVP-FDT effusions in the left and right
hands was similar; moreover, there was a correlation be-
tween the left and right hands in the presence of PIPJVP-
FDT effusions. These results suggest that the appearance
of PIPJVP-FDT effusions may not be associated with the
amount and duration of joint movement, which is sup-
ported by the fact that the prevalence of PIPJVP-FDT
effusions in younger participants was comparable with
that in older people in HOA patients and healthy controls.
A limitation of this study is the lack of homogeneity
regarding the age in the three study groups. Ultrasound is
more sensitive than conventional radiography in the di-
agnosis of HOA [12-14]. What’s more, the average age in
the HOA group is more than 50 years old. Therefore, it is
very difficult to find age-matched controls without PIPJ
osteophytes on ultrasound. Moreover, we have verified
that age has no effect on the presence of PIPJVP-FDT
effusions, so the impact of age differences among three
groups may be negligible. Another limitation is the larger
female group in the HOA population. HOA patients were
recruited through outpatient clinics, suggesting that more
women are willing to seek help from doctors than men.
This factor contributes to the fact that women outnumber
men in our study. In addition, the prevalence of HOA in
females is indeed significantly higher than that in males.
A Framingham analysis of incidence of HOA showed
an age-standardized prevalence of 44.2% in women and
37.7% in men, respectively [15]. At age 63 years, the
HOA prevalence was 86% in females and 67% in males,
and the difference was statistically significant (p=0.005)
[14]. In addition, our study could not demonstrate the va-
lidity and interobserver reliability of PIPJVP-FDT effu-
sions on ultrasound, so these will need further investiga-
tion in a new cohort study.
In conclusion, to the best of our knowledge, this is
the first study that demonstrated that the presence of
PIPJVP-FDT effusions is a very common phenomenon
in HOA patients and healthy individuals and this may
be related to the movement intensity and the anatomical
structure of joint rather than inflammation, degeneration
and age. However, the exact mechanism and significance
of PIPJVP-FDT effusions remain unknown and further
investigations are required.
Conflicts of interest: none.
Acknowledgments: This work was supported
by the National Nature Science Foundation of Chi-
na (81800622); the Natural Science Foundation of
the Jiangsu Higher Education Institutions of China
(18KJB320020); the Suzhou Health and Key Talent Pro-
ject (GSWS2019011); and the Jiangsu social develop-
ment project (BE2019663).
180 Leixi Xue et al Effusions between the volar plate of the proximal interphalangeal joint and the flexor digitorum tendon
References
1. Saito S, Suzuki Y. Biomechanics of the volar plate of the
proximal interphalangeal joint: a dynamic ultrasonographic
study. J Hand Surg Am 2011;36:265-271.
2. Bayer T, Schweizer A, Müller-Gerbl M, Bongartz G.
Proximal interphalangeal joint volar plate configuration
in the crimp grip position. J Hand Surg Am 2012;37:899-
905.
3. Fukae J, Kon Y, Henmi M, et al. Change of synovial vas-
cularity in a single finger joint assessed by power doppler
sonography correlated with radiographic change in rheu-
matoid arthritis: comparative study of a novel quantita-
tive score with a semiquantitative score. Arthritis Care Res
(Hoboken) 2010;62:657-663.
4. Backhaus M, Burmester GR, Gerber T, et al. Guidelines for
musculoskeletal ultrasound in rheumatology. Ann Rheum
Dis 2001;60:641-649.
5. Sato J, Ishii Y, Noguchi H, Takeda M. Sonographic appear-
ance of the flexor tendon, volar plate, and A1 pulley with
respect to the severity of trigger finger. J Hand Surg Am
2012;37:2012-2020.
6. Bruyn GA, Hanova P, Iagnocco A, et al. Ultrasound defini-
tion of tendon damage in patients with rheumatoid arthritis.
Results of a OMERACT consensus-based ultrasound score
focussing on the diagnostic reliability. Ann Rheum Dis
2014;73:1929-1934.
7. Wakefield RJ, Balint PV, Szkudlarek M, et al. Musculoskel-
etal ultrasound including definitions for ultrasonographic
pathology. J Rheumatol 2005;32:2485-2487.
8. Aletaha D, Neogi T, Silman AJ, et al. 2010 rheumatoid ar-
thritis classification criteria: an American College of Rheu-
matology/European League Against Rheumatism collabo-
rative initiative. Ann Rheum Dis 2010;69:1580-1588.
9. Long C 2nd, Conrad PW, Hall EA, Furler SL. Intrinsic-ex-
trinsic muscle control of the hand in power grip and preci-
sion handling. An electromyographic study. J Bone Joint
Surg Am 1970;52:853-867.
10. Landsmeer JM. Power grip and precision handling. Ann
Rheum Dis 1962;21:164-170.
11. Morisawa S. Functional disability and hand deformity in
rheumatoid arthritis. Ryumachi 1984;24:18-24.
12. Sivakumaran P, Hussain S, Ciurtin C. Comparison between
Several Ultrasound Hand Joint Scores and Conventional
Radiography in Diagnosing Hand Osteoarthritis. Ultra-
sound Med Biol 2018;44:544-550.
13. Keen HI, Wakefield RJ, Grainger AJ, Hensor EMA, Em-
ery P, Conaghan PG. Can ultrasonography improve on
radiographic assessment in osteoarthritis of the hands? A
comparison between radiographic and ultrasonographic de-
tected pathology. Ann Rheum Dis 2008;67:1116-1120.
14. Abraham AM, Pearce MS, Mann KD, Francis RM, Birrell
F. Population prevalence of ultrasound features of osteoar-
thritis in the hand, knee and hip at age 63 years: The New-
castle thousand families birth cohort. BMC Musculoskelet
Disord 2014;15:162.
15. Haugen IK, Englund M, Aliabadi P, et al. Prevalence, in-
cidence and progression of hand osteoarthritis in the gen-
eral population: The Framingham Osteoarthritis Study. Ann
Rheum Dis 2011;70:1581-1586.
Original papers Med Ultrason 2021, Vol. 23, no. 2, 181-187
DOI: 10.11152/mu-2819

Assessment of testes with two-dimensional Shear Wave Elastography

in patients with operated inguinal hernia
Mehmet Sedat Durmaz1, Fatih Ateş1, Serdar Arslan2, Turgay Kara3, Funda Gökgöz Durmaz4,
Mehmet Ali Eryılmaz5, Kemal Arslan5

1Selçuk University, Medicine Faculty, Department of Radiology, Konya, 2Istanbul University, Cerrahpasa Medicine
Faculty, department of Radiology, Istanbul, 3Health Sciences University, Training and Research Hospital, Depart-
ment of Radiology, Konya, 4Karatay Community Health Center Family Medicine Konya, 5Health Sciences University,
Training and Research Hospital, Department of General Surgery, Konya, Turkey

Abstract
Aim: We compared the two-dimensional shear-wave elastography (2D-SWE) values between the testes with same side
operated inguinal hernia (IH) and the contralateral testes, as well as the testes of healthy volunteers without IH. Material and
methods: A total of 189 participants (117 unilateral [117 testes] and 8 bilateral operated IH patients [16 testes] and 64 healthy
volunteers [128 testes]), providing a total of 378 testicles, were investigated prospectively. All patients underwent B-mode
ultrasonography (US) and 2D-SWE examinations. Operation type, the period between diagnosis and operation, the period
since the operation, testes volumes, and 2D-SWE values were compared. Results: The B-mode US finding of the testes were
normal in all participants. The mean testes’ volume of same side operated IH was significantly lower comparing to contralat-
eral testes and the healthy group (p<0.001). The 2D-SWE values of the testes with same side operated IH were significantly
higher comparing to the contralateral testes and the healthy group (p<0.001). There was a statistically significant correlation
between 2D-SWE values and IH severity, as well as the duration of the hernia (p=0.001). There was no significant correlation
between the IH severity and testes volume (p=0.285). No significant difference was found between the direct and indirect IH
in terms of testicular volume and SWE values and between the duration of the hernia, the time after sugery, testicular volume
and SWE values according to operation techniques (p>0.005). Conclusions: The 2D-SWE can be used as an effective imag-
ing method to evaluate testicular stiffness with objective numerical values, to estimate the severity of histologic damage in
patients with operated IH.
Keywords: inguinal hernia; Shear Wave Elastography; testes; ultrasonography

Introduction
Inguinal hernia (IH) is an important public health
problem that requires surgical treatment and has an in-
cidence rate of approximately 5 to 10% [1,2]. There are
Received 12.09.2020  Accepted 15.12.2020
Med Ultrason
2021, Vol. 23, No 2, 181-187
Corresponding author: Fatih Ateş MD
Selçuk University, Medicine Faculty,
Department of Radiology, Konya, Turkey
313 Ardıclı Mahallesi, Celal Bayar Cad.,
Selçuklu, 42250 Konya, Turkey
Phone: +90 5468375296
Fax: +903325121653
E-mail: fatih_ates81@hotmail.com
two main subtypes of IH, direct and indirect, the last be-
ing more common [3]. IH can lead to serious complica-
tions, including testicular damage and atrophy resulting
from the narrowing of the blood vessels in the spermatic
cord [4]. Testicular damage due to IH surgery has been
well documented being reported to range from 1 to 14%
[3]. The testicular blood supply may return to normal as
the pressure effect of the hernia-induced spermatic cord
is removed after the operation [5,6]. However, testicular
damage can be caused by injury to the spermatic cord and
vas deferens during surgery, postoperative infection, and
foreign body inflammatory responses to the mesh, scar,
and granulation tissue [7-9].
Progressive histopathological changes, such as the
reduction of the germ cell count, delay germ cell matura-
182 Mehmet Sedat Durmaz et al Assessment of testes with 2D-SWE in patients with operated inguinal hernia
tion and atrophy can occur due to the effects of the IH
and possible postoperative changes in the inguinal area
[10]. As a consequence, the testis becomes stiffer due
to parenchymal fibrosis [11]. Ultrasonography (US) is
a commonly used imaging technique to assess testicular
pathologies. Testicular morphology, volume, and vas-
cularity can be evaluated with US. However, US results
may indicate no testicular damage even though histologi-
cal testicular damage is present [11,12]. Testicular biopsy
is the gold standard technique for evaluating histological
damage. However, this is no longer recommended be-
cause of the high likelihood that the procedure will lead
to complications [11].
Two-dimensional shear wave elastography (2D-
SWE), a US imaging technique that provides quantitative
measurement of tissue stiffness, provides an alternative
technique for estimating histopathological abnormalities
that cannot be detected by conventional US [11]. The
2D-SWE can evaluate the testis stiffness to estimate the
level of fibrosis resulting from damage has been reported
[11,12]. Parenchymal damage in the testis could be as-
sessed using a quantitative evaluation of the degree of
parenchymal fibrosis caused by IH and possible postop-
erative changes in the inguinal area.
In this study, we quantitatively evaluated the stiffness
of testes parenchyma using 2D-SWE on the same side
as the IH operation and we compared the results with
normal testes. To the best of our knowledge, no previ-
ous studies have focused on the evaluation of posibile
testicular damage in patients with operated IH using the
2D-SWE technique.
Materials and methods
This study was conducted between December 2017
and January 2020, after the approval of the local research
Ethics Committee. All subjects were informed about the
study protocol and written informed consent was ob-
tained before the procedure.
Table I. Number of testes included into the study, according to grouping of testes.
Classification
Symptomatic Inguinal canal proximal Direct IH (33)
group (grade 1) (14)
History of
Inguinal canal middle
operation
(grade 2) (60) Indirect IH (100)
for IH
Inguinal canal distal
or scrotal (grade 3) (59)
Asymptomatic Normal opponent of IH operation testicle
group
Bilateraly normal testes
Inguinal hernia (IH), endoscopic totally extraperitoneal (TEP).
A total of 189 participants and 378 testes (1:1 ratio of
left to right) were included in the study. Group A, total-
ling 133 testes, was composed of 117 patients who had
undergone unilateral IH surgery (117 testes) and 8 pa-
tients who had undergone bilaterally IH surgery (16 tes-
tes). Group A was divided into three subgroups according
to the hernia repair operation technique applied: preperi-
toneal alloplasty with a posterior open inguinal approach
(Kugel); open mesh hernia repair (Lichtenstein); and
endoscopic totally extraperitoneal (TEP) mesh hernia
repair. There were no reported intraoperative complica-
tions in this group of patients. Group A was further sub-
divided into IH types, either direct or indirect, accord-
ing to the surgical findings, and the level of IH location
(proximal [grade I], middle [grade II] and distal inguinal
canal, almost in scrotum [grade III]). The duration of the
hernia in these patients and the period between the IH op-
eration and the US examination were recorded. Group B
was composed of 117 testes located on the opposite side
of the site of IH surgery. Group C was composed of 128
testes from 64 healthy volunteers. The groups character-
istics are summarized in Table I.
Patients with any of the following traits or conditions
were excluded from the study: less than 18 years of age;
a history of any lower abdominal-pelvic surgery or or-
chiectomy; newly diagnosed IH; recurrent or failed IH
repair; previously emergency exploration for complica-
tions of IH (e.g., irreducible hernia, bowel obstruction,
and strangulation); previous chemotherapy/radiotherapy
treatment; previous abnormal US findings, such as focal
lesion in the testicle or epididymis, varicocele, hydro-
cele, epididymitis, and orchitis.
US and 2D-SWE
All participants underwent B-mode US and 2D-SWE
examinations of the testes in the supine position. The US
and 2D-SWE examinations were performed with a high-
frequency (4–14 MHz) linear array transducer (Aplio
500, Canon Medical System Corporation, Tustin, CA).
Each patient was examined under the same standard cir-
Nr. of testes
Lichtenstein operation history (80) Group A 133 (35.19%)
TEP operation history (35)
Kugel operation history (18)
Group B 117 (30.95%)
Group C 128 (33.86%)
 Med Ultrason 2021; 23(2): 181-187 183

Fig 1. A 52-year-old male healthy volunteer. The right testis was measured in three dimensions and testes volume was calculated
as 16.4 cm³. The B-mode US findings of the testis were in normal parenchymal structure (a). Quantitative 2D-SWE values were
measured in the transverse plane at the level of the testicular hilus by drawing manually contours of the entire testis structure on
propagation mode with a free ROI (b). The quantitative elasticity values were measured as 6.1 kPa (c) and 1.41 m/s (d).

cumstances. US and 2D-SWE were carried out by a radi- Statistical analysis

ologist with 15 years’ experience using US and six years
of experience using 2D-SWE. The pulse repetition fre-
quency, focal zone, gain and wall filter were adjusted to
obtain the optimal image for each participant. The ingui-
nal region was evaluated for the presence or recurrence
of hernia in all participants before evaluating the testes.
The examination of the testes was started with a B-mode
US. Each testicle was measured in three dimensions.
Testes volume was calculated by US device software
(testicular volume = width × height × depth × 10³ ×
0.523).
Quantitative 2D-SWE measurements were taken for
each participant in the transverse plane at the level of the
testicular hilus, using a magnified view as much as possi-
ble, by manually drawing the contours of the entire testis
structure using a free region of interest (ROI) in all testes
(fig 1). Three images of the testes were obtained by plac-
ing the US probe very lightly on the scrotum and then
the average of these three measurements was calculated
as the SWE value of the testis. The quantitative elasticity
values were measured in both meters/second (m/s) and
kilopascal (kPa).
The A, B, and C groups were compared in terms of the
2D-SWE values and testes volume. Among the testes in
Group A, we assessed the correlation between 2D-SWE
values and the duration of hernia and the period since the
inguinal hernia operation. Furthermore, 2D-SWE values
were compared among Group A patients according to the
type of hernia and surgical procedure, and also the grade
of hernia.
The Statistical Package for Social Sciences (SPSS)
version 22.0 (Inc., Chicago, Illinois, USA) software was
used to evaluate the data. The descriptive statistics were
expressed as mean, SD, minimum-maximum values,
frequency, and percentile. One-way analysis of variance
(ANOVA) and Student’s t-test were used to evaluate the
differences between the groups. The normality assump-
tion for the distribution of continuous variables was
tested using the Kolmogorov–Smirnov test. Pearson’s
correlation analysis was used to evaluate the relation-
ship between 2D-SWE values and the degree of hernia,
duration of hernia, and the period since the operation. A
p-value of less than 0.05 was accepted as statistically sig-
nificant.
Results
The mean age, testis volume and 2D-SWE values for
each of the three groups (A, B, and C) are summarized in
Table II. There was no statistically significant difference
in the mean age among the three groups (p=0.406). The
B-mode US of the testes were normal in all participants
and no recurrent–failed repair of IH in those who under-
went IH surgery was observed. There was a statistically
significant difference between both testes’ volume and
2D-SWE values (kPa and m/s) across the three groups.
A post hoc Tukey test indicated that the mean testes’ vol-
ume was significantly lower and 2D-SWE values were
significantly higher in group A comparing with Groups
B and C.
184 Mehmet Sedat Durmaz et al Assessment of testes with 2D-SWE in patients with operated inguinal hernia

There were no statistically significant differences be-
tween groups B and C in terms of testes volume and 2D-
SWE values (Table III).
There was no significant difference between the du-
ration of the hernia, the period passed after the IH op-
eration, testicular volume and SWE values according to
operation techniques (p>0.005). In addition, there was no
significant difference between the direct and indirect IH
with respect to age, duration of hernia, the period since
the IH operation, testicular volume and SWE values
(p>0.005, Table IV).
IH was located in proximal (grade I), middle (grade
II) and distal, almost in scrotum (grade III) in 14 (10.5%),
60 (54.5%) and 59 (44.4%) patients, respectively. There
was no statistically significant difference among these
three groups in terms of age, duration of hernia and the
period since the IH operation (p>0.005). There was a sig-
nificant positive correlation between the degree of her-
nia and 2D-SWE values of testes for both kPa and m/s
(r=0.643, p=0.001 and r=0.614, p=0.001, respectively).
There was no correlation between the grade of hernia and
testes volume (r =-0.093, p=0.285).
There was a positive correlation between the duration
of the hernia and 2D-SWE values for both kPa and m/s
(r=0.806 p=0.001 and r=0.736, p=0.001, respectively).
There was no significant correlation between the duration
of hernia and testis volume (r=-0.004, p=0.965) and be-
tween the time after the IH operation, 2D-SWE values for
both kPa and m/s (r=0.113, p=0.197 and r=0.081, p=0.351,
respectively) and testes volume (r=0.014, p=0.869).
Discussion
This study focuses on whether IH or surgery to repair
IH has any effect on testicular stiffness and testicular vol-
ume. We found that on the side of IH surgery the testes
were smaller and stiffer. The stiffness of testes was cor-
related with the grade of IH and duration of hernia but
no significant correlation was found between testicular
volume and grade of IH or the duration of the hernia.
Our results suggest that SWE can be used as an effec-
tive imaging technique for assessing testicular damage
in these patients. The significant correlation between 2D-
SWE values and the degree and the duration of the hernia
supports the hypothesis that the degree of testicular dam-
age changes according to the effect of pressure and the
duration of the pressure on the spermatic cord and vas
deferens.

Table II. The mean age, testes volume and 2D-SWE values of testes according to the group which we used.
Totaly Group A Group B Group C p
Age (year) 45.50±11.20 48.11±11.55 47.72±11.62 40.64±8.60 0.406
Volume (cm3) 15.07±4.21 13.60±3.83 15.71±4.58 15.97±3.80 0.001
Mean 2D-SWE kPa 7.04±1.81 8.54±2.08 6.22±1.08 6.30±0.90 0.001
value m/s 1.44±0.16 1.56±0.16 1.37±0.10 1.38±0.13 0.001
Two-dimensional shear-wave elastography (2D-SWE), inguinal hernia (IH), kilopascal (kPa) and meters/second (m/s).

Table III. The p value when the Group A, B and C were compared between each other acording to testis volume and 2D-SWE values.
A versus B A versus C A versus (B+C) B versus C
Testes volume 0.001 0.001 0.001 0.622
Mean 2D-SWE kPa 0.001 0.001 0.001 0.520
value m/s 0.001 0.001 0.001 0.607
Two-dimensional shear-wave elastography (2D-SWE), inguinal hernia (IH), kilopascal (kPa) and meters/second (m/s).

Table IV. The mean age, volume and 2D-SWE values of testes in Group A according to type of inguinal hernia and operation type
Direct IH Indirect IH p Lichtenstein TEP Kugel p
Age (year) 49.39±9.13 47.69±12.25 0.465 47.27±13.14 50.37±9.73 47.44±5.43 0.543
Duration of IH (month) 4.90±2.33 4.40±2.33 0.280 4.71±2.23 4.14±2.14 4.44±3.09 0.297
Time from surgery (month) 4.27±2.21 4.23±2.44 0.929 3.78±2.28 5.37±2.17 4.05±2.55 0.828
Testes volume (cm3) 14.24±3.51 13.39±3.92 0.268 13.61±4.01 13.41±3.86 13.95±3.06 0.893
Mean 2D-SWE kPa 9.02±2.51 8.38±11.90 0.122 8.67±2.03 8.26±1.29 8.46±3.28 0.620
value m/s 1.61±0.20 1.54±0.14 0.080 1.58±0.16 1.54±0.11 1.54±0.21 0.414
Two-dimensional shear-wave elastography (2D-SWE), inguinal hernia (IH), endoscopic totally extraperitoneal (TEP), kilopascal (kPa) and
meters/second (m/s).

Since the effect of pressure is removed after the op-
eration, the testicular damage can be prevented by return-
ing the testicular blood supply to normal [6,13]. Howev-
er, IH repair operations can lead to complications, which
may arise during surgery or during the postoperative pe-
riod that may cause damage to the spermatic cord and
vas deferens [14]. Moreover, a foreign body inflamma-
tory response to the mesh may result in testicular damage
by causing fibrosis in the tissues surrounding the sper-
matic cord and vas deferens [7-9,14]. The incidence of
testicular damage following IH surgery has been reported
to range from 1 to 14% [3]. Thus, both IH and IH repair
surgery can cause testicular damage [14].
Assessment of damage to the testicle with serum hor-
mone levels and sexual function is not reliable, especially
in cases where unilateral IH is present [12]. B-mode US
is the most commonly used initial imaging technique in
the diagnosis of testicular damage, but B-mode US can-
not provide sufficient information for a reliable diagnosis
in every case and, in particular, has low sensitivity for
evaluating testicular damage [14,15]. Testicular atrophy
and a decrease in testicular volume is an important in-
dicator of testicular damage in patients with IH and re-
paired IH, with a reported incidence of 0.5% [16]. Some
studies have evaluated the effectiveness of IH on testicu-
lar perfusion with Doppler US using a resistive index
(RI) [15,17]. These studies have shown the testis with IH
has a significantly higher RI [18] or that there is no sig-
nificant difference in the RI values [14,15,19-21]. Stud-
ies have also shown that there is no significant alteration
of testicular perfusion after IH operation using Doppler
US [21,22]. These results suggest that Doppler US and
RI are not reliable to evaluate testicular damage. In our
study, B-mode US aspect of the testicles was normal in
all cases. However, the statistically significant higher
testicular stiffness values obtained by 2D-SWE in those
patients who had undergone IH surgery showed testicular
damage. For these reasons, the efficacy of B-mode US
in the assessment of testicular damage is limited. In our
study, the testes volumes were significantly lower on the
side of IH surgery. This finding supports that testicular
volume may be used for the evaluation of testicular dam-
age. However, there was a lack of a significant correla-
tion between testicular volume and the grade of hernia
and duration of hernia, suggesting that taken separately,
the testicular volume cannot be used to assess testicular
damage reliably.
The effectiveness of 2D-SWE to evaluate stiffness for
estimating the degree of fibrosis as a result of testicular
damage has been shown in many studies [11,12,23]. In
these studies, SWE measurements were made in lim-
ited areas with ROI in constant small circle or square
Med Ultrason 2021; 23(2): 181-187 185
forms. This method of evaluation limits the repeatabil-
ity and reduces the reliability of SWE values [11]. We
used the average SWE values obtained by drawing the
whole testis parenchyma with free ROI in the transverse
plane. We consider this measurement method superior,
more applicable, reliable and more repeatable than other
methods used, in diagnosis and follow-up, under daily
out-patient practice. A single study evaluated the testicles
of patients with postoperative IH using compressive elas-
tography and showed that testicle stiffness was higher
on the operated side due to testicular damage [24]. The
shortcomings of compressive elastography are its opera-
tor dependency and the semi-quantitative nature of the
data, which may limit its reproducibility. However, 2D-
SWE is less operator-dependent, real-time, reproducible
and allows the quantitative evaluation of the target tissue
[11,12,23]. Elastic value E (kPa) is calculated using the
equation E =3ρ(m/s)² (ρ refers to the tissue density, with
approximated value in human body as 1 g/cm³ and m/s
refers to the shear wave propagation velocity) [25]. The
quantitative elasticity values both measured in kPa and
m/s revealed good diagnostic performance, but the speci-
fity of the standart deviation and area under the ROC
curve of the measurement in kPa were found significantly
higher compared to the measurement in m/s [26]. Since
both SWE values (m/s and kPa) could be obtained with
our US device, we measured both in our study. Thus, we
think that our findings can be compared with future stud-
ies using m/s or kPa.
Many studies have compared testicular damage aris-
ing from different operating techniques. In these studies,
testicular damage was compared with reference to sexual
function, testicular volume and hormone levels [7,27,28].
For IH repair surgery, especially with respect to surgical
techniques using mesh, there are studies that support [7-
9, 21] or do not [19,27,29] the testicular damage caused
by a marked foreign body inflammatory and immuno-
logical response to the mesh. In our hospital, mainly Li-
chtenstein, Kugel and TEP IH repair surgery techniques
are used. When these three surgical techniques were
compared, no significant difference was found between
testicular volume and 2D-SWE values and, therefore, the
degree of testicular damage. In addition, there was no
significant difference between the testicular volume and
2D-SWE values between operations on direct or indirect
hernia types.
There are some limitations in our study. In this study,
we included patients who had undergone an IH operation.
If we examined the testicles at regular intervals before
and after the operation with SWE, we could separately
evaluate the possible effects of IH and IH operation on
testicular damage; this may be the subject of future re-
186 Mehmet Sedat Durmaz et al Assessment of testes with 2D-SWE in patients with operated inguinal hernia
search. A biopsy was not performed in any of the patients
due to potential complications. Since the patients did not
present with infertility complaints, laboratory tests (e.g.,
semen analysis, hormone levels) were not requested. All
examinations were performed by an experienced radiolo-
gist, and this study did not evaluate the possibility of in-
terobserver variability.
In conclusion, our study suggests that the 2D-SWE
technique can be effectively used as an imaging method
for evaluating testicular damage, using objective numeri-
cal values in patients with postoperative IH. Long-term
follow-up studies with a larger number of patients are
required to confirm the efficacy of using 2D-SWE for
routine evaluation in this patient group.
Conflict of interest: none
References
1. Ballas K, Kontoulis T, Skouras CH, et al. Unusual findings
in inguinal hernia surgery: Report of 6 rare cases. Hip-
pokratia 2009;13:169-171.
2. Beddy P, Ridgway PF, Geoghegan T, et al. Inguinal Hernia
Repair Protects Testicular Function: A Prospective Study
of Open and Laparoscopic Herniorraphy. J Am Coll Surg
2006;203:17–23.
3. Kumar S, Dikshit P. Effect of bilateraly Lichtenstein hernia
repair on male gonadal function: A prospective study. Ind J
Sci Res Tech 2015;3:66-71.
4. Ozdamar MY, Karakus OZ. Testicular ischemia caused
by incarcerated inguinal hernia in infants: incidence,
conservative treatment procedure, and follow-up. Urol J
2017;14:4030-4033.
5. Bittner R, Schwarz J. Inguinal hernia repair: current surgi-
cal techniques. Langenbecks Arch Surg 2012;397:271–282.
6. Bulus H, Dogan M, Tas A, Agladıoglu K, Coskun A. The
effects of Lichtenstein tension-free mesh hernia repair on
testicular arterial perfusion and sexual functions. Wien Klin
Wochenschr 2013;125:96-99.
7. Hallen M, Westerdahl J, Nordin P, Gunnarsson U, Sandb-
lom G. Mesh hernia repair and male infertility: A retrospec-
tive register study. Surgery 2012;151:94-98.
8. Hallen M, Sandblom G, Nordin P, Gunnarsson U, Kvist U,
Westerdahl J. Male infertility after mesh hernia repair: A
prospective study. Surgery 2011;149:179-184.
9. Kolbe T, Hollinsky C, Walter I, Joachim A, Rulicke T. In-
fluence of a new self-gripping hernia mesh on male fertility
in a rat model. Surg Endosc 2010;24:445-461.
10. Sucullu I, Filiz AI, Sen B, et al. The effects of inguinal her-
nia repair on testicular function in young adults: a prospec-
tive randomized study. Hernia 2010;14:165–169.
11. Durmaz MS, Sivri M, Sekmenli T, Kocaoğlu C, Çiftci İ.
Experience of using shear wave elastography imaging in
evaluation of undescended testes in children: Feasibil-
ity, reproducibility, and clinical potential. Ultrasound Q
2018;34:206–212.
12. Hattapoğlu S, Göya C, Arslan S, et al. Evaluation of postop-
erative undescended testicles using point shear wave elas-
tography in children. Ultrasonics 2016;72:191-194.
13. Bansal VK, Krishna A, Manek P, et al. A prospective rand-
omized comparison of testicular functions, sexual functions
and quality of life following laparoscopic totally extra-per-
itoneal (TEP) and trans-abdominal pre-peritoneal (TAPP)
inguinal hernia repairs. Surg Endosc 2017;31:1478–1486.
14. Lal P, Bansal B, Sharma R, Pradhan G. Laparoscopic TEP
repair of inguinal hernia does not alter testicular perfusion.
Hernia 2016;20:429–434.
15. Aguilar-García J, Cano-González HA, Martínez-Jiménez
MA, de la Rosa-Zapata F, Sánchez-Aguilar M. Unilateral
Lichtenstein tension-free mesh hernia repair and testicular
perfusion: a prospective control study. Hernia 2018;22:479-
482.
16. Akbulut G, Serteser M, Yücel A, et al. Can laparoscopic
hernia repair alter function and volume of testis? Rand-
omized clinical trial. Surg Laparosc Endosc Percutan Tech
2003;13:377–381.
17. Chan MS, Teoh AY,  Chan KW, Tang YC,  Ng EK,  Leong
HT. Randomized double-blinded prospective trial of fibrin
sealant spray versus mechanical stapling in laparoscopic
total extraperitoneal hernioplasty. Ann Surg 2014;259:432–
437.
18. El-Awady SE, Elkholy AA. Beneficial effect of inguinal
hernioplasty on testicular perfusion and sexual function.
Hernia 2009;13:251–258.
19. Ramadan SU, Gokharman D, Tuncbilek I, et al. Does the
presence of mesh have an effect on the testicular blood flow
after surgical repair of indirect inguinal hernia? J Clin Ul-
trasound 2009;37:78–81.
20. Basha MAA, Saber S, El-Hamid M Abdalla AA, et al. As-
sessment of the testicular vascularity after inguinal herni-
otomy in children: a prospective color Doppler study. Acta
Radiol 2020;61:128-135.
21. Stula I, Druzijanic N, Srsen D, et al. Influence of inguinal
hernia mesh repair on testicular flow and sperm autoim-
munity. Hernia 2012;16:417–424.
22. Nath P, Dey S, Karim T, Jain A, Katiyar VK, Patel G. Study
of testicular perfusion after Lichtenstein hernioplasty in un-
complicated inguinal hernia. Int Surgy J 2018;5:1104-1110.
23. Erdogan H, Durmaz MS, Arslan S, et al. Shear Wave Elas-
tography Evaluation of Testes in Patients with Varico-
cele. Ultrasound Q 2020;36:64-68.
24. Zaharko VP, Nakonechnyy АY, Fedus VR. The significance
of qualitative compressive elastography in boys with hernia
inguinale.  Pharma Innovation 2016;5:26-28.
25. Durmaz MS, Arslan S, Özbakır B, et al. Effectiveness
of Shear Wave Elastography in the diagnosis of acute
pancreatitis on admission. Med Ultrason 2018;20:278-
284.
26. Youk JH, Son EJ, Park AY, Kim JA. Shear-wave elastogra-
phy for breast masses: local shear wave speed (m/sec) ver-
sus Young modulus (kPa). Ultrasonography 2014;33:34-39.

27. Aydede H, Erhan Y, Sakarya A, Kara E, Ilkgul O, Can M.
Effect of mesh and its location on testicular flow and sper-
matogenesis in patients with groin hernia. Acta Chir Belg
2003;103:607-610.
28. Palabiyik FB, Cimilli T, Kayhan A, Toksoy N. Do the ma-
nipulations in pediatric inguinal hernia operations affect
Med Ultrason 2021; 23(2): 181-187 187
the vascularization of testes? J Pediatr Surg 2009;44:788-
790.
29. Çelebi S, Yıldız A, Üçgül A, et al. Do open repair and dif-
ferent laparoscopic techniques in pediatric inguinal hernia
repairs affect the vascularization of testes? J Pediatr Surg
2012;47:1706-1710.
Original papers Med Ultrason 2021, Vol. 23, no. 2, 188-193
DOI: 10.11152/mu-2723

Congenital complete atrioventricular block from literature to clinical

approach – a case series and literature review
Liliana Gozar1, Claudiu Mărginean2, Amalia Făgărășan1, Iolanda Muntean1, Andreea
Cerghit-Paler3, Dorottya Miklósi3, Rodica Togănel1

1Department of Pediatric Cardiology, 2Department of Obstetrics and Gynecology, 3Pediatric Cardiology, Emergency
Institute for Cardiovascular Diseases and Transplantation, “George Emil Palade” University of Medicine, Pharmacy,
Science and Technology, Târgu Mureș, Romania

Abstract
Aim: Congenital atrioventricular block (CAVB) is an immunological condition, secondary to the transfer of maternal Ig
G antibodies from seropositive mothers. Although the presence of these antibodies is high among pregnant women, the preva-
lence of this fetal pathology is low. The aim of this paper is to analyze a series of cases with intrauterine diagnosis of CAVB
and to present their follow-up protocol. Material and method: In the period between 2013-2020, five fetuses were diagnosed
and followed up in the Pediatric Cardiology Clinic. In each of the cases, assessment of the hemodynamic status was done by
calculation of the fetal cardiovascular profile score (CVPS). In the last cases the follow-up protocol was supplemented with
longitudinal speckle tracking evaluation of the ventricular function. Results: In the present series, intrauterine death occurred
in one case; in another case resumption of atrioventricular conduction was observed. Epicardial pacemaker implantation was
required in three of the patients. Conclusion: Completing the evaluation of ventricular function with the longitudinal speckle
tracking method in fetuses and newborn patients with congenital atrioventricular block may play an important role in establish-
ing therapeutic behavior.
Keywords: fetal; congenital atrioventricular block; echocardiography

Introduction auto­immune etiology and is secondary to the transpla-

Congenital atrioventricular block can be defined as
an impairment of atrioventricular conduction until com-
plete cessation. Thus, ventricular activity is independent
of atrial activity, causing a significant decrease in the fe-
tal heart rate, which can lead to intrauterine heart failure
and even death. It is known that this fetal pathology is of
Received 17.07.2020  Accepted 17.10.2020
Med Ultrason
2021, Vol. 23, No 2, 188-193
Corresponding author: Claudiu Mărginean
Obstetrics and Gynecology Department,
“George Emil Palade”
University of Medicine,
Pharmacy, Science and Technology
38 Gheorghe Marinescu street,
540139, Târgu Mureș, Romania
Phone: +40-722505311
E-mail: marginean.claudiu@gmail.com
cental passage of maternal antibodies, especially SSA /
Ro and SSB / La [1,2].
Postnatal diagnosis of complete atrioventricular
block (CAVB) is made by standard 12 lead electrocar-
diography, which for technical reasons cannot be per-
formed on the fetus. Thus, echocardiography remains the
main diagnostic tool of this intrauterine rhythm disorder.
Using the conventional M mode, obtained by performing
a section that includes both the atrial and the ventricu-
lar walls, atrio-ventricular activity and dissociation may
be analyzed. Simultaneous recording of pulsed Doppler
waves in an artery and a vein, respectively the aorta and
the superior vena cava, has proven useful in the diagnosis
of this arrhythmia. In this registration the atrial activity is
rendered by the atrial reverse in the superior vena cava,
and the ventricular activity by the aortic flow. This meth-
od is useful in determining the atrioventricular interval-
the correspondent of the PR interval on the electrocardio-

Fig 1. M mode recording: slow ventricular rate and complete
dissociation between atrial and ventricular contractions.
gram, and it is of crucial importance to be monitored in
pregnant women with a positive serology [3].
Given the very low frequency of this congenital pa-
thology, the controversial data currently existing in the
literature on treatment and evolution, we consider it im-
portant to share our experience on this topic, by present-
ing the cases from our clinical experience, respectively
their immediate and long-term evolution.
Material and methods
In the period between 2013-2020, five fetuses were
diagnosed and followed with CAVB in the Pediatric Car-
diology Clinic from the Emergency Institute of Cardio-
vascular Diseases and Heart Transplant of Târgu-Mureș.
In 4 of the cases mothers were asymptomatic in terms
of autoimmune disease, but with positive serology with
anti Ro/SSA antibodies, and in one case the mother was
known to have Sjögren’s syndrome. In each case, the
Fig 3. Doppler flow in the ductus venosus with atrial reversal
Med Ultrason 2021; 23(2): 188-193 189
Fig 2. Pulsed Doppler recording in the ascending aorta and
superior vena cava: dissociation between atrial and ventricular
contraction
following echocardiographic diagnostic protocol was
performed: establishing the anatomy of the fetal heart
and excluding any structural abnormalities through a
segmental sequential analysis; establishing the relation-
ship between atrial and ventricular activity, using M
mode (fig 1) and pulsed Doppler recorded in an artery
and a vein, most commonly in the aorta and superior
vena cava (fig 2); determination of the atrial and ven-
tricular rates; hemodynamic evaluation of the fetus by
calculating the fetal cardiovascular profile score (CVPS)
[4,5].
CVPS was determined taking into account each of the
5 echocardiographic elements: 1. hydrops (effusion -1;
skin edema -2), 2. venous Doppler pulsation (atrial rever-
sal in ductus venosus -1 fig3; umbilical vein atrial pulsa-
tion - 2), 3. umbilical artery Doppler flow (absent end-
diastolic flow -1, reversed diastolic flow -2), 4. heart size
(cardiothoracic ratio 0.35 -50% -1; cardiothoracic area>
50% -2 - fig 4), 5. cardiac function (RV / LV shortening
Fig 4. Cardiothoracic ratio
190 Liliana Gozar et al Congenital complete atrioventricular block from literature to clinical approach

fraction <0.28 or holosystolic tricuspid regurgitation -1,

mitral regurgitation or monophasic ventricular diastolic
filling -2) [4].
To assist in the quantification of the ventricular func-
tion, in two of the cases we applied speckle tracking,
by using a Qlab 13 software. Considering that the new
speckle tracking analysis software allows the delimita-
tion of the cardiac cycle without the need for a simulta-
neous electrocardiographic recording, the marking of the
cardiac cycle being based upon the closure and aperture
of the mitral valve (fig 5), we applied this method of as- Fig 5. Delimitation of the cardiac cycle for speckle tracking
sessing ventricular function in fetuses and newborn pa- analysis, using M mode
tients with atrioventricular block. The analysis is made
in a four chamber view, recorded with a frame rate of at
least 80Hz (fig 6, fig 7)

Results

The clinical characteristics and the evolution of the

cases are presented in Table I.
The therapeutic approach of the fetuses with com-
plete atrioventricular block consisted of oral Dexametha-
sone, 4mg/day in a single dose for a period of 6 weeks.
If an improvement was obtained, more precisely if the Fig 6. Four chamber view: longitudinal strain curves, in the
regression of the complete atrioventricular block to a fetal period
second-degree atrioventricular block or even to periods
of sinus rhythm was observed, the treatment was con-
tinued until the end of the pregnancy. If no enhancement
was detected in atrioventricular conduction, dexametha-
sone therapy was discontinued. In the case of congenital
AVB with low ventricular rate, we initiated betamimetic
therapy (Salbutamol) for a period of time. No side effects
of corticosteroid therapy were observed in any of the pa-
tients after birth.
Postnatal therapy involved the following: administra-
tion of Dopamine and Adrenaline- if the heart rate was
extremely low and there were any signs of heart failure;
followed by implantation of a pacemaker after a few
days, with epicardial pacing wires placed on the left ven- Fig 7. Four chamber view: longitudinal strain curves, in the
tricle and device programming in VVI mode. Therapy neonatal period
and evolution of the cases are shown in table II.
Discussion
Table I. Clinical characteristics and the evolution of the cases
Case Gestational Heart Fetal Hydrops The first case of atrioventricular block was described
no. age rate cardiovascular in 1901 by Morqui. In 1972 the first studies were pub-
(weeks) profile score lished, which reported the incidence of this pathology to
1 23 50 bpm 5 Yes be about 1 in 22,000 live births. [6,7].
2 24 68 bpm 9 No Regarding the prevalence of this bradyarrhythmia
among newborns from mothers with Ro / SSA antibod-
3 26 60 bpm 9 No
ies, the first prospective study was performed by Brucato,
4 24 50 bpm 7 No
who followed 100 seropositive women, respectively 118
5 24 70 bpm 9 No pregnancies. From the total number of pregnancies fol-

Table II. Therapy and evolution of the cases
Case no. Intrauterine treatment Intrauterine evolution
1 Dexamethasone Intrauterine death
2 Dexamethasone for 15 weeks Partial recovery of
atrioventricular conduction implant at 14 days of age
3 Dexamethasone for 6 weeks Unchanged
4 Dexamethasone and Salbutamol Slow unfavourable
for 6 weeks
Hydroxychloroquine
5 Dexamethasone for 6 weeks Unchanged
lowed, 2 fetuses were diagnosed with congenital com-
plete heart block (CCHB), thus concluding that the
prevalence of this pathology was low, up to 1.8-2% [8].
According to a study published by Tsubai, the risk of
atrioventricular block is higher in the presence of an in-
creased titre of maternal antibodies [9].
The occurrence of this major conduction disorder is
secondary to cardiac injury resulting from the transpla-
cental passage of maternal IgG antibodies. Although the
number of seropositive mothers is high, the frequency of
complete atrioventricular block in the fetus is still low
[1,10].
A study published by Askanase et al in 2002 high-
lights the progressive nature of the rhythm disorders
in fetuses of seropositive mothers. Maternal antibodies
cross the placenta and produce inflammatory changes in
the fetal myocardium, followed by changes in myocytes
and fibrosis. In certain situations, the process continues
even after the disappearance of maternal antibodies, as
evidenced by the progressive postnatal character: evolv-
ing from a minor disorder, prolonged PR interval high-
lighted at birth, to the appearance of complete atrioven-
tricular block during childhood [10].
Prenatal diagnosis of fetal arrhythmias is difficult and
involves an indirect assessment of fetal heart activity. It
is mandatory that the diagnostic protocol includes both
M-mode and pulsed Doppler examination, due to the fact
that a second-degree atrioventricular block may be diffi-
cult to differentiate from blocked atrial extrasystoles [3].
Intrauterine fetal demise was estimated at 6% of fe-
tuses with this pathology. After birth, death is rare and
often due to cardiomyopathy associated with atrioven-
tricular block [11].
Regarding the therapy of this pathology, the data
published in literature are limited and somewhat con-
troversial. The most common treatment options include
steroids and beta-stimulation, immunoglobulins and
hydroxychloroquine. Administration of dexamethasone
may improve pleuro-pericardial effusions and intra­
Med Ultrason 2021; 23(2): 188-193 191
Postnatal treatment After birth evolution
- -
Monitoring and pacemaker Favourable
Pacemaker implant at 7 days Favourable
of age
Dopamine Pacemaker implant Favourable
at 6 days of age
No Favourable
uterine evolution [12]. Studies published so far in the
literature do not clarify the role of steroid therapy in
first-degree atrioventricular block. It is difficult to prove
the progression of first-degree atrioventricular block (di-
agnosed echocardiographically in M-mode or by pulsed
Doppler as a prolongation of the atrioventricular interval)
to second-degree atrioventricular block and even more
difficult to prove the role of corticosteroid therapy in this
setting [13-15].
Most fetuses diagnosed with second-degree atrioven-
tricular block have developed complete atrioventricular
block [10,16] and there are extremely few cases in which
an improvement in atrioventricular conduction has been
achieved, once CAVB has been diagnosed. Data pub-
lished in the literature regarding corticosteroid therapy
in the setting of second-degree atrioventricular block are
also controversial. Some studies report the recovery of
atrioventricular conduction after steroid therapy in the
context of a second-degree block [12,17,18].
In a study consisting of 214 cases with high-degree
congenital heart block, published by Levesque et al [19],
the role of corticosteroids in the recovery of atrioventric-
ular conduction was not proven. In the aforementioned
study a case of complete atrioventricular block that re-
covered atrioventricular conduction without treatment
was also reported [19]. Of the 5 cases we followed, case
2 had a similar evolution, with the recovery of atrioven-
tricular conduction. At the age of 24 weeks, the fetus had
a complete atrioventricular block, but at gestational age
of 32 weeks, a partial resumption of atrioventricular con-
duction was observed. At birth, Holter ECG monitoring
revealed long periods of sinus rhythm alternating with
short periods of third-degree atrioventricular block, with
supra-Hisian escape rhythm. Although the newborn was
asymptomatic, at 4 weeks of age, pacemaker implanta-
tion was performed. Upon follow-up, currently at the
age of 6, the patient has long episodes of sinus rhythm
with short periods of pacemaker rhythm. This evolution
with the regression of a complete atrioventricular block
192 Liliana Gozar et al Congenital complete atrioventricular block from literature to clinical approach
is unusual and extremely rarely described in the literature
[20]. To the best of our knowledge, this is the second case
reported in the literature, in which an important recovery
of atrioventricular conduction was obtained.
Cuneo et al draws attention to the existence of a tran-
sitional period until the appearance of complete atrioven-
tricular block and emphasizes on the importance of in-
trauterine diagnostics and early initiation of therapy [21].
The last of the presented cases showed a favourable
evolution throughout the pregnancy, with a heart rate
of 70 bpm, without hemodynamic changes. During the
neonatal period, heart rates of 80 bpm were maintained,
displaying an aspect of complete atrioventricular block,
with a supra-Hisian escape rhythm. Furthermore, the as-
sessment of ventricular function through speckle track-
ing, performed postnatal, showed very good longitudinal
strain parameter values, but with a slight asynchrony in
the contraction of the ventricular segments.
Despite numerous controversial data regarding cor-
ticosteroid therapy, it should not be discouraged [18].
Moreover, Sonnesson et al recently published a pro-
spective study and concluded that steroid therapy has a
positive effect on both the intrauterine and postnatal evo-
lution, on the rate of the escape rhythm and could help in-
crease the age at which the pacemaker is implanted [22].
Given the controversy surrounding intrauterine ther-
apy, the monitoring the fetal heart with complete AVB is
of extreme importance. Risk factors for an unfavourable
evolution are: ventricular rate below 55 bpm, the appear-
ance of endomyocardial fibroelastosis and low shortening
fraction. With the help of the speckle tracking method,
the ventricular function can be evaluated more faithfully,
as well as any early evidence of a possible ventricular
dysfunction. In case 4, although the heart rate and the
fetal cardiovascular profile score (CVPS) were low, hav-
ing a severely altered ejection fraction, global and seg-
mental longitudinal strain and strain rates showed values
close to normal, indicating a favourable prognosis [23].
Furthermore, in the last case, the good ventricular func-
tion, with very good longitudinal global and segmental
strain values, helped us make the decision to delay the
implantation of the pacemaker. It is known that complete
atrioventricular block leads in time to the appearance of
dilated cardiomyopathy, thus evaluation of ventricular
function through speckle tracking can bring useful infor-
mation to help therapeutic decision-making.
None of the 5 patients that were born presented any
side effects from the administered corticosteroid therapy,
and their evolution following the pacemaker implanta-
tion was favourable, with the normalization of echocardi-
ographic parameters of the left ventricle; proving that the
placement of epicardial electrodes on the left ventricle in
pediatric patients with complete atrioventricular block is
extremely important [24].
The limitations of this study are given primarily by
the low number of cases with complete congenital atrio-
ventricular block and the fact that the speckle tracking
evaluation method has not yet entered into clinical prac-
tice, thus proving its importance in the functional evalu-
ation of the fetal heart.
Conclusions
Because the prevalence of this pathology is very low,
each clinical case must be analyzed in the light of cur-
rent data available in the literature. The effectiveness of
corticosteroid therapy is uncertain. Echocardiographic
follow-up and completion of the evaluation protocol with
longitudinal, global and segmental speckle tracking has
an important role in the therapeutic decision.
Conflict of interest: none
References
1. Clancy RM, Kapur RP, Molad Y, Askanase AD, Buyon JP.
Immunohistologic evidence supports apoptosis, IgG depo-
sition, and novel macrophage/fibroblast crosstalk in the
pathologic cascade leading to congenital heart block. Ar-
thritis Rheum 2004;50:173-182.
2. Bordachar P, Zachary W, Ploux S, Labrousse L, Haissa-
guerre M, Thambo JB. Pathophysiology, clinical course,
and management of congenital complete atrioventricular
block. Heart Rhythm 2013;10:760-766.
3. Hunter LE, Simpson JM. Atrioventricular block during fe-
tal life. J Saudi Heart Assoc. 2015;27:164-178.
4. Falkensammer CB, Paul J, Huhta JC. Fetal congestive heart
failure: correlation of Tei-index and Cardiovascular-score.
J Perinat Med 2001;29:390-398.
5. Huhta JC. Guidelines for the evaluation of heart fail-
ure in the fetus with or without hydrops. Pediatr Cardiol
2004;25:274-286.
6. Michaëlsson M, Engle MA. Congenital complete heart
block: an international study of the natural history. Cardio-
vasc Clin 1972;4:85-101.
7. Kertesz NJ, Fenrich AL, Friedman RA. Congenital complete
atrioventricular block. Tex Heart Inst J. 1997;24(4):301-
307.
8. Brucato A,  Frassi M,  Franceschini F,  et al. Risk of con-
genital complete heart block in newborns of mothers with
anti-Ro/SSA antibodies detected by counter immunoelec-
trophoresis: a prospective study of 100 women. Arthritis
Rheum 2001;44:1832-1835.
9. Tsuboi H, Sumida T, Noma H, et al. Maternal predictive
factors for fetal congenital heart block in pregnant moth-
ers positive for anti-SS-A antibodies. Mod Rheumatol
2016;26:569-575.

10. Askanase AD, Friedman DM, Copel J, et al. Spectrum and
progression of conduction abnormalities in infants born
to mothers with anti-SSA/Ro-SSB/La antibodies. Lupus
2002;11:145-151.
11. Clowse MEB, Eudy AM, Kiernan E, et al. The prevention,
screening and treatment of congenital heart block from neo-
natal lupus: a survey of provider practices. Rheumatology
(Oxford) 2018;57(suppl_5):v9-v17.
12. Jaeggi ET, Fouron JC, Silverman ED, Ryan G, Smallhorn
J, Hornberger LK. Transplacental fetal treatment improves
the outcome of prenatally diagnosed complete atrioven-
tricular block without structural heart disease. Circulation
2004;110:1542-1548.
13. Friedman DM, Kim MY, Copel JA, et al; PRIDE Investi-
gators.  Utility of cardiac monitoring in fetuses at risk for
congenital heart block: the PR Interval and Dexametha-
sone Evaluation (PRIDE) prospective study, Circulation
2008;117:485-493.
14. Jaeggi ET, Fouron JC, Silverman ED, Ryan G, Smallhorn
J, Hornberger LK. Transplacental fetal treatment improves
the outcome of prenatally diagnosed complete atrioven-
tricular block without structural heart disease. Circulation
2004;110:1542-1548.
15. Rein AJ, Mevorach D, Perles Z, et al. Early diagnosis and
treatment of atrioventricular block in the fetus exposed to
maternal anti-SSA/Ro-SSB/La antibodies: a prospective,
observational, fetal kinetocardiogram-based study. Circula-
tion 2009;119:1867-1872.
16. Buyon JP, Hiebert R, Copel J, et al. Autoimmune-associated
congenital heart block: demographics, mortality, morbidity
and recurrence rates obtained from a national neonatal lu-
pus registry. J Am Coll Cardiol 1998;31:1658-1666.
17. Saleeb S, Copel J, Friedman D, Buyon JP. Comparison of
treatment with fluorinated glucocorticoids to the natural
Med Ultrason 2021; 23(2): 188-193 193
history of autoantibody-associated congenital heart block:
retrospective review of the research registry for neonatal
lupus. Arthritis Rheum 1999;42:2335-2345.
18. Ciardulli A, D’Antonio F, Magro-Malosso ER, et al. Ma-
ternal steroid therapy for fetuses with second-degree im-
mune-mediated congenital atrioventricular block: a system-
atic review and meta-analysis. Acta Obstet Gynecol Scand
2018;97:787-794.
19. Levesque K, Morel N, Maltret A, et al. Description of
214 cases of autoimmune congenital heart block: Results
of the French neonatal lupus syndrome. Autoimmun Rev
2015;14:1154-1160.
20. Bordachar P, Zachary W, Ploux S, Labrousse L, Haissa-
guerre M, Thambo JB. Pathophysiology, clinical course,
and management of congenital complete atrioventricular
block. Heart Rhythm 2013;10:760-766.
21. Cuneo BF, Ambrose SE, Tworetzky W. Detection and suc-
cessful treatment of emergent anti-SSA-mediated fetal
atrioventricular block. Am J Obstet Gynecol 2016;215:527-
528.
22. Sonesson SE, Ambrosi A, Wahren-Herlenius M. Benefits of
fetal echocardiographic surveillance in pregnancies at risk
of congenital heart block: single-center study of 212 anti-
Ro52-positive pregnancies. Ultrasound Obstet Gynecol
2019;54:87-95.
23. DeVore GR, Cuneo B, Klas B, Satou G, Sklansky M. Com-
prehensive Evaluation of Fetal Cardiac Ventricular Widths
and Ratios Using a 24-Segment Speckle Tracking Tech-
nique. J Ultrasound Med 2019;38:1039-1047.
24. Tissot C, Aggoun Y, Rimensberger PC, et al. Left ventricu-
lar epicardial VVI pacing for a congenital complete heart
block with severe myocardial dysfunction: shall epicar-
dial pacing wires be positioned left? Int J Cardiol 2007;
116:e7-e9.
Review Med Ultrason 2021, Vol. 23, no. 2, 194-202
DOI: 10.11152/mu-2652

Ultrasonography techniques in the preoperative diagnosis of parotid

gland tumors – an updated review of the literature
Sebastian Stoia1, Grigore Băciuț1, Manuela Lenghel2, Radu Badea3, Mihaela Băciuț1, Simion
Bran1, Cristian Dinu1

1Department of Maxillofacial Surgery and Implantology, 2Department of Radiology, 3Department of Medical Imaging
“Prof. Dr. Octavian Fodor” Regional Institute of Gastroenterology, “Iuliu Hațieganu” University of Medicine and
Pharmacy, Cluj-Napoca, Romania

Abstract
Preoperative diagnosis of parotid tumors plays a crucial role in selecting and planning the surgical treatment. Ultrasound
(US) with its modern techniques can contribute to the differential diagnosis of parotid tumors. This paper aims to achieve a
comprehensive review of the ultrasound techniques used in the differential diagnosis of parotid tumors, based on the latest
literature data. Considering that most parotid gland tumors are located in the superficial lobe, US is frequently the first imaging
technique used for the diagnosis of parotid tumors. Sonoelastography can provide additional data on the elasticity of parotid
tumors, but there is an overlap between malignant and benign parotid tumors parameters. Contrast-enhanced ultrasound adds
value to conventional ultrasound and allows a more complete characterization of parotid tumors. Many authors have reported
promising results using contrast-enhanced ultrasound in the differential diagnosis of parotid tumors. Multiparametric ultra-
sound with a careful and systematic approach usually allows an accurate differential diagnosis of parotid tumors.
Keywords: ultrasound; elastography; contrast-enhanced ultrasound (CEUS); diagnosis; parotid gland tumors

Introduction Benign parotid tumors are frequently located in the

Salivary glands tumors account for 2 to 6% of all
head and neck tumors. Among these, 70-80% are located
in the parotid glands, 70-80% are benign and 20-30%
are malignant. Parotid tumors have a wide histologi-
cal variety, the pleomorphic adenoma (PMA) being the
most common benign tumor (60-80%), followed by the
Warthin tumor (WT) (15-30%). Mucoepidermoid carci-
noma is the most encountered malignant tumor of the pa-
rotid glands (30%) [1-4].
Received 01.06.2020  Accepted 25.09.2020
Med Ultrason
2021, Vol. 23, No 2, 194-202
Corresponding author: Manuela Lenghel
Department of Radiology, Faculty of
Medicine, “Iuliu Hațieganu” University of
Medicine and Pharmacy,
2-4 Clinicilor street,
400006, Cluj-Napoca, Romania
Phone: +40740423025
E-mail: pop.lavinia@umfcluj.ro
superficial lobe, being well delimited, mobile, do not
modify the covering skin and have a firm-elastic consist-
ency. Signs and symptoms such as: pain, rapid growth,
invasion of surrounding tissues, ulceration of the overly-
ing skin, facial nerve palsy, otalgia, cervical adenopathy
and weight loss should raise malignancy suspicion. One
third of malignant tumors have a clinical behavior similar
to benign tumors [2,3].
Imaging investigations play an essential role in the
preoperative differential diagnosis of parotid gland tu-
mors providing information about the specific tumor
localization in the gland and the relationships with sur-
rounding structures. The preoperative differentiation be-
tween benign and malignant lesions and the knowledge
of the histological type of the parotid tumors is extremely
important for the surgical planning: malignant tumors re-
quire extensive surgery in accordance with oncological
principles, while in the case of benign tumors less inva-
sive surgical treatment is performed, which may differ
depending on the tumor histological type [5].

Fig 1. a) Grayscale appearance for pleomorphic adenoma: well-defined margins, lobulated contour, acoustic enhancement, subcap-
sular location in the superficial lobe without capsular effraction; b) grayscale appearance for Warthin tumor: well defined lesion with
necrotic areas; c) low-grade malignant lesion, well defined, irregular contour, calcification and perilesional round hypoechoic lymph
node (histological proven as a collision tumor of acinic cell carcinoma and salivary duct carcinoma)
Currently, cross-sectional imaging investigations, in-
cluding modern US techniques, are routinely used and
studied for the differential diagnosis of parotid tumors,
with promising results published in literature. There are
a few papers published focused on the evaluation of US
techniques used in the differential diagnosis of parotid
tumors. This paper aims to achieve an updated, objective
and comprehensive review of the US techniques used in
the diagnosis of parotid tumors based on the latest litera-
ture data.
Diagnostic value of ultrasound techniques (US)
Current US techniques can contribute to the differ-
ential diagnosis of parotid lesions. In 80-90% of cases
tumors are localized in the superficial lobe [6-8]. Usually
US is the initial method of choice for parotid tumors di-
agnosis, due to a large number of advantages (low-cost,
accessible, non-invasive and non-irradiating technique).
The diagnostic accuracy is very high for the differen-
tiation between tumors and other glandular pathologies
such as cystic, infectious or inflammatory lesions. US
can make the differential diagnosis between benign and
malignant tumors, can locate the intra- or extra-glandular
lesions and is very helpful for the guidance of fine nee-
dle aspiration biopsy (FNAB). The technique presents a
series of limitations such as: dependence of the opera-
tor, lack of standardized sections, low specificity in as-
sessment of parotid tumors histological type, difficulty
in evaluating tumors located in the deep parotid lobe or
posterior to bone structures [1,5,7-14].
Grayscale and Doppler criteria
The following parameters can be assessed with the
conventional B-mode US: axial diameter, borders, shape,
acoustic enhancement, echotexture, echogenicity and the
presence of anechoic areas. Doppler US provides addi-
Med Ultrason 2021; 23(2): 194-202 195
tional information related to vascularization of parotid
tumors [7,5,15]. On US, pleomorphic adenoma (PMA)
typically presents well circumscribed margins, predomi-
nantly homogeneous echotexture, acoustic enhancement,
internal focal calcifications (rarely), anechoic / cystic
degeneration areas. A Warthin tumor (WT) usually has
well-defined margins, internal cystic changes, hyper-
echoic internal septation, a more inhomogeneous struc-
ture and acoustic enhancement. Small, low-grade malig-
nant tumors may appear well defined, homogeneous and
hypoechoic, while large, high-grade malignant tumors
present irregular, poorly defined margins, heterogeneous
echotexture; extra-glandular spread and pathologic cervi-
cal adenopathy can be identified (fig 1, fig 2) [4,8,16,17].
On Doppler US, WTs and malignancies have a rich
capsular and/or internal vascularization. PMA shows re-
duced vascularization [4,8,16,17].
Fischer et al in 2010 describes a predominant distri-
bution of the vessels in the periphery rather than in the
center for PMA, compared to the WT, in which the over-
all tumor vascularization is much higher, without signifi-
Fig 2. Illustration of malignant features identified in parotid
tumors: a) infiltrative contour, inhomogeneous lesion; b) ext-
racapsular spread (the arrow is showing the disruption of the
parotid capsule)
196 Sebastian Stoia et al Ultrasonography techniques in the preoperative diagnosis of parotid gland tumors
Fig 3. Doppler US: a) pleomorphic adenoma with reduced vascularization; b) highly vascularized tumor (histological proven – a
Warthin tumor) with capsular and internal vascularization; c) anarchic distribution of vessels in the case of a malignant lesion (his-
tological proven – salivary duct carcinoma).
cant differences between the central and the peripheral
part of the tumor (p<0.05) (fig 3) [16].
Rong et al (2014) was able to differentiate PMA from
WT based on B-mode Doppler US characteristics [13].
Rzepakowska et al (2017) managed to differentiate PMA
from WT [sensitivity (Se) 61.5%, specificity (Sp) 81.5%,
accuracy 73.1%] as well as malignant tumors from be-
nign lesions (Se 60%, Sp 95.2%, accuracy 90.3%) using
B-mode and Doppler US [7].
Gerwel et al (2015) and Kovačević (2010) et al
showed that an ill-defined margin is the most representa-
tive US parameter for malignant tumors (Sp 86%, Se
37.5%) while acoustic enhancement is a benign charac-
teristic; simultaneous evaluation of all US parameters
can increase the method sensitivity [18,19].
In the literature a diagnostic accuracy between 61.8-
64.7%, a Se between 38.9-88% and a Sp between 67.4-
90.1% were reported using only the conventional US pa-
rameters in differentiating malignant from benign parotid
tumors [18,20,21].
Several authors propose that in cases of benign le-
sions confined to the superficial parotid lobe suggested
by US and confirmed by the cytological result of FNAB,
no further investigations are required [4,9,10,14].
Cheng et al (2019) showed that the US assessment of
the minimum distance between the parotid fascia and the
parotid tumor (MFTD) with a cutoff value of 2.4 mm can
be used to differentiate the location of benign tumors in
the parotid gland; distance of less than 2.4 mm correlates
with the tumor localization in the superficial lobe of the
parotid gland [22]. On the other hand, for a more precise
and objective localization of tumors in the parotid glands
with the help of the US, a schematic diagram based on
marking the tumor location in relation to the external ear
structures was proposed. The use of this diagram helps to
standardize and organize the US preoperative report [23].
There is still a considerable overlap between be-
nign and low-grade malignancies [8]. Therefore, differ-
entiating between malignant and benign tumors based
on B-mode and Doppler US is difficult and can lead to
misdiagnosis. Zhang et al (2018) have showed that the di-
agnostic accuracy (96.1 vs 82.4%) of DWI-MRI was in-
creased when ADC values were used in conjunction with
US Doppler flow imaging features to differentiate parotid
tumors compared to single usage of ADC values [24].
Conventional US is a non-invasive examination with
a facile and quick access for the clinicians, often being
the first recommended for patients with parotid lesions.
US completes the clinical exam, allowing in the majority
of the cases a general evaluation of the parotid tumor and
an overview in establishing the clinical final diagnosis. In
addition, it has high Se and Sp for the differentiation from
other non-neoplastic lesions of the parotid glands. US al-
lows the characterization of malignant and benign tumors
based on specific US criteria for each tumor entity, but
has a low Se (38.9-88%) in the differential diagnosis of
parotid tumors, due to an overlap between benign and
low-grade malignant tumors, according to most literature
data [5,8,15,16,18,20,21,25-28]. Mansour et al (2012)
suggested in their study that vascular pattern and acous-
tic enhancement are the most significant parameters in
differentiating PMA from WT, when using the US [29].
In addition to conventional US, modern techniques
are currently available and allow the assessment of elas-
ticity or other biomechanical tissue properties. These
techniques are reunited under the name of multiparamet-
ric ultrasonography (MPUS), adding value to the con-
ventional US and completing some of the disadvantages
of the method. MPUS includes the ultrasound elastogra-
phy (USE) with the two techniques- strain elastography
(SE) and shear wave elastography (SWE) as well as the
contrast-enhanced ultrasound (CEUS) [5,8,10,20,30-32].
Ultrasound elastography
Ultrasound elastography (USE) is a non-invasive
technique that allows a quantitative or qualitative evalu-
ation of the soft tissue elasticity.

The use of USE in the differential diagnosis of parot-
id tumors is controversial. At present, the effectiveness of
USE, as the only examination technique, is not proven in
the differential diagnosis of parotid tumors. Many studies
have shown that malignant tumors have higher stiffness
values than benign ones and PMAs are in generally stiff-
er than WTs, but there is a large overlap between these
findings [10,11,26,28,33].
USE brings new information on the characterization
of parotid tumors, but nevertheless it has been shown that
adding it to conventional US does not improve signifi-
cantly the differential diagnosis between parotid tumors
[29]. Initial results looking purely at USE techniques in
the parotid glands diagnosis have been disappointing.
Larger prospective studies are needed to determine the
value of this investigation. USE recorded a pooled Se of
67%, a Sp of 64% and a diagnosis odds ratio of 8.00 in
a meta-analysis [34] that evaluated the value of this in-
vestigation in the differential diagnosis of parotid tumors
[9,26,34,35,36].
Dumitriu et al (2010) could not identify in their study
a sonoelastographic pattern characteristic for the PMA
[37]. Although the same authors found a statistically
significant difference between the elastography scores
of malignant tumors compared to benign lesions, the de-
tailed analysis was not possible to establish a cutoff point
for this scores, more likely due to the high stiffness of
PMAs (fig 4) [38].
Fig 4. Strain elastography technique showing increased stiffness in both pleomorphic adenoma (a) and a malignant tumor (b)
Fig 5. Shear-wave elastography. Illustration of increased stiffness of pleomorphic adenoma (a) compared with Warthin tumor (b)
Med Ultrason 2021; 23(2): 194-202 197
Mansour et al (2015) demonstrated that PMA pre-
sents significantly increased stiffness compared to mono-
morphic adenoma, malignant tumors or cystic lesions;
however, this criterion has not been shown to be effective
in differentiating between malignant and benign tumors
[32].
On the other hand, many authors obtained promising
results regarding the use of USE in the differential diag-
nosis of parotid tumors.
Mansour et al in 2012, using SWE with the mean
ARFI velocity values, obtained a statistically significant
difference (p<0.003) between PMA and WT (2.23±0.53
m/s vs 2.58±0.77 m/s) [29].
According to Bhatia et al (2012) the mean shear mod-
ulus of benign lesions was 18.3 kPa which overlapped
considerably over malignant lesions (13.5 kPa). Howev-
er, PMAs shown to be significantly stiffer compared with
WTs (22.5 kPa versus 16.9 kPa) [28] (fig 5).
Wierzbicka et al (2013) used USE to differentiate be-
nign from malignant parotid tumors. Malignant tumors
had a mean stiffness of 146.6 kPa as opposed to benign
tumors that had a mean stiffness of 88.7 kPa. The results
were statistically significant (p <0.001) and the authors
concluded that adding USE to US improves the differen-
tial diagnosis of parotid tumors [39].
Cortcu et al evaluated the USE in discrimination of
malignant and benign parotid tumors based on strain ra-
tio (SR). Median SR value for benign tumors was 1.11
198 Sebastian Stoia et al Ultrasonography techniques in the preoperative diagnosis of parotid gland tumors
Fig 6. Illustration of strain ratio (SR) measurement: a) Warthin tumor, b) malignant lesion
and 2.75 for malignant tumors, with a statistically signifi-
cant p value. For the SR cutoff value of 2.1, the Se was
83.3%, Sp was 97%, positive predictive value- PPV was
83.3%, negative predictive value - NPV was 97%, and
the diagnostic accuracy was 94% (fig 6) [11].
Klintworth et al (2012) showed that parotid tumors
may have certain elastographic patterns: garland sign (re-
ticular distribution of stiff tissue in the whole tumor) for
malignant tumors, dense core sign (very stiff central zone
with softer neighboring tissue) - characteristic of PMAs
and half-half sign (a stiff area located in the superficial
half of the lesion while the deeper part has a softer as-
pect) – for WT [40].
In a recent study, Cantisani et al showed that benign
tumors had a mean elasticity contrast index (ECI) of 2.31
and malignant tumors had a mean value of 5.55. The ECI
value >3.5 was the cutoff point for the differentiation of
malignancies (90.5% diagnostic accuracy). The authors
concluded that USE with ECI index improves US Se in
discriminating parotid tumors [20].
Virtual touch imaging quantification (VTIQ) rep-
resents a two-dimensional new form of shear wave im-
aging which allows combining quantitative and relative
stiffness imaging in a single screen. The investigation
also allows the user to draw a two-dimensional region of
interest (ROI). VTIQ includes several share wave modes
such as: velocity, quality, travel time, and displacement.
VTIQ in combination with routine US techniques pro-
vides additional data useful in distinguishing benign
from malignant parotid tumors [41,42].
Zengel et al (2017) using VTIQ showed that malig-
nancies had much lower quality and the share wave ve-
locity was higher than 6.8 m/s in comparison with be-
nign lesions [41]. Using the same investigation, Liu et a.
(2018) obtained higher values of share wave velocity for
malignant tumors compared to benign ones; the cutoff
value was 2,445 m/s with a Sp of 80% and a Se of 91.8%
[31]. Altinbas et al (2017) looking for the diagnostic
value of SE in parotid tumors differentiation presented
the following results: mean elasticity score for benign
tumors was 2.75±0.95, compared with malignant tumors
score of 3.44±0.85; (p<0.034) [43]. Heřman et al (2017)
showed in their study that the coefficient of stiffness
variability (CSV) which represents the ratio between the
maximum and minimum stiffness values recorded with
SWE is the strongest predictor of malignancy and allows
parotid tumor differentiation between benignity and ma-
lignancy (p <0.0001) [27].
Regarding the modern US techniques, the use of elas-
tography in the differential diagnosis of parotid tumors
is still controversial due to the overlap between the pa-
rameters. An important overlap between the values of
elasticity of PMA and malignant tumors was reported in
the literature [10,38]. Moreover, adenolymphomas show
a great variety of histological structure, which is reflect-
ed in the wide range of SWE elasticity modulus of these
tumors reported by Huang et al (2018) [44]. However,
most authors accept that, in general, malignant tumors
are stiffer than benign lesions, and PMAs are stiffer than
WTs, but this is not a strong enough criterion to allow the
differentiation of parotid tumors [10,11,26,28]. Zhang
et al (2019) in a meta-analysis presented a Sp of 67%
and a Se of 64% recorded by USE for the differentiation
of parotid tumors [34]. Many authors have investigated
methods to improve USE diagnostic accuracy in differ-
entiating parotid tumors. Thus, Mansour et al. (2019)
[29] succeeded in differentiating PMA from WT using
conventional US together with SWE. Cantisani et al
(2017) [20] using a cutoff value >3.5 for the elasticity
contrast index obtained significant results in differentiat-
ing malignant parotid tumors from benign lesions, with
an accuracy of 90.5%.
Herman et al (2017) obtained encouraging results
for differentiating parotid tumors with the help of a co-
efficient of stiffness variability (CSV) [27]. Cortcu et al
(2018) managed to differentiate between malignancy and
benignity using USE based on strain ratio cutoff value of
2.1 [11]. Encouraging results in the evaluation of parotid
tumors with the help of USE have been obtained also by
other authors [11,31,39], which shows that this investiga-

tion has a potential in differentiating parotid tumors that
must be investigated by further studies. USE, especially
SE is an operator-dependent investigation, which may
explain the large variations between the results of the
published studies.
Contrast-enhanced ultrasound
CEUS allows a quantitative analysis of the microvas-
cular perfusion of tumoral tissue, representing a reproduc-
ible measurement of perfusion kinetics after intravenous
administration of the contrast agent. It is a non-invasive
investigation with all the advantages of conventional US.
The contrast agent has a good safety profile, with a low
risk of side effects, and can be administered even to pa-
tients with impaired renal function due to its elimination
through the respiratory tract (sulphur hexafluoride gas).
The standard CEUS protocol involves injecting 4.8 ml of
contrast agent into a peripheral vein followed by dynam-
ic monitoring of the contrast agent passage through the
ROI for minimum 120 seconds. The conventional gray-
scale US images are displayed on the screen simultane-
ously with CEUS images. Quantitative parameters such
as time-intensity curves (TIC) and enhancement patterns,
but also a number of semi-quantitative parameters such
as area under the curve (AUC), time to peak (TTP) and
mean transit time (MTT) can be evaluated using CEUS
investigation [8,10,25].
Wei et al (2013) assessed the morphologic and dis-
tribution features of micro-vascularity of the tumors and
classified CEUS imaging into three types: type 1 - dif-
fuse homogeneous enhancement; type 2 - heterogeneous
enhancement with 3 subdivisions: a - more than 50%
enhanced areas with well-defined margin, b - dotted or
splattered vessels or less than 50% enhanced areas with
well-defined margin and c - heterogeneous enhanced, le-
sions with ill-defined margin; and type 3 - no enhance-
ment and iso-enhancement with 2 subtypes: a - no en-
hancement in lesions and b - iso-enhancement between
lesions and surrounding tissues. The authors concluded
that types 1 and 3 are representative for benign tumors and
type 2, especially subtype c can suggest malignancy [45].
Many studies published in the literature have shown
improved results with the using of CEUS for the differen-
tial diagnosis of parotid tumors (fig 7) [5,16,25,32].
Data from the literature show that the use of multi-
modal US to differentiate parotid tumors increased the
Se of malignant tumors detection from 77% to 91% but
decreased the Sp from 98% to 81% compared to conven-
tional US [32].
Steinhart et al (2003) showed that using US contrast
agents in combination with Doppler US, PMA recorded
Med Ultrason 2021; 23(2): 194-202 199
a significantly stronger perfusion enhancement with in-
creased Doppler signal area. The authors concluded that
the Doppler signal area can show different courses de-
pending on tumor histology [46].
Fischer et a. (2010) were successful in differentiating
PAM from WT based on the TTP parameter (26.8±11.1
versus 22.6±5.1 sec, p<0.05). Also, following the AUC
analysis, the PMA registered much lower values com-
pared to the WT [16].
According to Klotz et al (2013) malignant tu-
mors register prolonged MTT (17.94±1.62 s versus
14.86±0.65 s; p<0.05) and AUC values (584.9±143 ver-
sus 400.62±53.85; p<0.05) compared to benign lesions.
Regarding the benign tumors, WT showed higher AUC
values than PMA (515.4± 1.26 versus 285.82±36.44;
p<0.05) [25].
Klotz et al (2014) using a new quantification software
(Vuebox) for CEUS, were able to differentiate malignant
tumors from benign ones (p<0.05) as well as WT from
PMA (p<0.05) based on the following parameters: AUC
- 528.6±183.3 vs. 174.4±52.9 – malign vs. benign and
302.8±36.2 – WT vs 65.6±13.8 – PMA, peak enhance-
ment - 155.3±69.5 vs. 38.2±9.0 – malign vs. benign and
68.8±11.4 – WT vs 12.2±1.8 – PMA, wash-in-rate and
wash-in perfusion index. On the other hand, they did not
find significant differences for mean transit time (MTT)
and rise time (RT) [47].
Küstermeyer et al (2016) using an innovative method
of analysis the perfusion characteristics in 8 ROIs based
on CEUS, obtained a significant difference between be-
nign parotid tumors compared to malignant ones [48]. On
the other hand, Badea et al (2013) could not find signifi-
cant differences between the circulatory bed of benign
and malignant tumors using CEUS investigation [12].
The CEUS parameters - TIC, TTP, MTT and AUC
were successful in the differential diagnosis of parotid
tumors [5,16,25]. Klotz et al (2013) showed in their study
that MTT and AUC showed significantly higher values
for malignant tumors compared to benign ones. The
circulation of the contrast agent through abnormal and
disordered vessel architecture of cancer lesions may ex-
plain these increased values; on the other hand, it seems
that the reduced perfusion of benign tumors may occur
due to the regular architecture of the vascular systems of
these tumors [25]. Klotz et al (2014) [47] with the help
of a new software and the parameters such as: AUC, PE,
WiR, WiPI manage to differentiate malignant parotid tu-
mors from benign ones with CEUS (fig 8).
The preoperative differential diagnosis of parotid tu-
mors based on US investigations is challenging for clini-
cians due to the great histological variety of parotid tu-
mors. According to the latest World Health Organization
200 Sebastian Stoia et al Ultrasonography techniques in the preoperative diagnosis of parotid gland tumors
classification from 2017 [1], 20 malignant and 11 benign
tumor histology are described, to which another 9 histo-
logical tumor types are added. The histological and US
characteristics that overlap between many tumor entities,
as well as the lack of clear guidelines for the differential
Fig 7. Illustration of CEUS pattern in a histological proven
Warthin tumor: peripheral inhomogeneous enhancement in the
arterial phase (a) and intralesional vascular septa and wash-out
in the late phase (b); c) the enhancement curve of the lesion is
suggestive for benign lesion.
Fig 8. CEUS appearance in a case with malignant parotid le-
sion: rapid tumoral uptake of the contrast during the arterial
phase (a), followed by a rapid wash-out at the end of the arterial
phase (b). The enhancement curve of the lesion is suggestive
for malignancy, with similar pattern for the surrounding meta-
static adenopathy (c).
diagnosis of parotid tumors contribute to the controver-
sies of this topic. In cases of inconclusive results, addi-
tional investigations and/or a second imaging opinion
could help establish the correct diagnosis.
The use of MPUS provides promising results for pa-
rotid tumors differential diagnosis. Data from the litera-
ture shows high Se of up to 91% and Sp by combining
USE, CEUS and conventional US in differentiating pa-
rotid tumors; however, US still cannot replace the MRI
scan [32,49]. Further studies are necessary in order to
define the diagnostic value of the MPUS and to draw
clear indications of the method in the current diagnosis
of parotid tumors.
Based on the literature review [4-5,7-10,13,16-
20,25,28,30-32,45], a guide for the US differential di-
agnosis of parotid tumors is proposed by our group and
presented in fig 9.
Fig 9. US guide for the differential diagnosis of parotid gland
tumors

Limitations
This literature review is not without limitations. Many
of the revised studies are retrospective, performed on
small groups of patients. In these studies, there is no unity
in the experience of radiologists or pathologists who have
interpreted the investigation reports. The statistical meth-
ods used differ from one study to another. There are large
variations between the US equipment and protocols. In
some studies, there was an unequal distribution between
the evaluated malignant and benign parotid tumors.
Conclusions
The most important statistically sustained conclusion
of this review is that if the US of a parotid tumor located in
the superficial lobe suggests benignity and the FNAB report
confirms the diagnosis and succeeds in specifying the tumor
histological type, no further investigations are necessary.
The preoperative differential diagnosis of parotid tu-
mors is challenging and plays a crucial role in the choice
of therapeutic strategy. Establishing a correct preoperative
diagnosis with the US techniques requires a systematic and
careful approach with the correlation of the clinical exam
and the patient history. The experience of the US operator
and the knowledge of the US limits are of great importance.
Ultrasonography adds important information to the
clinical exam related to the tumor location and exten-
sion, has an important role in the differential diagnosis
between benign and malignant lesions and guides the
clinician for the need of further investigations and the
therapy strategy.
The value of USE alone is not proven in the differen-
tial diagnosis of parotid tumors. However, the USE pro-
vides additional data on the elasticity of parotid tumors.
Adding CEUS to conventional US increases the differ-
ential diagnosis accuracy of parotid tumors. MPUS al-
lows the differential diagnosis of parotid tumors, but ad-
ditional studies are required to validate the results. When
the diagnosis of parotid tumors is inconclusive with the
MPUS, MRI becomes mandatory.
Conflict of interest: none
References
1. Thielker J, Grosheva M, Ihrler S, Wittig A, Guntinas-Lichi-
us O. Contemporary Management of Benign and Malignant
Parotid Tumors. Front Surg 2018;5:39.
2. Tartaglione T, Botto A, Sciandra M, et al. Diagnosi differ-
enziale dei tumori parotidei: Quali caratteristiche di riso-
nanza magnetica considerare? Acta Otorhinolaryngol Ital
2015;35:314–320.
3. Zhan KY, Khaja SF, Flack AB, Day TA. Benign Parotid Tu-
mors. Otolaryngol Clin North Am 2016;49:327–342.
Med Ultrason 2021; 23(2): 194-202 201
4. Howlett DC, Kesse KW, Hughes DV, Sallomi DF. The role
of imaging in the evaluation of parotid disease. Clin Radiol
2002;57:692–701.
5. Cantisani V, David E, Sidhu PS, et al. Parotid Gland Le-
sions: Multiparametric Ultrasound and MRI Features. Ul-
traschall Med 2016;37:454–471.
6. Kato H, Kanematsu M, Mizuta K, Aoki M. Imaging find-
ings of parapharyngeal space pleomorphic adenoma in
comparison with parotid gland pleomorphic adenoma. Jpn
J Radiol 2013;31:724–730.
7. Rzepakowska A, Osuch-Wójcikiewicz E, Sobol M, Cruz R,
Sielska-Badurek E, Niemczyk K. The differential diagno-
sis of parotid gland tumors with high-resolution ultrasound
in otolaryngological practice. Eur Arch Otorhinolaryngol
2017;274:3231–3240.
8. David E, Cantisani V, De Vincentiis M, et al. Contrast-en-
hanced ultrasound in the evaluation of parotid gland lesions:
An update of the literature. Ultrasound 2016;24:104–110.
9. Afzelius P, Nielsen MY, Ewertsen C, Bloch KP. Imaging
of the major salivary glands. Clin Physiol Funct Imaging
2016;36:1–10.
10. Bhatia KSS, Dai YL. Routine and Advanced Ultrasound
of Major Salivary Glands. Neuroimaging Clin N Am
2018;28:273–293.
11. Cortcu S, Elmali M, Tanrivermis Sayit A, Terzi Y. The Role
of Real-Time Sonoelastography in the Differentiation of
Benign from Malignant Parotid Gland Tumors. Ultrasound
Q 2018;34:52–57.
12. Badea AF, Bran S, Tamas-Szora A, Floareş A, Badea R,
Baciut G. Solid parotid tumors: An individual and in-
tegrative analysis of various ultrasonographic criteria.
A prospective and observational study. Med Ultrason
2013;15:289–298.
13. Rong X, Zhu Q, Ji H, Li J, Huang H. Differentiation of
pleomorphic adenoma and Warthin’s tumor of the pa-
rotid gland: Ultrasonographic features. Acta Radiol
2014;55:1203–1209.
14. Brennan PA, Herd MK, Howlett DC, Gibson D, Oeppen
RS. Is ultrasound alone sufficient for imaging superficial
lobe benign parotid tumours before surgery? Br J Oral
Maxillofac Surg 2012;50:333–337.
15. Liu Y, Li J, Tan YR, Xiong P, Zhong LP. Accuracy of diag-
nosis of salivary gland tumors with the use of ultrasonogra-
phy, computed tomography, and magnetic resonance imag-
ing: A meta-analysis. Oral Surg Oral Med Oral Pathol Oral
Radiol 2015;119:238-245.e2.
16. Fischer T, Paschen CF, Slowinski T, et al. Differentiation
of parotid gland tumors with contrast-enhanced ultrasound.
Rofo 2010;182:155–162.
17. Lingam RK, Daghir AA, Nigar E, Abbas SAB, Kumar
M. Pleomorphic adenoma (benign mixed tumour) of the
salivary glands: Its diverse clinical, radiological, and his-
topathological presentation. Br J Oral Maxillofac Surg
2011;49:14–20.
18. Gerwel A, Kosik K, Jurkiewicz D. US in preoperative
evaluation of parotid gland neoplasms. Otolaryngol Pol
2015;69:27-33.
202 Sebastian Stoia et al Ultrasonography techniques in the preoperative diagnosis of parotid gland tumors
19. Kovačević DO, Fabijanić I. Sonographic diagnosis of pa-
rotid gland lesions: Correlation with the results of sono-
graphically guided fine-needle aspiration biopsy. J Clin
Ultrasound 2010;38:294–298.
20. Cantisani V, David E, De Virgilio A, et al. Prospective
evaluation of Quasistatic Ultrasound Elastography (USE)
compared with Baseline US for parotid gland lesions: Pre-
liminary results of elasticity contrast index (ECI) evalua-
tion. Med Ultrason 2017;19:32–38.
21. Wu S, Liu G, Chen R, Guan Y. Role of ultrasound in the
assessment of benignity and malignancy of parotid masses.
Dentomaxillofacial Radiol 2012;41:131–135.
22. Cheng PC, Chang CM, Huang CC, et al. The diagnostic
performance of ultrasonography and computerized tomog-
raphy in differentiating superficial from deep lobe parotid
tumours. Clin Otolaryngol 2019;44:286–292.
23. Luczewski L, Golusinski P, Pazdrowski J, et al. The ultra-
sound examination in assessment of parotid gland tumours:
The novel graphic diagram. Eur Arch Otorhinolaryngol
2013;270:2129–2133.
24. Zhang W, Zuo Z, Luo N, et al. Non-enhanced MRI in
combination with color Doppler flow imaging for improv-
ing diagnostic accuracy of parotid gland lesions. Eur Arch
Otorhinolaryngol 2018;275:987–995.
25. Klotz LV, Gürkov R, Eichhorn ME, et al. Perfusion char-
acteristics of parotid gland tumors evaluated bycontrast-
enhanced ultrasound. Eur J Radiol 2013;82:2227–2232.
26. Westerland O, Howlett D. Sonoelastography techniques in
the evaluation and diagnosis of parotid neoplasms. Eur Ra-
diol 2012;22:966–969.
27. Heřman J, Sedláčková Z, Vachutka J, et al. Differential
Diagnosis of Parotid Gland Tumors: Role of Shear Wave
Elastography. Biomed Res Int 2017;2017:9234672.
28. Bhatia KSS, Cho CCM, Tong CSL, Lee YYP, Yuen EHY,
Ahuja AT. Shear wave elastography of focal salivary gland
lesions: Preliminary experience in a routine head and neck
US clinic. Eur Radiol 2012;22:957–965.
29. Mansour N, Stock KF, Chaker A, Bas M, Knopf A. Evalua-
tion of parotid gland lesions with standard ultrasound, color
duplex sonography, sonoelastography, and acoustic radia-
tion force impulse imaging - A pilot study. Ultraschall Med
2012;33:283–288.
30. Mansour N, Hofauer B, Knopf A. Ultrasound Elastog-
raphy in Diffuse and Focal Parotid Gland Lesions. ORL
2017;79:54–64.
31. Liu G, Wu S, Liang X, Cui X, Zuo D. Shear Wave Elastog-
raphy Improves Specificity of Ultrasound for Parotid Nod-
ules. Ultrasound Q 2018;34:62–66.
32. Mansour N, Bas M, Stock KF, Strassen U, Hofauer B,
Knopf A. Multimodal Ultrasonographic Pathway of Parotid
Gland Lesions. Ultraschall Med 2017;38:166-173.
33. Yerli H, Eski E, Korucuk E, Kaskati T, Agildere AM. So-
noelastographic qualitative analysis for management of sal-
ivary gland masses. J Ultrasound Med 2012;31:1083–1089.
34. Zhang YF, Li H, Wang XM, Cai YF. Sonoelastography for
differential diagnosis between malignant and benign parot-
id lesions: a meta-analysis. Eur Radiol 2019;29:725–735.
35. Celebi I, Mahmutoglu AS. Early results of real-time quali-
tative sonoelastography in the evaluation of parotid gland
masses: A study with histopathological correlation. Acta
Radiol 2013;54:35–41.
36. Karaman CZ, Başak S, Polat YD, et al. The role of real-time
elastography in the differential diagnosis of salivary gland
tumors. J Ultrasound Med 2019;38:1677–1683.
37. Dumitriu D, Dudea SM, Botar-Jid C, Bǎciuţ G. Ultrasono-
graphic and sonoelastographic features of pleomorphic ad-
enomas of the salivary glands. Med Ultrason 2010;12:175–
183.
38. Dumitriu D, Dudea S, Botar-Jid C, Bǎciuţ M, Bǎciuț G. Re-
al-time sonoelastography of major salivary gland tumors.
AR Am J Roentgenol 2011;197:924–930.
39. Wierzbicka M, Kałużny J, Szczepanek-Parulska E, et al.
Is sonoelastography a helpful method for evaluation of pa-
rotid tumors? Eur Arch Otorhinolaryngol 2013;270:2101–
2107.
40. Klintworth N, Mantsopoulos K, Zenk J, Psychogios G, Iro
H, Bozzato A. Sonoelastography of parotid gland tumours:
Initial experience and identification of characteristic pat-
terns. Eur Radiol 2012;22:947–956.
41. Zengel P, Notter F, Reichel CA, Clevert DA. Does Virtual
Touch IQ elastography help to improve the preoperative di-
agnosis of parotid tumors: A prospective trial. Clin Hemor-
heol Microcirc 2017;67:425–434.
42. Zengel P, Notter F, Clevert DA. Does acoustic radia-
tion force elastography improve the diagnostic capabil-
ity of ultrasound in the preoperative characterization of
masses of the parotid gland? Dentomaxillofac Radiol
2018;47:20180068.
43. Altinbas NK, Gundogdu Anamurluoglu E, Oz II, et al. Re-
al-time sonoelastography of parotid gland tumors. J Ultra-
sound Med 2017;36:77–87.
44. Huang K, Gao N, Wang X, Bian D, Liu Y, Zhai Q. Clini-
cal Application of Real-Time Shear Wave Elastography in
Identifying the Histological Components of Parotid Adeno-
lymphoma. Ultrasound Q 2018;34:47–51.
45. Wei X, Li Y, Zhang S, et al. Evaluation of microvasculari-
zation in focal salivary gland lesions by contrast-enhanced
ultrasonography (CEUS) and Color Doppler sonography.
Clin Hemorheol Microcirc 2013;54:259–271.
46. Steinhart H, Zenk J, Sprang K, Bozzato A, Iro H. Con-
trast-enhanced color Doppler sonography of parotid
gland tumors. Eur Arch Otorhinolaryngol 2003;260:344–
348.
47. Klotz LV, Ingrisch M, Eichhorn ME, et al. Monitoring
parotid gland tumors with a new perfusion software for
contrast-enhanced ultrasound. Clin Hemorheol Microcirc
2014;58:261–269.
48. Küstermeyer J, Klingelhöfer G, Welkoborsky HJ. Analysis
of Perfusion Parameters within Salivary Gland Tumors us-
ing Contrast Enhanced Ultrasound. Laryngorhinootologie
2016;95:688-693.
49. Knopf A, Mansour N, Chaker A, Bas M, Stock K. Mul-
timodal ultrasonographic characterisation of parotid gland
lesions - A pilot study. Eur J Radiol 2012;81:3300–3305.
Review Med Ultrason 2021, Vol. 23, no. 2, 203-212
DOI: 10.11152/mu-2659

The diagnostic accuracy of ultrasound in the detection of foot and

ankle fractures: a systematic review and meta-analysis
Jiangfeng Wu*, Yunlai Wang*, Zhengping Wang
* the authors shared the first authorship

Department of Ultrasound, Affiliated Dongyang Hospital of Wenzhou Medical University, Dongyang, Zhejiang,
China

Abstract
Aims: Foot and ankle injuries are a common presenting complaint in the emergency department. The diagnosis of foot
and ankle fractures is conventionally accomplished through X-rays. Whether ultrasound (US) can be considered as a primary
scanning modality is still a controversial issue; therefore, we did a meta-analysis to synthesize the diagnostic performance of
ultrasound for foot and ankle fractures. Material and methods: A comprehensive search was carried out to identify studies
in which patients with clinically suspected foot and ankle fractures were assessed by US. Two investigators independently
screened the literature and extracted the data. Any discrepancies were resolved via discussion. Study quality was assessed by
the Quality Assessment of Diagnostic Accuracy Studies 2 tool, and pooled sensitivity and specificity of various US findings
were determined. Results: Ten studies with a total of 1065 patients were included. There was significant heterogeneity across
the included studies. The pooled sensitivity, specificity, positive likelihood ratio, negative likelihood ratio, and diagnostic odds
ratio for the diagnosis of foot and ankle fractures by US were 0.96 (95% confidence interval [CI], 0.90-0.99), 0.94 (95% CI,
0.88-0.97), 15.0 (95% CI, 7.9-28.6), 0.04 (95% CI, 0.02-0.11), and 367 (95% CI, 101-1338), respectively. Furthermore, the
summary receiver operating characteristic area under the curve was calculated to be 0.99. Conclusions: Ultrasound has an
excellent diagnostic performance for foot and ankle fractures and should be considered as a primary and radiation-free scan-
ning modality in the diagnosis of foot and ankle fractures.
Keywords: ultrasound; foot and ankle fractures; meta-analysis; systematic review; diagnostic accuracy

Introduction tients with foot and ankle fractures to prevent long-term

Foot and ankle injuries are a common presenting
complaint to the emergency department (ED), which oc-
cur as buckling or blunt trauma and generally lead to a
strain, sprain, or, more rarely, fracture [1,2]. Although
such injuries are not usually fatal, the functions of the
affected limb can be imperilled, and therefore, early di-
agnosis and treatment are exceedingly significant for pa-
Received 06.06.2020  Accepted 25.09.2020
Med Ultrason
2021, Vol. 23, No 2, 203-212
Corresponding author: Jiangfeng Wu
Department of Ultrasound, Affiliated Dongyang
Hospital of Wenzhou Medical University
60 Wuning West Road, Dongyang (322100),
Zhejiang, China
E-mail: wjfhospital@163.com
Phone: 18257937213
complications [3].
The diagnosis of foot and ankle fractures is conven-
tionally accomplished through X-ray, which is usually
considered as the standard reference [4]. Whereas, the
X-ray entails exposure to ionizing radiation with its at-
tendant possible carcinogenic and teratogenic effects
[5,6]. Alternative imaging modalities, such as computed
tomography (CT), magnetic resonance imaging (MRI)
and ultrasound (US), have been considered to improve
the diagnostic accuracy of foot and ankle fractures [7,8].
However, selection of CT examination as the standard
reference will bring overmuch and sometimes unneces-
sary ionizing radiation to patients and MRI is related to
increased and unnecessary time and expense [5,7].
In the recent years, ultrasound has emerged as a pos-
sible alternative for the diagnosis of bone fractures in the
ED, which has been reported to have a high accuracy in
204 Jiangfeng Wu, Yunlai Wang et al The diagnostic accuracy of ultrasound in the detection of foot and ankle fractures
pediatric elbow, distal radius, metacarpal, phalanx, ankle
and metatarsal fractures [9–12]. One of the main advan-
tages of ultrasound is the lack of ionizing radiation. Fur-
thermore, ultrasound can be immediately performed and
is easily accessible in the ED, reducing diagnostic delays
and the time to the initiation of management. Additional
advantages of ultrasound include the relative easiness to
teach [13], reduced pain experience [14], repeatability,
portability, and the provision of additional information
regarding the musculoskeletal system [15–17].
Whether ultrasound can be considered as a primary
scanning modality is still a controversial issue, as the
diagnostic accuracy of ultrasound for detecting foot and
ankle fractures is variable across different studies. Prior
studies assessing the accuracy of US in the diagnosis of
foot and ankle fractures have been published, with the
sensitivity ranging from 83% to 100% and the specificity
ranging from 76% to 100% [18–20]. To our knowledge,
no studies have comprehensively evaluated the literature
on foot and ankle fractures diagnosis using ultrasound.
Hence, we did a meta-analysis to synthesize the diagnos-
tic performance of ultrasound for foot and ankle fractures.
Material and methods
Meta-analysis principles
This meta-analysis was conducted according to the
Preferred Reporting Items for Systematic Reviews and
Meta-Analyses (PRISMA) guidelines, which include 27
items and provide specific guidance for the reporting of
systematic reviews [21].
Search strategy
Pubmed, EMBASE and Cochrane Library were sys-
tematically searched to identify potentially eligible stud-
ies from inception to March 2020. Computer searches
were carried out using the Medical Subject Heading and
keywords. Detailed search terms are provided in supple-
mentary file 1. The bibliographies of identified studies
and review articles were manually screened to expand
the number of eligible studies. Only studies in English,
which satisfied the inclusion criteria, were included.
Inclusion and exclusion criteria
Two researchers (JW and YW) independently
screened the titles and the abstracts of the potentially eli-
gible studies. Before identifying the literature, inclusion
and exclusion criteria were defined to increase validity
and reproducibility. Any disagreements between the two
researchers were resolved via discussion with the senior
author (ZW).
The inclusion criteria were as follows: (1) randomised
control trials and prospective studies were included;
(2) studies involving patients with clinically suspected
foot and ankle fractures; (3) the accuracy of ultrasonog-
raphy in the diagnosis of foot and ankle fractures was
evaluated and (4) a reference standard was adopted to
confirm foot and ankle fractures, including X-ray, com-
puted tomography, and/or magnetic resonance imaging.
The exclusion criteria were as follows: (1) case re-
ports, letters, guidelines, consensus statements, and un-
published articles; (2) studies that contained an over-
lapped population and (3) studies without sufficient data
to construct diagnostic 2x2 tables.
Data extraction
Two researchers (JW and YW) independently ex-
tracted the relevant data from the included studies us-
ing a pre-designed data collection form. Any discrepan-
cies were resolved via discussion with the senior author
(ZW). For eligible studies, the following items were ex-
tracted: last name of the first author, year of publication,
country, study type, study setting, blinding method, US
equipment, probe frequency, sample size, number with
fractures, fracture prevalence, fracture site, mean age,
gender, US operator specialty, examiner training, US
diagnostic criteria, standard reference, time between ul-
trasonography and the standard reference, true positives,
true negatives, as well as false positives and false nega-
tives of US in the diagnosis of foot and ankle fractures.
Study quality assessment
The Quality Assessment of Diagnostic Accuracy
Studies-2 (QUADAS-2) tool [22] was utilized to evaluate
the risk of bias and methodological quality. The quality
of each included study was evaluated by an appraisal of
the risk of bias of four domains and clinical applicability
of three domains of the study characteristics. Four do-
mains consisted of patient selection, index test, reference
standard and flow and timing. Each domain was evalu-
ated for risk of bias, and the first three domains were
evaluated for applicability. The processing of the quality
assessment was performed utilizing RevMan 5.3 soft-
ware (Nordic Cochrane Centre, Copenhagen, Denmark).
Statistical analysis
The present meta-analysis was conducted by Stata
12.0 (Stata Corporation, College Station, Texas). All sta-
tistical analyses were performed by one investigator, who
has experience in performing meta-analysis. The sum-
mary estimates of sensitivity, specificity, positive likeli-
hood ratio (PLR), negative likelihood ratio (NLR) and
diagnostic odds ratio (DOR) with corresponding 95%
confidence intervals (CIs) were calculated using a bivari-
ate random effect model in the present analysis, which
indicated the accuracy of US in the diagnosis of foot and
ankle fractures. Meanwhile, the summary receiver op-
erator curve (SROC) was constructed and the area under
the curve (AUC) was calculated. An AUC close to 0.5

shows a poor test, while an AUC of 1.0 demonstrates an
excellent diagnostic test [23]. We applied the spearman
correlation analysis to determine whether a threshold ef-
fect is present, with p < 0.05 representing a threshold ef-
fect. The Cochrane Q test and the inconsistency index
(I2) were used to assess the heterogeneity among differ-
ent studies with a p-value < 0.1 or I2 > 50% considered
significant for heterogeneity [24].
Meta-regression analyses utilizing several covari-
ates were carried out to investigate the potential causes
of heterogeneity: country (Turkey versus countries other
than Turkey), sample size (>100 versus ≤100), fracture
prevalence (>30% versus ≤30%), year published (2009-
2013 versus 2014-2019), reference standard (only X-ray
versus including CT or MRI), ultrasonographic operator
(emergency physician versus others), ultrasonographic
training (yes versus others), blinding method (double
blinding versus others), and bedside ultrasound (yes ver-
sus others). The Deeks’ funnel plot asymmetry test was
applied to assess publication bias [25], through a p value
>0.05 denoting no significant publication bias.
Results
Study selection
The initial search for studies which assessed the di-
agnostic performance of US for foot and ankle fractures
provided 1818 studies, of which 1217 relevant studies
remained after removing 601 duplicate studies. We re-
Fig 1. Preferred Reporting Items for Systematic Reviews and
Meta-Analyses (PRISMA) flowchart of the search process
Med Ultrason 2021; 23(2): 203-212 205
viewed 1217 titles and abstracts and then excluded 1176
studies because it was obvious from the title or abstract
that they were not relevant to this meta-analysis. Full text
of the remaining studies was reviewed, and another 41
studies were excluded. Finally, 10 original research stud-
ies were included in the present meta-analysis [18-20,26-
32]. Manual searching of the reference cited in these 10
studies did not yield any additional relevant studies. Fig-
ure 1 shows a flow diagram summarizing the literature.
Characteristics of the included studies
The 10 included studies with a total of 1065 patients
were published between 2009 and 2019 and written in
English. Five studies were conducted in Turkey [20,28-
31], 1 was performed in France [18], 1 in Israel [19], 1
in England [26], 1 in Sweden [27], and 1 in Iran [32]. All
were prospective observational studies. The number of
patients in the study ranged from 37 to 246; 48.2% of pa-
tients were male and the mean age ranged from 8.1 to 52.7
years. The prevalence of foot and ankle fractures ranged
from 10%-40.4%. Nine studies [19,20,26-32] were con-
ducted in ED and 1 was performed in the rheumatology
department [18]. Emergency physicians performed the US
examination in 5 studies [20,28,30-32], an experienced
rheumatologist in 1 study [18], a pediatric radiologist in
one study [19], an orthopedic surgeon in one study [27], an
ED member in 1 study [26] and a sonographer in 1 study
[29]. Double blinding between the standard reference and
index tests was found in 7 studies [18,20,26,27,29,31,32],
single blinding of the standard reference to US results in
2 studies [19,28] and 1 study [30] did not report a blind-
ing method. In all studies the time interval between US
and the standard reference was not given except for the
study by Banal et al who declared US and the standard
reference were performed on the same day [18]. Three
studies included ankle fractures [19,27,31], 3 included
foot or ankle fractures [26,28,29] and 4 only included
metatarsal fractures [18,20,30,32]. Seven studies used
the X-ray as the reference standard [19,20,26-28,30,32],
1 used MRI [18] and 2 used X-ray or CT [29,31]. The
probe frequency ranged from 5 to 15 MHz. Table I and
II epitomizes the data extracted from the included stud-
ies. More details are showed in supplementary file 2.
Quality assessment
The quality assessment results of the risk of bias and
applicability concerns of the selected studies were pre-
sented graphically in figure 2.
With respect to the patient selection domain, 1 study
was considered as having high bias because the sam-
ple was selected by the nonrandom purposive sampling
method [32].
206 Jiangfeng Wu, Yunlai Wang et al The diagnostic accuracy of ultrasound in the detection of foot and ankle fractures
Table I. Primary data extracted from included studies for meta-analysis
Author Year Study Country Sample Number with Fracture Male/ Age Training
type size (n) fractures (n) (%) female (mean±SD)
Banal [18] 2009 Pro France 37 12 32.4 9/28 52.7±14.1 No
Simanovsky [19] 2009 Pro Israel 41 11 26.8 14/27 8.1 No
Canagasabey [26] 2011 Pro England 110 11 10.0 65/45 31.8±12.7 Yes
Hedelin [27] 2013 Pro Sweden 122 23 18.8 58/64 42 (18-92) Yes
Ekinci [28] 2013 Pro Turkey 131 20 15.2 64/67 37.2±15.44 Yes
Atilla [29] 2014 Pro Turkey 246 76 30.9 105/141 37 (median age) Yes
Yesilaras [20] 2014 Pro Turkey 84 34 40.4 36/48 36±15 No
Kozaci [30] 2017 Pro Turkey 72 28 38.8 48/24 33±18 Yes
Ozturk [31] 2018 Pro Turkey 120 42 35.0 57/63 40.8±19.3 Yes
Ebrahimi [32] 2019 Pro Iran 102 31 30.3 58/44 35.14±14.32 NR
Pro, prospective; NR, not reported; SD, standard deviation

Concerning the index test domain, 1 study [30] was

considered as “unknown” because the blinded status was
not explicitly reported; 2 other studies [27,28] were also
labelled as “unknown” because they did not explicitly
report the diagnostic threshold. Regarding the reference
standard domain, 3 studies [19,28,30] were considered as
“unknown” because the blinded status was not explicitly
reported. With regard to the flow and timing domain, 9
studies were considered as “unknown” because they did
not definitely report the time interval between US and the
reference standard [19,20,26-32].
Regarding applicability, for patient selection, index
test, and reference standard domains, all studies were
considered to have low concerns.
Data synthesis
Overall, ultrasound had a 0.96 (95% CI, 0.90-0.99)
sensitivity and 0.94 (95% CI, 0.88-0.97) specificity for
the diagnosis of foot and ankle fractures (fig 3). The
pooled PLR, NLR, and DOR of US were 15.0 (95% CI,
7.9-28.6), 0.04 (95% CI, 0.02-0.11), and 367 (95% CI, Fig 2. Quality assessment of the included studies using QUA-
101-1338), respectively and the post-test probability was DAS-2 tool
79% and 1%, respectively (fig 4). Significant heterogene-
ity was found for the sensitivity (I2 = 51.43%, p=0.03) other potential sources of heterogeneity. The covariables
and specificity (I2 = 81.45%, p=0.00). The AUC under the included the locale (Turkey versus countries other than
SROC curve for the value of US in the diagnosis of foot Turkey), number of patients (>100 versus ≤100), fracture
and ankle fractures was 0.99 (fig 5). The Spearman corre- prevalence (>30% versus ≤30%), year published (2009-
lation coefficient was determined to be -0.127 (p=0.726), 2013 versus 2014-2019), reference standard (only radio-
which indicated no significant threshold effect among the graph versus including CT or MRI), ultrasonographic
individual studies. operator (emergency physician versus others), ultrasono-
The Deeks’ funnel plot asymmetry test demonstrated graphic training (yes versus others), blinding method
that the studies were distributed symmetrically with a p (double blinding versus others), and bedside ultrasound
value of 0.8 (fig 6), which indicated that there was no (yes versus others).
significant publication bias in the present meta-analysis. Among the various potential covariates, the blinding
Meta regression and subgroup analyses method (double blinding versus others) was associated
Due to the significant heterogeneity among studies, with the heterogeneity of the sensitivity. With respect
meta-regression analysis was then conducted to explore to the covariate of the country, studies in Turkey had a
 Med Ultrason 2021; 23(2): 203-212 207
Table II. Characteristics of the included studies
Author Fre- Fracture Blind- Refer- US Study US US diagnostic Time
quency site ing ence operator setting equipment criteria between
(MHz) standard reference
and US
Banal 7.5–13 Metatar- Double MRI Experienced Rheumatol- An Esaote Hypoechoic The
[18] sal blinded rheumatolo- ogy depart- Technos periosteal elevation, same
gists ment MP system cortical disrup- day
tion, and increased
vascularity
Simanovsky 5–12 Ankle Single X-ray A pediatric Emergency HDI 5000 Discontinuity of the NR
[19] blinded radiologist department machine echogenic cortical
line, cortical depres-
sion, and periosteal
elevation
Canagasabey NR Foot or Double X-ray An emer- Emergency NR A significant frac- NR
[26] ankle blinded gency department ture was defined as
department having a breadth
member greater than 3 mm
Hedelin 15 Ankle Double X-ray Orthopedic Emergency An NR NR
[27] blinded surgeons department M-Turbo
Sonosite
system
Ekinci [28] 10 Foot or Single X-ray An mergen- Emergency A Logiq NR NR
ankle blinded cy physician department Book XP
device
Atilla [29] 10 Foot Double X-ray or A sonogra- Emergency Mindray Cortical disruption NR
and/or blinded CT pher department M5 or stepping or axial
ankle deviation on the
bone surface
Yesilaras 7.5-10 The fifth Double X-ray An emergen- Emergency Mindray Cortical disruption NR
[20] meta- blinded cy physician department M5
tarsal
Kozaci 7.5 Meta- NR X-ray Emergency Emergency Esaote Cortical disruption NR
[30] tarsal physicians department Firenze
Italy
Ozturk 10 Malleo- Double X-ray or Emergency Emergency Mindray A cortical irregular- NR
[31] lus blinded CT physicians department M7 ity on one or more
plane
Ebrahimi 10 Meta- Double X-ray Emergency Emergency NR Presence of cortical NR
[32] tarsal blinded medicine department disruption or
specialist stepping or axial
deviation of
the bone surface
NR, not reported; US, ultrasound; CT, computed tomography; MRI, magnetic resonance imaging

higher pooled specificity to countries other than Turkey Discussion

(specificity: 0.96 and 0.88). All other sensitivity and
specificity of subgroup analysis were similar.
The subgroup analysis of the eight studies [18,20,26-
29,31,32] only including patients older than 14 years,
demonstrated a pooled sensitivity of 0.96 (CI: 0.89–
0.99) and specificity 0.94 (CI: 0.87–0.97). Furthermore,
the subgroup analysis of the four studies [18,20,30,32]
with respect to metatarsal fractures evidenced a pooled
sensitivity of 0.94 (CI: 86–98) and specificity 0.90 (CI:
0.74–0.97). The results of the meta-regression are shown
in Table III.
So far, many studies have shown that ultrasound has a
well diagnostic value for fractures in different sites. Gor-
don et al showed that point-of-care ultrasound had rela-
tively high sensitivity of 0.91 and specificity of 0.96 for
diagnosis in detecting skull fractures after pediatric head
trauma [33]. Zhao et al demonstrated that ultrasound had
an excellent diagnostic value for hand fractures, with a
pooled sensitivity of 0.91 and specificity of 0.96. More-
over, ultrasound was recommended as a first-line and
radiation-free modality in detecting hand fractures [12].
208 Jiangfeng Wu, Yunlai Wang et al The diagnostic accuracy of ultrasound in the detection of foot and ankle fractures
Fig 3. Forest diagrams of US diagnosing foot and ankle frac-
tures, showing sensitivity and specificity
Fig 4. Fagan nomogram for detecting foot and ankle fractures
Lee et al reported that ultrasound sensitivity varied from
0.88 to 0.99 and specificity varied from 0.82 to 0.94 in el-
bow fracture in pediatric patients with trauma [9]. Charti-
er et al suggested that point-of-care ultrasound sensitiv-
ity varied from 0.65 to 1.00 and specificity varied from
0.79 to 1.00 in long bone fractures [34]. Douma-den et
al reported that ultrasound had a perfect accuracy for the
diagnosis of distal forearm fractures with a sensitivity of
0.97 and specificity of 0.95, particularly in children [35].
Fig 5. Summary receiver operating characteristics curve of US
for foot and ankle fractures
Fig 6. Funnel plot of US for foot and ankle fractures
So far, this is the first meta-analysis to evaluate the di-
agnostic performance of US for foot and ankle fractures.
This meta-analysis provided a pooled sensitivity of 0.96
(95% CI, 0.90-0.99), specificity of 0.94 (95% CI, 0.88-
0.97) and DOR of 367 (95% CI, 101-1338), respectively,
with an SROC AUC of 0.99 (95% CI, 0.97-0.99). The
findings of the present meta-analysis reveal that US has
an excellent diagnostic value for foot and ankle fractures.
Furthermore, the meta-analysis reveals a high PLR of
15.0 (95% CI, 7.9-28.6) and a low NLR of 0.04 (95% CI,
0.02-0.11), suggesting that the diagnostic test performs
well in correctly identifying the true disease conditions
in patients with foot and ankle fractures. The previous
review in adults identified the foot and ankle as the site of
highest sensitivity and specificity across multiple studies,
 Med Ultrason 2021; 23(2): 203-212 209
Table III. Meta-regression and subgroup analyses
Covariate No. of Studies Sensitivity (95% CI) P Value Specificity (95% CI) P Value
Locale 0.89 0.16
Turkey 5 0.96 (0.92-1.00) 0.96 (0.94-0.99)
Countries other than Turkey 5 0.96 (0.91-1.00) 0.88 (0.82-0.93)
No. of patients 0.99 0.05
≤ 100 4 0.96 (0.91-1.00) 0.91 (0.82-0.99)
> 100 6 0.96 (0.91-1.00) 0.95 (0.91-0.99)
Fracture prevalence, % 0.44 0.10
≤ 30 4 0.99 (0.96-1.00) 0.95 (0.89-1.00)
> 30 6 0.94 (0.89-0.99) 0.93 (0.87-0.98)
Year published 0.71 0.33
2009-2013 5 0.97 (0.92-1.00) 0.93 (0.86-0.99)
2014-2019 5 0.96 (0.91-1.00) 0.94 (0.89-0.99)
Reference standard 0.59 0.51
Only radiograph 7 0.97 (0.94-1.00) 0.94 (0.89-0.99)
Including CT or MRI 3 0.92 (0.84-1.00) 0.92 (0.84-1.00)
US operator 0.90 0.48
Emergency physician 5 0.97 (0.95-1.00) 0.95 (0.91-0.99)
Others 5 0.91 (0.85-0.96) 0.91 (0.85-0.98)
US training 0.93 0.40
Yes 6 0.96 (0.92-1.00) 0.94 (0.90-0.99)
Others 4 0.96 (0.91-1.00) 0.92 (0.85-1.00)
Blinding method 0.81 0.04
Double blinding 7 0.95 (0.91-1.00) 0.92 (0.87-0.97)
Others 3 0.98 (0.93-1.00) 0.97 (0.92-1.00)
POCUS 0.83 0.18
Yes 6 0.96 (0.92-1.00) 0.93 (0.88-0.98)
Others 4 0.96 (0.89-1.00) 0.94 (0.88-1.00)
POCUS, point-of-care ultrasound; US, ultrasound; CI, confidence interval

with values ranging from 85.9 to 100% and 86.4 to 100%
respectively [11]. This is consistent with the results of
our study.
As we all know, ultrasound is an operator-dependent
technology [36]. Therefore, it is of the highest impor-
tance to ensure that operators acquire sufficient train-
ing and practice with this technology. With regard to the
subgroup analysis of US training, operators with extra
ultrasonographic training had a comparable diagnostic
performance to others such as experience operators or
not reported (sensitivity: 0.96 and 0.96; specificity: 0.94
and 0.92). However, the training protocols were vari-
able among studies and the most of the training courses
ranged from 30 minutes to 2 days [26,27,29-31]. It is un-
defined what the optimal training protocol is, so further
prospective studies are required to determine the optimal
training protocol and learning curve for this technology.
According to the eligible studies, the main ultrasonic
characteristics of foot and ankle fractures included cor-
tical disruption or stepping, axial deviation of the bone
surface and cortical depression, which directly indicate
fractures [19,20,26-32]. Other ultrasonic findings such
as hypoechoic periosteal elevation, swollen soft tissues
and increased vascularity, which are often not visualized
by radiography also support foot and ankle fractures in-
directly [18]. So further prospective studies with larger
sample sizes are required to identify a well-defined pro-
tocol for ultrasound in detecting foot and ankle fractures.
Significant heterogeneity had been observed in this
meta-analysis (sensitivity: I2 = 51.43%, p=0.03; specific-
ity: I2 = 81.45%, p=0.00) and meta-regression analyses
showed that the blinding method accounted for part of
the significant sources of heterogeneity in terms of sen-
sitivity. However, there were other factors which might
involve the significant heterogeneity. Other factors such
as specialties of ultrasonic operators, different experi-
ence levels and different equipment might also play an
important role in heterogeneity among studies. However
meta-regression analyses could not be performed to ex-
plore the significant sources of heterogeneity according
to other factors referred above because of the insufficient
information in the included studies.
210 Jiangfeng Wu, Yunlai Wang et al The diagnostic accuracy of ultrasound in the detection of foot and ankle fractures
Most of the included studies chose conventional
X-ray as the reference standard, which is not the golden
standard to detect fractures. Because occult fractures ac-
count for 2%-36% in a conventional X-ray on account
of overlapping structures, under-mineralized ossification
centres and non-perpendicular X-ray beam to the fracture
line [35]. Several studies have demonstrated that multi-
planar capabilities of ultrasound might make it superior
to radiography in the detection of occult fractures that the
X-ray has missed [37,38]. Banal et al found that in cases
of normal radiographs, US is indicated in the diagnosis
of metatarsal bone stress fractures with a sensitivity of
0.83 and a specificity of 0.76 [18]. The prospective study
of ultrasonographic evaluation of radiographically nega-
tive ankle injuries in a pediatric population performed
by Simanovsky et al demonstrated that ultrasound is ef-
fective in the detecting radiographically silent fractures
of the pediatric ankle with a sensitivity of 1.00 and a
specificity of 0.97 [19]. In future, studies adopting CT or
MRI as the golden standard are required to evaluate the
diagnostic performance of ultrasound in detecting radio-
graphically occult fractures.
Atilla et al showed that the sensitivity and specific-
ity of US for foot and ankle fractures could vary accord-
ing to the fracture site. US had excellent sensitivity and
specificity in the diagnosis of the fifth metatarsal frac-
tures with a sensitivity of 1.00 and specificity of 0.96,
which might be explained by the superficial site of the
fifth metatarsal and its smooth contours and ease of view-
ing from different planes by US; while the sensitivity and
specificity of US for detecting navicular fractures was
relatively lower (sensitivity: 0.40 and specificity: 0.93)
as the dorsal surface of the navicular bone is irregular and
could only be viewed in the dorsal plane [29]. Our sub-
group analysis of the 4 studies only including metatarsal
fractures [18,20,30,32] showed pooled sensitivity of 0.95
and specificity of 0.90. However, we faced insufficient
data to perform more subgroup analyses regarding dif-
ferent fracture sites such as ankle fractures or navicular
fractures.
Foot and ankle injuries are almost universally as-
sessed by the Ottawa Foot and Ankle Rules, which have
a reported sensitivity of 97.9% to 99.8% and a specificity
of 28.8% and 42.3% for foot and ankle fractures respec-
tively [39-42]. The poor specificity of the rules indicates
that about 60% of patients who undergo a radiograph
do not have a fracture, with many patients exposed to
unnecessary harmful ionizing radiation. Hedelin et al
suggested ultrasound-guided triage seemed to be able to
decrease the need for radiographic imaging in patients
with ankle trauma [27]. Similarly, Canagasabey et al sug-
gested that US examination had been employed prior to
the X-ray, the quantity of radiographs needed would have
fallen by 80.9% [26]. Consequently, US is increasingly
being considered as a first-line modality in the primary
response to emergency situations and could decrease the
need for radiographic imaging in patients with foot and
ankle fractures.
It is important to consider some limitations with re-
spect to this study. First, a relatively small number of
studies were included in the present meta-analysis as a
result of the limited relevant high-quality studies, and the
literature search merely included studies written in Eng-
lish. Anyway, we were able to acquire several important
conclusions with respect to the diagnostic value of US
and related factors. Second, most of the eligible studies
failed to report the precise duration between the stand-
ard reference and US examination except for the study
by Banal et al who declared that US and dedicated MRI
examinations of the metatarsal bones were performed the
same day [18]. However, the reference standard and US
performed without a narrow time frame may not increase
the performance bias because fractures will not change
over time in the short term. Third, no study evaluated
intraobserver or interobserver variability; however, it is
important to do so because ultrasound is an operator-
dependent modality. Therefore, more further studies are
required to evaluate the intraobserver or interobserver
variability. Finally, most of the included studies had
methodological limitations, especially in domains such
as patient selection, the index test, reference standard and
flow and timing, and therefore improvements in the fu-
ture study design are required to accurately address the
issue under investigation.
Conclusions
In summary, this comprehensive meta-analysis dem-
onstrates that ultrasound has an excellent diagnostic
performance for foot and ankle fractures and should be
considered as a primary and radiation-free scanning mo-
dality in the diagnosis of foot and ankle fractures. How-
ever, the conclusion of this meta-analysis on the strength
of a small quantity of studies that met the specific in-
clusion criteria should be interpreted with caution. Large
prospective international multicenter studies are still re-
quired to support the present conclusion.
Conflict of interest: none
References
1. Shibuya N, Davis ML, Jupiter DC. Epidemiology of foot
and ankle fractures in the United States: an analysis of the
National Trauma Data Bank (2007 to 2011). J Foot Ankle
Surg 2014;53:606–608.

2. Crowley SG, Trofa DP, Vosseller JT, et al. Epidemiology of
foot and ankle injuries in National Collegiate Athletic As-
sociation men’s and women’s ice hockey. Orthop J Sports
Med 2019;7:2325967119865908.
3. Wood JN, Henry MK, Berger RP, et al. Use and utility of
skeletal surveys to evaluate for occult fractures in young
injured children. Acad Pediatr 2019;19:428–437.
4. Nery C, Raduan F, Baumfeld D. Foot and ankle injuries in
professional soccer players: diagnosis, treatment, and ex-
pectations. Foot Ankle Clin. 2016;21:391–403.
5. Ait-Ali L, Andreassi MG, Foffa I, Spadoni I, Vano E, Pi-
cano E. Cumulative patient effective dose and acute radi-
ation-induced chromosomal DNA damage in children with
congenital heart disease. Heart 2010;96:269–274.
6. Brenner DJ, Doll R, Goodhead DT, et al. Cancer risks at-
tributable to low doses of ionizing radiation: assessing what
we really know. Proc Natl Acad Sci U S A 2003;100:13761–
13766.
7. LiMarzi GM, Scherer KF, Richardson ML, et al. CT and
MR imaging of the postoperative ankle and foot. Radio-
graphics 2016;36:1828–1848.
8. Beard NM, Gousse RP. Current ultrasound application in
the foot and ankle. Orthop Clin North Am 2018;49:109–
121.
9. Lee SH, Yun SJ. Diagnostic performance of ultrasonogra-
phy for detection of pediatric elbow fracture: a meta-analy-
sis. Ann Emerg Med 2019;74:493–502.
10. Douma-den Hamer D, Blanker MH, Edens MA, et al. Ultra-
sound for distal forearm fracture: a systematic review and
diagnostic meta-analysis. PLoS One 2016;11:e0155659.
11. Champagne N, Eadie L, Regan L, Wilson P. The effective-
ness of ultrasound in the detection of fractures in adults
with suspected upper or lower limb injury: a systematic
review and subgroup meta-analysis.  BMC Emerg Med
2019;19:17.
12. Zhao W, Wang G, Chen B, et al. The value of ultrasound for
detecting hand fractures: A meta-analysis. Medicine (Balti-
more) 2019;98:e17823.
13. Mircea PA, Badea R, Fodor D, Buzoianu AD. Using ul-
trasonography as a teaching support tool in undergraduate
medical education - time to reach a decision. Med Ultrason
2012;14:211–216.
14. Saul T, Ng L, Lewiss RE. Point-of-care ultrasound in the
diagnosis of upper extremity fracture-dislocation. A picto-
rial essay. Med Ultrason 2013;15:230–236.
15. Sconfienza LM, Albano D, Allen G, et al. Clinical indica-
tions for musculoskeletal ultrasound updated in 2017 by
European Society of Musculoskeletal Radiology (ESSR)
consensus. Eur Radiol 2018;28:5338–5351.
16. Zayat AS, Md Yusof MY, Wakefield RJ, Conaghan PG,
Emery P, Vital EM. The role of ultrasound in assessing
musculoskeletal symptoms of systemic lupus erythemato-
sus: a systematic literature review. Rheumatology (Oxford)
2016;55:485–494.
17. Zhang CM, Zhang JF, Xu J, Guo YL, Wang G, Yang LH.
Musculoskeletal ultrasonography for arthropathy assess-
ment in patients with hemophilia: A single-center cross-
Med Ultrason 2021; 23(2): 203-212 211
sectional study from Shanxi Province, China.  Medicine
(Baltimore) 2018;97:e13230.
18. Banal F, Gandjbakhch F, Foltz V, et al. Sensitivity and spec-
ificity of ultrasonography in early diagnosis of metatarsal
bone stress fractures: a pilot study of 37 patients. J Rheu-
matol 2009;36:1715–1719.
19. Simanovsky N, Lamdan R, Hiller N, Simanovsky N. So-
nographic detection of radiographically occult fractures
in pediatric ankle and wrist injuries.  J Pediatr Orthop
2009;29:142–145.
20. Yesilaras M, Aksay E, Atilla OD, Sever M, Kalenderer O.
The accuracy of bedside ultrasonography as a diagnostic
tool for the fifth metatarsal fractures.  Am J Emerg Med
2014;32:171–174.
21. Shamseer L, Moher D, Clarke M, et al. Preferred report-
ing items for systematic review and meta-analysis proto-
cols (PRISMA-P) 2015: elaboration and explanation. BMJ
2015;350:g7647. 
22. Whiting PF, Rutjes AW, Westwood ME, et al. QUADAS-2:
a revised tool for the quality assessment of diagnostic ac-
curacy studies. Ann Intern Med 2011;155:529–536.
23. Hanley JA, McNeil BJ. The meaning and use of the area
under a receiver operating characteristic (ROC) curve. Ra-
diology 1982;143:29–36.
24. Higgins JP, Thompson SG, Deeks JJ, Altman DG. Meas-
uring inconsistency in meta-analyses. BMJ 2003;327:557–
560.
25. Deeks JJ, Macaskill P, Irwig L. The performance of tests
of publication bias and other sample size effects in system-
atic reviews of diagnostic test accuracy was assessed. J Clin
Epidemiol 2005;58:882–893.
26. Canagasabey MD, Callaghan MJ, Carley S. The sono-
graphic Ottawa Foot and Ankle Rules study (the SOFAR
study). Emerg Med J 2011;28:838–840.
27. Hedelin H, Goksör LÅ, Karlsson J, Stjernström S. Ultra-
sound-assisted triage of ankle trauma can decrease the need
for radiographic imaging. Am J Emerg Med 2013;31:1686–
1689.
28. Ekinci S, Polat O, Günalp M, Demirkan A, Koca A. The
accuracy of ultrasound evaluation in foot and ankle trau-
ma. Am J Emerg Med 2013;31:1551–1555.
29. Atilla OD, Yesilaras M, Kilic TY, et al. The accuracy of
bedside ultrasonography as a diagnostic tool for fractures
in the ankle and foot.  Acad Emerg Med 2014;21:1058–
1061.
30. Kozaci N, Ay MO, Avci M, et al. The comparison of ra-
diography and point-of-care ultrasonography in the di-
agnosis and management of metatarsal fractures.  Injury
2017;48:542–547.
31. Ozturk P, Aksay E, Oray NC, Bayram B, Basci O, Tokgoz
D. Emergency physician accuracy using ultrasonography
to diagnose lateral malleolar fracture.  Am J Emerg Med
2018;36:362–365.
32. Ebrahimi M, Habibzadeh SR, Ahmadi SR, Khajeh Nasiri
S, Kaveh MM, Foroughian M. Diagnostic Accuracy of Ul-
trasonography in Diagnosis of Metatarsal Bone Fracture; a
Cross Sectional Study. Arch Acad Emerg Med 2019;7:e49.
212 Jiangfeng Wu, Yunlai Wang et al The diagnostic accuracy of ultrasound in the detection of foot and ankle fractures
33. Gordon I, Sinert R, Chao J. The utility of ultrasound in 38. Allen GM, Wilson DJ, Bullock SA, Watson M. Extrem-
detecting skull fractures after pediatric blunt head trauma: ity CT and ultrasound in the assessment of ankle inju-
systematic review and meta-analysis.  Pediatr Emerg Care ries: occult fractures and ligament injuries.  Br J Radiol
2020, doi:10.1097/PEC.0000000000001958.  2020;93:20180989.
34. Chartier LB, Bosco L, Lapointe-Shaw L, Chenkin J. Use of 39. Bachmann LM, Kolb E, Koller MT, Steurer J, ter Riet G.
point-of-care ultrasound in long bone fractures: a system- Accuracy of Ottawa ankle rules to exclude fractures of the
atic review and meta-analysis. CJEM 2017;19:131–142. ankle and mid-foot: systematic review. BMJ 2003;326:417.
35. Douma-den Hamer D, Blanker MH, Edens MA, et al. Ultra- 40. Auleley GR, Ravaud P, Giraudeau B, et al. Implementation
sound for distal forearm fracture: a systematic review and of the Ottawa ankle rules in France. A multicenter rand-
diagnostic meta-analysis. PLoS One 2016;11:e0155659. omized controlled trial. JAMA 1997;277:1935–1939.
36. Scheel AK, Schmidt WA, Hermann KG, et al. Interobserver 41. Lucchesi GM, Jackson RE, Peacock WF, Cerasani C,
reliability of rheumatologists performing musculoskeletal Swor RA. Sensitivity of the Ottawa rules. Ann Emerg Med
ultrasonography: results from a EULAR “Train the train- 1995;26:1–5.
ers” course. Ann Rheum Dis 2005;64:1043–1049. 42. Beckenkamp PR, Lin CC, Macaskill P, Michaleff ZA, Ma-
37. Kwee RM, Kwee TC. Ultrasound for diagnosing ra- her CG, Moseley AM. Diagnostic accuracy of the Ottawa
diographically occult scaphoid fracture.  Skeletal Radiol Ankle and Midfoot Rules: a systematic review with meta-
2018;47:1205–1212. analysis. Br J Sports Med 2017;51:504–510.
Review Med Ultrason 2021, Vol. 23, no. 2, 213-219
DOI: 10.11152/mu-2727

Understanding the role of echocardiography in patients with

obstructive sleep apnea and right ventricular subclinical myocardial
dysfunction – comparison with other conditions affecting RV
deformation
Ioana Maria Chetan1, Bianca Domokos Gergely 2, Adriana Albu3, Raluca Tomoaia1, Doina
Adina Todea2

1HeartInstitute “N. Stancioiu”, 2Pulmonology Department, 32nd Internal Medicine Department, “Iuliu Hatieganu”
University of Medicine and Pharmacy Cluj Napoca, Romania

Abstract
Despite efforts to treat obstructive sleep apnea syndrome (OSA), the condition remains an important risk factor in the
development of cardiovascular disease. Early detection of RV dysfunction with novel echocardiographic techniques (speckle
tracking echocardiography) may be useful in preventing progression to pulmonary hypertension, with subsequent heart failure
and cardiovascular death. Echocardiography is the method of choice for the evaluation of OSA consequences on the heart.
Although standard echocardiographic parameters are routinely used in these patients, there are several limitations in the early
detection of RV dysfunction. The main concerns are the complex geometry of RV and the impact of pre- and afterload on RV
myocardium, which cannot be assessed through standard measurements. The aim of this review is to highlight the utility of
advanced echocardiographic parameters in the identification of OSA patients with subclinical myocardial dysfunction, which
are at risk of developing heart failure and later adverse events. Speckle tracking echocardiography might provide higher sensi-
tivity in unmasking alterations in RV function when conventional echocardiographic methods cannot detect them. Therefore,
this method has a major role in the detection of early stages of RV dysfunction, along with better risk stratification and better
timing in the initiation of therapy.
Keywords: obstructive sleep apnea; speckle tracking; echocardiography

Introduction tor, alongside smoking, alcohol consumption and male

Obstructive sleep apnea (OSA) is a potentially dan-
gerous sleep disorder, with an incidence of about 4-7% in
the general adult population [1]. It is characterized by the
presence of intermittent episodes of partial or complete
obstruction of the upper airways during sleep, leading to
intermittent hypoxemia, fragmentation of sleep, snoring
and daytime sleepiness. Obesity is an important risk fac-
Received 13.07.2020  Accepted 15.10.2020
Med Ultrason
2021, Vol. 23, No 2, 213-219
Corresponding author: Bianca Domokos Gergely
“Leon Daniello” Pneumology Hospital,
6 BP Hasdeu street,
400332 Cluj-Napoca, Romania
Phone: +40741499968
E-mail: biancadomokos@yahoo.com
sex. The hallmark of OSA is the craniofacial morphol-
ogy, with its contribution to upper airway collapsibility
[2]. Some of these risk factors for OSA are also estab-
lished risk factors for cardiovascular diseases, making
it difficult to demonstrate a causal relationship between
OSA cardiac damage independent of these potentially
confounding factors [3].
In order to diagnose OSA, sleep studies (polysomnog-
raphy) are required; these can quantify the apnea-hypo-
pnea index (AHI), which places OSA in a severity class
(mild, moderate and severe, respectively 5-14, 15-30 and
>30 events/hour) [4]. A large number of epidemiologi-
cal evidences describes a link between OSA syndrome
and cardiovascular morbidity and mortality [1,4,5], evi-
dence that has prompted the American Heart Association
(AHA) to issue a scientific statement describing the need
214 Ioana Maria Chetan et al Echocardiography in patients with obstructive sleep apnea and RV subclinical myocardial dysfunction
to recognize this pathology as an important target for car-
diovascular disease risk reduction therapy [5]. Front-line
treatment of OSA implies stabilizing the upper airway
mechanically, by introducing a column of air via continu-
ous positive airway pressure (CPAP). Although CPAP is
considered the “gold standard” treatment, its results are
narrowed by poor patient adherence to, due to various
side effects [6]. Several forms of cardiovascular diseas-
es (CVD) including stroke, heart failure, hypertension,
coronary artery disease, arrhythmias (atrial fibrillation),
have been associated with OSA [7]. However, the exact
effect of OSA on cardiac structure and function is not
fully understood [8]. Previous reports have shown that
pulmonary hypertension (PH) occurs in up to 40% of pa-
tients with OSA along with heart or pulmonary disease,
but it may also develop in OSA patients in the absence of
known cardiopulmonary disorders. Although PH is mild
to moderate in OSA, it can lead to RV hypertrophy and
dysfunction. [9]. Podszus et al [10] found that OSA is
a predisposing factor for the development of pulmonary
hypertension, but there are still few and discrepant data
concerning the repercussions of OSA on the structural
and functional changes of the right heart [11].
A comprehensive echocardiography screening for
subclinical heart failure in OSA patients is needed. It is
well known that assessment of the right ventricle (RV)
by echocardiography in day-to-day practice is often
based on qualitative evaluation, thus being recognized
as the “forgotten chamber” [12]. It is important to ob-
tain a complete assessment of the RV, quantitative as
well as qualitative, using combined techniques. Three-
dimensional echocardiography (3DE) and speckle track-
ing echocardiography (STE) allow the evaluator to over-
come the limitations of conventional echocardiography.
The aim of this review was to highlight the utility of
advanced echocardiography parameters in identifying
OSA patients with subclinical myocardial dysfunction,
along with better risk stratification and well-timed initia-
tion of therapy.
Is 3D echocardiography a reality of daily clinical
practice when assessing the RV?
Most conventional methods of echocardiographic
evaluation of the RV function and dimensions are based
on approximate volumetric measurements. Such ap-
proaches are limited by the complex geometry of the
RV and by their angle or load dependency. As a rule,
the problem of the complex geometry of the RV could
be solved through real time 3DE and that of the angle/
load dependence through STE (fig 1). Compared to bi-
dimensional echocardiography (2DE), the real time 3DE
allows a better anatomical definition of the RV, with good
highlighting of the base, apex and outflow tract [11,13].
Fig 1. a) Representative right ventricular (RV) quantification
using EchoPAC software. After tracing the endocardial border
at end-diastole and end-systole b) 3D reconstruction of the RV
was automatically realized; c) 3D RV ejection fraction (3D RV
EF) and 3D RV fractional area change (32.4 %) in a patient with
moderate obstructive sleep apnea. 3D-RVEF=36.1%. EDV in-
dicates end-diastolic volume; EF, ejection fraction; ESV, end-
systolic volume; RV, right ventricular; RVEF, RV ejection
fraction; SV, stroke volume; TAPSE, tricuspid annular plane
systolic excursion; FAC, fractional area change; Dd, diameter.
Novel techniques using software based on RV 3DE ac-
quisitions have proven the utility to accurately and re-
producibly assess the RV dimensions and function [14].
A first study to assess 3DE derived global and regional
RV shape parameters in normal subjects and in patients
with PH was developed by Addetia et al [15]. The au-
thors showed that in PH, RV expresses modifications in
regional curvature evident in apical and septal free-wall
regions and in the outflow tract, when compared to nor-
mal RV. In this study, the outflow tract was found to be
rounder, with more convex apical and body portions of
the septum, bulging into the left ventricle (LV) at end-
diastole and end-systole, while the apical free wall was
found to be more flattened. These findings demonstrate
that curvature analysis using 3DE allows a better quanti-
tative assessment of RV remodeling, due to the tracking
of changes in the regional RV shape [15]. The feasibility,
accuracy, and reproducibility of RV volumes and ejection
fraction (EF) assessment using real-time 3D echocardio-
graphic imaging were pointed out in several studies. For
example, the value of 3DE in the evaluation of RV struc-
ture and function in patients with congenital heart disease
(CHD) was assessed by van der Zwaan et al [16] who used
cardiac magnetic resonance imaging (CMR) as a refer-
ence. They have compared 3DE derived RV volumes and
EF with 2DE derived measurements for RV assessment
and demonstrated that 3DE showed a higher specificity
to exclude RV myocardial dysfunction. These findings
are in agreement with other published studies. Moreo-
ver, Nagata et al showed that in 446 patients, 3DE RV-
EF was independently corelated with cardiac outcomes

in subjects with diverse backgrounds. Furthermore, 3DE
RV-EF offered additional values over clinical risk factors
and other echocardiographic parameters, in prediction of
future cardiovascular adverse outcomes [17].
Other advantages of 3DE are the relatively short time
required for off-line analysis of acquisitions and the good
quality of 3D images used for reconstructions, which
were demonstrated by the study of Tamborini et el [18].
Several previous studies have proven the reproducibility
and fidelity of 3DE in the evaluation of RV systolic func-
tion and global volume [19-20]. For example, Zhou et al
[3] concluded that using real time 3DE in the assessment
of RV systolic function is an important key in the nonin-
vasive evaluation of severity in OSA; furthermore, they
have demonstrated patients with OSA, RV global EF as-
sessed by 3DE was lower than in controls with an inverse
relationship to the severity of OSA as assessed by AHI.
Similar results were obtained by Vitarelli et al [21], who
demonstrated that RVEF assessed by 3DE was lower in
subjects with moderate-sever OSA compared with con-
trols, irrespective of the presence of PH. Moreover, the
authors have concluded that changes in parameters were
independently associated with AHI. Interestingly, after
using CPAP for 4 months, the subgroup with severe OSA
showed a significantly lower pulmonary artery pressure
than that of patients at baseline and an improvement in
RV dysfunction with an increase of RV EF.
It is well known that CMR represents the gold stand-
ard in terms of quantitative assessment of the RV; how-
ever, its common use in daily practice is limited because
it is expensive, time-consuming and sometimes contrain-
dicated. In a number of studies, the accuracy of 3DE
was compared with CMR, and they were in good con-
cordance [22]. Jenkins et al confirmed that 3DE has less
variation and higher correlation to CMR than all 2DE
techniques, when being compared in patients after acute
myocardial infarction [23]. Also, when comparing 3DE
and 2DE with CMR, 3DE EF was the most reliable echo-
cardiographic variable for recognizing CMR-derived
RV EF <50% [24]. On the other hand, Kjaergaard et al
[25], found at best modest correlations between 3DE
and CMR, concluding that tricuspid annular plane sys-
tolic excursion (TAPSE) is the best option for RV func-
tion assessment for daily clinical purposes. However, it
is important to keep in mind that TAPSE assesses only
longitudinal RV shortening, which does not represent the
RV global function, especially if extensive regional ab-
normalities are present [26].
In order to use 3DE for RV assessment it is important
to remember that the RV cavity borders can be difficult
to define, given the numerous trabeculations of RV free
wall. This may be the explanation for the differences
Med Ultrason 2021; 23(2): 213-219 215
between 3DE and CMR in the evaluation of RV func-
tion and volumes and also the reason why there are cases
when RV chamber is underestimated using 3DE [27].
Speckle tracking echocardiography
in the evaluation of the RV:
the future is here, what`s next?
Conventional and Tissue-Doppler Echocardiogra-
phy allow the evaluation of RV function through frac-
tional area change (FAC), myocardial performance index
(MPI), TAPSE and lateral S’ wave by tissue-Doppler.
STE is a relatively new technique used in the evaluation
of myocardial function and is less angle independent.
As a principle of this method, random noise is filtered
out, while preserving unique stable myocardial features,
mentioned as speckles. These blocks of speckles can
be unmasked within the myocardium based on frame-
by-frame tracking, simultaneously in several regions of
a plan; therefore, the parameters of myocardial func-
tion (velocity, strain, strain rate), are being evaluated in
the longitudinal, radial and circumferential ventricular
axis. Basically, information on both global and segmen-
tal myocardial deformation is provided by this method
[28]. However, these parameters are limited by load and
angle dependency and do not assess all the axis of RV
myocardial contractility. Several studies that have used
these conventional echocardiographic parameters re-
ported changes in RV function and structure in patients
with OSA [29,30]. Moreover, some studies revealed no
abnormalities of RV structure and function in OSA pa-
tients [31-32]. Thus, it remains a question when the other
techniques are necessary, in order to detect more subtle
changes in RV deformation, which might be present in
patients with OSA. Novel echocardiographic techniques
including 2DE and 3DE speckle tracking methods might
therefore provide higher sensitivity in unmasking altera-
tions in RV function and structure, in the early stages of
RV dysfunction, when conventional echocardiographic
methods cannot detect them [9,33,34]. These methods
proved their utility in the assessment of RV function in
acute conditions. For example, Trivedi et al in a retro-
spective study conducted on patients with pulmonary
embolism (PE), demonstrated that RV function assessed
by 2D RV free wall strain (FWS) was reduced in PE pa-
tients compared with controls. Adding RV FWS to the
existing parameters of RV size and function significantly
improved the sensitivity and specificity of the diagnosis
of PE and might play a role in guiding treatment. [35]
More recently Li et al evaluated in 120 patients suffering
from COVID-19, the relationship between echocardiog-
raphy variables (including 2D RV LS) and clinical vari-
ables to predict patient prognosis including mortality. RV
LS predicted mortality, with a sensitivity and specificity
216 Ioana Maria Chetan et al Echocardiography in patients with obstructive sleep apnea and RV subclinical myocardial dysfunction
of 94.4 % and 64.7% respectively. The prognostic value
of RV LS was better than other indicators of RV function
used alone or in combination, and it was maintained in
univariate and multivariate Cox regression analysis [36].
2DE speckle tracking of the RV
in patients with OSA
RV function plays an important role in the morbidity
and mortality of patients with OSA. Early detection of
RV dysfunction before the onset of pulmonary hyperten-
sion is crucial in preventing progression to heart failure
and cardiac death [37]. This became possible using 2D
speckle tracking echocardiography. Studies showed that
RV longitudinal strain had a predicted value in several
cardiovascular diseases (CHD, heart failure, valvular
heart disease, PH and OSA). RV strain in OSA patients
has not been yet rigorously and exhaustively explored,
but there are investigations on this subject with encourag-
ing results. [38]. Recent studies have shown that patients
with OSA have shown impairment of global longitudinal
strain (GLS-RV) and lateral wall strain (LS-RV) in the
presence of preserved RV-EF. As Buonauro et al [39],
concluded in their study, OSA patients showed signifi-
cant reduction in RV GLS (18.2±2.4%, p<0.001) and LS-
RV (17.8±4.8%, p<0.001), but not in septal longitudinal
strain (SLS-RV) (18.8±3.6 %, p=0.229) compared with
healthy controls. RV-GLS correlated with OSA severity,
in the absence of important alteration of other echocar-
diographic parameters of RV systolic function, such as
3DE-EF or TAPSE. Li et al [40] showed significantly
reduced RV LS and strain rates of RV apical segments
in patients with mild OSA compared to controls, results
confirmed by the study of Kepez et al [11]. D’Andrea
et al [41], revealed a reduction in both LV and RV GLS
(-13.8±5.2 %, p<0.01), as well as RV LS in OSA patients
compared with controls. In concordance with these find-
ings, Hammerstingl et al [42], found that RV 2DE GLS
(-14.5 ± 8.2 %, p<0.0001) was significantly lower in sub-
jects with higher AHI, compared with controls; and was
positively associated with the severity of OSA. More-
over, conventional echocardiography parameters, such as
TAPSE, RV MPI, were not significantly impaired in these
patients, supposing that they are not sufficiently sensitive
in detecting subclinical changes of RV function in OSA.
RV dysfunction might be the expression of pulmonary
hypertension. This hypothesis is supported by several
studies, which demonstrate that patients with OSA and
right ventricular systolic pressure (RVSP) ≥30 mmHg,
show substantially lower GLS-RV (18.2±2.4%, p<0.001),
LLS-RV (17.8± 4.8%, p<0.001), compared with patients
with RVSP<30 mmHg (GLS-RV=22.7±4.8%, LLS-
RV=25.1±7.0%) [28]. However, these data are in op-
position with the results obtained by Altekin et al [37],
who did not find any differences in stain or strain rate
at patients with OSA and pulmonary hypertension. Con-
cerning RV strain in OSA patients during CPAP therapy,
free wall RV strains and global strain improved after
CPAP treatment, even in the presence of unchanged con-
ventional parameters [39]. Kim et al showed that CPAP
therapy, unlike sham therapy significantly improved RV
GLS and also reduced RV dimension [43]. Other data re-
port strain improvement under CPAP therapy in apical
segments, but not in the medial or basal ones, explain-
ing why RV global longitudinal strain (average of apical,
medial and basal segments) remained unchanged after 6
months of CPAP therapy [42].
3D speckle tracking in the evaluation of the RV:
a new era
The RV has a complex pattern of contraction, which
occurs along three anatomically relevant axes: longitudi-
nal shortening with tricuspid annulus traction towards the
apex, anteroposterior shortening by stretching the free
wall over the interventricular septum and radial move-
ment of the free wall, also known as the “bellows effect“.
Although it is well known that the subepicardial layer of
the myocardium is rich in circumferential myofibers, in
the daily echocardiographic evaluation of the right ven-
tricle, non-longitudinal motion directions are often omit-
ted [44]. Therefore, the mix of deformation marks pro-
vides insight into the pathophysiologic mechanics of RV
dysfunction. In comparison to 2D STE, the advantage of
using 3D STE consists of the possibility of simultaneous
assessment of various wall motion indices, in particular:
radial strain (RS), longitudinal strain (LS) and circumfer-
ential strain (CS) in the entire RV myocardium. Another
unique parameter assessed by 3DE STE is the area change
ratio (ACR) or area strain in a regional or entire track-
ing area [45]. The clinical utility and feasibility of the
3D STE of RV technique were evaluated by Atsumi et al
[46], in both an experimental and clinical study, by using
sonomicrometry crystals implanted in the RV endocar-
dium, in order to validate 3D STE data. The authors ob-
tained significant correlation between the two techniques.
The PH group had significantly lower RV global defor-
mation parameters than the control group. These findings
confirm the results of previous studies that used 3D STE
modified for RV software, showing that PH patients have
reduced RV strain values, poorer outcomes, lowered
ACR, LS and CS and increased mortality risk compared
with controls [47]. On the other hand, Liu et al evalu-
ated and compared the value of RV LS detected by both
2DE and 3D STE in the risk stratification of patients with
pre-capillary PH. Authors demonstrated that 2D RV LS is
better than 3D RV LS for the detections of patients with
pre-capillary hypertension intermediate-high risk [48].

3D speckle tracking echocardiography for the
RV: are regional deformation parameters also
necessary?
The myofibers in the myocardium are a complex 3D
network, arranged as a multiple helical display with im-
portant functional effect. The orientation of these my-
ofibers is important because it helps to understand the
pattern of RV contraction in 3D space [42]. Knowing that
3D STE allows a better perception of RV wall dynamic/
motion, the question is which of the 3D strain parameters
is more dependable.
Atsumi et al [46] showed in their experimental study
(10 sheep), that RV ACR and RV LS were significantly
lower in moderate pulmonary artery banding (peak RV
pressure>40 mmHg) and in severe pulmonary artery
banding (peak RV pressure >60 mmHg) compared to
baseline values; however, RV CS differed significantly
only between baseline and sever PAB, revealing that
longitudinal contraction may be more reactive to pres-
sure overload. The difference between LS and CS was
explained by the RV myofibers arrangement: the endo-
cardial layer contains longitudinal myofibers, the epicar-
dial layer comprises circumferential myofibers, whereas
the epicardial fiber is continuous with the left ventricu-
lar myocardium. Therefore, circumferential contraction
might be less affected by pressure overload solely in the
RV due to its preservation by LV contraction.
On the other hand, in a study comparing 97 patients
suffering from PHT with 60 controls, Smith et al [47]
demonstrated that all PHT patients had reduced RV
strain indices compared with controls; RV AS and CS
were strongly correlated to RV EF (p<0.001). LS and RS
were also correlated to RV EF, but lesser. Moreover, the
authors evaluated mortality at 24 months, showing that
RV AS, CS and LS predicted it most accurately. Further-
more, only AS, and in a lesser level, age, were predictors
of death, underling the superiority of 3D STE RV – AS
over other parameters. A possible explication for why
RV CS may be more affected in the group of patients
with PHT, is that important structural changes happen
during the development of PHT, including a change in
fiber orientation and increased wall stress, leading to re-
duced circumferential contraction (fig 2).
3D speckle tracking of RV in patients with OSA:
future directions
The impact of OSA on the right ventricular structure
and function has been demonstrated, since PAH, impair-
ment of RV systolic function and RV dyssynchrony have
an important influence on cardiovascular morbidity and
mortality. 3DE speckle tracking allows a better and more
comprehensive assessment of the RV wall motion. More-
over, this novel technique is able to detect alterations in
Med Ultrason 2021; 23(2): 213-219 217
Fig 2. Representative 3D RV strain images in two patients with
obstructive sleep apnea. To asses global and longitudinal RV
systolic function, we used a 6 segment RV model basal RV free
wall, mid RV free wall, apical RV wall, apical septum, mid sep-
tum, and basal septum). In both subjects left ventricle function
was normal. a) 3D RV speckle-tracking bulls eye in a patient
with severe obstructive sleep apnea and PH (decrease of GPSL
RV = -15). b) 3D RV speckle-tracking bulls eye in a patient with
mild to moderate obstructive sleep apnea without PH (GPSL
RV = -28). GPSL, global peak longitudinal strain; PH, pulmo-
nary hypertension; RV, right ventricular.
subclinical stages of RV dysfunction. Furthermore, it
might be useful in evaluating the efficacy of continuous
positive air pressure therapy in improving cardiac func-
tion. Thus, an earlier initiation of OSA therapy can re-
duce cardiovascular risk.
Although these data provide motivation for using the
more recent echocardiographic techniques for the assess-
ment of cardiac function in patients with OSA and their
response to therapy, several issues need further investi-
gation in order to be accepted as mainstream methods
for the quantitative assessment of RV function. Speckle
tracking imaging is a more sensitive technique for detect-
ing subclinical alterations in ventricular functions that
contrarily may be missed by conventional echocardiog-
raphy. The favourite deformation parameter is the global
longitudinal strain due to its reproducibility and feasibil-
ity. 3DE is useful in the determination of ejection frac-
tion, RV volumes and intracardiac anatomy. It is strongly
recommended to develop and validate a dedicated echo-
cardiography program for each specific OSA lesion and
to combine deformation imaging with 3DE [49].
Conclusion
Deterioration of RV mechanics occurs before func-
tional and structural heart damage. Therefore, recent
echocardiographic techniques seem to be the cornerstone
of detecting subtle cardiac changes in the RV. Since RV
demonstrated an important predictive value in several
pathologies, it should be more widely used in the evalu-
218 Ioana Maria Chetan et al Echocardiography in patients with obstructive sleep apnea and RV subclinical myocardial dysfunction
ation of patients with OSA, which might present early
changes in RV contraction, even in the presence of an
apparently preserved systolic function. Risk stratifica-
tion in these patients using modern echocardiography
techniques might bring future information regarding the
impact on the cardiac function of the disease and might
also offer guidance for early therapy in order to prevent
further damage on the systolic RV function.
Conflict of interest: none
References
1. Dong R, Dong Z, Liu H, Shi F, Du J. Prevalence, Risk Fac-
tors, Outcomes, and Treatment of Obstructive Sleep Apnea
in Patients with Cerebrovascular Disease: A Systematic Re-
view. J Stroke Cerebrovasc Dis 2018;27:1471-1480.
2. Hersi AS. Obstructive sleep apnea and cardiac arrhythmias.
Ann Thorac Med 2010;5:10.
3. Zhou NW, Pan CZ, Kong DH, et al. A novel method for
sensitive determination of subclinical right ventricular sys-
tolic dysfunction in patients with obstructive sleep apnea.
Clin Respir J 2017;11:951-959.
4. Lee CHK, Leow LC, Song PR, Li H, Ong TH. Acceptance
and Adherence to Continuous Positive Airway Pressure
Therapy in patients with Obstructive Sleep Apnea (OSA)
in a Southeast Asian privately funded healthcare system.
Sleep Sci 2017;10:57-63.
5. Bradley TD, Floras JS. Obstructive sleep apnoea and its
cardiovascular consequences. Lancet 2009;373:82-93.
6. Broström A, Årestedt KF, Nilsen P, Strömberg A, Ulander
M, Svanborg E. The side-effects to CPAP treatment inven-
tory: the development and initial validation of a new tool
for the measurement of side-effects to CPAP treatment. J
Sleep Res 2010;19:603-611.
7. Shahar E, Whitney CW, Redline S, et al. Sleep-disordered
breathing and cardiovascular disease: cross-sectional re-
sults of the Sleep Heart Health Study. Am J Respir Crit Care
Med 2001;163:19-25.
8. Stradling J, Davies RJ. Sleep apnea and hypertension—
what a Mess! Sleep 1997;20:789-793.
9. Maripov A, Mamazhakypov A, Sartmyrzaeva M, et al.
Right Ventricular Remodeling and Dysfunction in Obstruc-
tive Sleep Apnea: A Systematic Review of the Literature
and Meta-Analysis. Can Respir J 2017;2017: 1587865.
10. Podszus T, Bauer W, Mayer J, Penzel T, Peter JH, Wichert
P. Sleep apnea and pulmonary hypertension. Klin Wochen-
schr 1986;64:131-134.
11. Kepez A, Niksarlioglu E, Hazirolan T, et al. Early Myo-
cardial Functional Alterations in Patients with Obstructive
Sleep Apnea Syndrome. Echocardiography 2009;26:388-
396.
12. Kossaify A. Echocardiographic Assessment of the Right
Ventricle, from the Conventional Approach to Speckle
Tracking and Three-Dimensional Imaging, and Insights
into the “Right Way” to Explore the Forgotten Chamber.
Clin Med Insights Cardiol 2015;9:65-75.
13. Addetia K, Muraru D, Badano LP, Lang RM. New Direc-
tions in Right Ventricular Assessment Using 3-Dimensional
Echocardiography. JAMA Cardiol 2019;4:936-944.
14. Narang A, Freed B. The Future of Imaging in Pulmonary
Hypertension: Better Assessment of Structure, Function,
and Flow. Adv Pulm Hypertens 2019;18:126-133.
15. Addetia K, Maffessanti F, Yamat M, et al. Three-dimen-
sional echocardiography-based analysis of right ventricular
shape in pulmonary arterial hypertension. Eur Heart J Car-
diovasc Imaging 2016;17:564-575.
16. van der Zwaan H, Helbing W, McGhie J, et al. Clinical
Value of Real-Time Three-Dimensional Echocardiography
for Right Ventricular Quantification in Congenital Heart
Disease: Validation With Cardiac Magnetic Resonance Im-
aging. J Am Soc Echocardiogr 2010;23:134-140.
17. Nagata Y, Wu VC, Kado Y, et al. Prognostic Value of
Right Ventricular Ejection Fraction Assessed by Tran-
sthoracic 3D Echocardiography. Circ Cardiovasc Imaging
2017;10:e005384.
18. Tamborini G, Brusoni D, Torres Molina JE, et al. Feasi-
bility of a New Generation Three-Dimensional Echocardi-
ography for Right Ventricular Volumetric and Functional
Measurements. Am J Cardiol 2008;102:499-505.
19. Renella P, Marx GR, Zhou J, Gauvreau K, Geva T. Feasibil-
ity and Reproducibility of Three-Dimensional Echocardio-
graphic Assessment of Right Ventricular Size and Function in
Pediatric Patients. J Am Soc Echocardiogr 2014;27:903-910.
20. Lang RM, Badano LP, Mor-Avi V, et al. Recommendations
for Cardiac Chamber Quantification by Echocardiography
in Adults: An Update from the American Society of Echo-
cardiography and the European Association of Cardiovas-
cular Imaging. J Am Soc Echocardiogr 2015;28:1-39.e14.
21. Vitarelli A, Terzano C, Saponara M, et al. Assessment of
Right Ventricular Function in Obstructive Sleep Apnea
Syndrome and Effects of Continuous Positive Airway Pres-
sure Therapy: A Pilot Study. Can J Cardiol 2015;31:823-
831.
22. Shimada YJ, Shiota M, Siegel RJ, Shiota T. Accuracy of
Right Ventricular Volumes and Function Determined by
Three-Dimensional Echocardiography in Comparison with
Magnetic Resonance Imaging: A  Meta-Analysis Study. J
Am Soc Echocardiogr 2010;23:943-953.
23. Jenkins C, Chan J, Bricknell K, Strudwick M, Marwick TH.
Reproducibility of Right Ventricular Volumes and Ejection
Fraction Using Real-time Three-Dimensional Echocardiog-
raphy. Chest 2007;131:1844-1851.
24. Knight DS, Grasso AE, Quail MA, et al. Accuracy and Re-
producibility of Right Ventricular Quantification in Patients
with Pressure and Volume Overload Using Single-Beat
Three-Dimensional Echocardiography. J Am Soc Echocar-
diogr 2015;28:363-374.
25. Kjaergaard J, Petersen CL, Kjaer A, Schaadt BK, Oh JK,
Hassager C. Evaluation of right ventricular volume and
function by 2D and 3D echocardiography compared to
MRI. Eur J Echocardiogr 2006;7:430-438.
26. Morcos P, Vick GW 3rd, Sahn DJ, Jerosch-Herold M, Shur-
man A, Sheehan FH. Correlation of right ventricular ejec-

tion fraction and tricuspid annular plane systolic excursion
in tetralogy of Fallot by magnetic resonance imaging. Int J
Cardiovasc Imaging 2008;25:263-270.
27. Zou H, Leng S, Xi C, et al. Three-dimensional biventricu-
lar strains in pulmonary arterial hypertension patients using
hyperelastic warping. Comput Methods Programs Biomed
2020;189:105345.
28. Mor-Avi V, Lang RM, Badano LP, et al. Current and
evolving echocardiographic techniques for the quantita-
tive evaluation of cardiac mechanics: ASE/EAE consensus
statement on methodology and indications endorsed by the
Japanese Society of Echocardiography. J Am Soc Echocar-
diogr 2011;24:277-313.
29. Kasikcioglu HA, Karasulu L, Tartan Z, Kasikcioglu E, Cu-
hadaroglu C, Cam N. Occult cardiac dysfunction in patients
with obstructive sleep apnea syndrome revealed by tissue
Doppler imaging. Int J Cardiol 2007;118:203-205.
30. Dursunoğlu N, Dursunoğlu D, Kılıç M. Impact of obstruc-
tive sleep apnea on right ventricular global function: sleep
apnea and myocardial performance index. Respiration
2005;72:278-284.
31. Otto ME, Belohlavek M, Romero-Corral A, et al. Compari-
son of Cardiac Structural and Functional Changes in Obese
Otherwise Healthy Adults With Versus Without Obstructive
Sleep Apnea. Am J Cardiol 2007;99:1298-1302.
32. Tugcu A, Guzel D, Yildirimturk O, Aytekin S. Evaluation of
Right Ventricular Systolic and Diastolic Function in Patients
with Newly Diagnosed Obstructive Sleep Apnea Syndrome
without Hypertension. Cardiology 2009;113:184-192.
33. Mendoza-Vázquez J, Steiner S, Esquinas AM. Acute and
chronic effects of noninvasive ventilation on left and right
myocardial function in patients with obstructive sleep ap-
nea syndrome: a speckle tracking echocardiographic study.
Echocardiography 2016;33:1623-1624.
34. Mandoli GE, De Carli G, Pastore MC, et al. Right cardiac
involvement in lung diseases: a multimodality approach
from diagnosis to prognostication. J Intern Med 2020 Sep
30. doi:10.1111/joim.13179.
35. Trivedi SJ, Terluk AD, Kritharides L, et al. Right ventricu-
lar speckle tracking strain echocardiography in patients
with acute pulmonary embolism. Int J Cardiovasc Imaging
2020;36:865-872.
36. Li Y, Li H, Zhu S, et al. Prognostic Value of Right Ven-
tricular Longitudinal Strain in Patients With COVID-19.
JACC Cardiovasc Imaging 2020 Apr 28. doi:10.1016/j.
jcmg.2020.04.014.
37. Altekin RE, Karakas MS, Yanikoglu A, et al. Determination
of right ventricular dysfunction using the speckle tracking
echocardiography method in patients with obstructive sleep
apnea. Cardiol J 2012;19:130-139.
Med Ultrason 2021; 23(2): 213-219 219
38. Tadic M, Cuspidi C, Grassi G, Mancia G. Obstructive
sleep apnea and cardiac mechanics: how strain could help
us? Heart Fail Rev 2020 Feb 3. doi:10.1007/s10741-020-
09924-0
39. Buonauro A, Galderisi M, Santoro C, et al. Obstructive
sleep apnoea and right ventricular function: A combined as-
sessment by speckle tracking and three-dimensional echo-
cardiography. Int J Cardiol 2017;243:544-549.
40. Li J, Wang Z, Li Y, et al. Assessment of regional right ven-
tricular systolic function in patients with obstructive sleep
apnea syndrome using velocity vector imaging. Medicine
2016;95:e4788.
41. D’Andrea A, Martone F, Liccardo B, et al. Acute and
Chronic Effects of Noninvasive Ventilation on Left and
Right Myocardial Function in Patients with Obstructive
Sleep Apnea Syndrome: A Speckle Tracking Echocardio-
graphic Study. Echocardiography 2016;33:1144-1155.
42. Hammerstingl C, Schueler R, Wiesen M, et al. Impact of
Untreated Obstructive Sleep Apnea on Left and Right Ven-
tricular Myocardial Function and Effects of CPAP Therapy.
PLoS One 2013;8:e76352.
43. Kim D, Shim CY, Cho YJ, et al. Continuous Positive Air-
way Pressure Therapy Restores Cardiac Mechanical Func-
tion in Patients With Severe Obstructive Sleep Apnea: A
Randomized, Sham-Controlled Study. J Am Soc Echocar-
diogr 2019;32:826-835.
44. Kovács A, Lakatos B, Tokodi M, Merkely B. Right ven-
tricular mechanical pattern in health and disease: beyond
longitudinal shortening. Heart Fail Rev 2019;24:511-520.
45. Jasaityte R, Heyde B, D’hooge J. Current State of Three-
Dimensional Myocardial Strain Estimation Using Echocar-
diography. J Am Soc Echocardiogr 2013;26:15-28.
46. Atsumi A, Seo Y, Ishizu T, et al. Right Ventricular Deforma-
tion Analyses Using a Three-Dimensional Speckle-Track-
ing Echocardiographic System Specialized for the Right
Ventricle. J Am Soc Echocardiogr 2016;29:402-411.e2.
47. Smith BC, Dobson G, Dawson D, Charalampopoulos
A, Grapsa J, Nihoyannopoulos P. Three-dimensional
speckle tracking of the right ventricle. J Am Coll Cardiol
2014;64:41-51.
48. Liu BY, Wu WC, Zeng QX, et al. Comparison of the capabil-
ity of risk stratification evaluation between two- and three-
dimensional speckle-tracking strain in pre-capillary pulmo-
nary hypertension. Pulm Circ 2019;9:2045894019894525.
49. Huntgeburth M, Germund I, Geerdink LM, Sreeram N,
Udink Ten Cate FEA. Emerging clinical applications of
strain imaging and three-dimensional echocardiography
for the assessment of ventricular function in adult con-
genital heart disease. Cardiovasc Diagn Ther 2019;9 (Suppl
2):S326-S345.
Pictorial essay Med Ultrason 2021, Vol. 23, no. 2, 220-225
DOI: 10.11152/mu-2580

First trimester fetal heart evaluation. A pictorial essay

Ioana Cristina Rotar1, Daniel Mureșan1, Claudiu Mărginean2, Dominic Gabriel Iliescu3,
Ștefania Tudorache3

11st
Department of Obstetrics and Gynecology, University of Medicine and Pharmacy “Iuliu Hațieganu” Cluj-Napoca,
2Department of Obstetrics and Gynecology, University of Medicine, Pharmacy, Sciences and Technology, Târgu Mureș,
3Department of Obstetrics and Gynecology, University of Medicine and Pharmacy, Craiova, Romania

Abstract
Classically fetal heart evaluation is mandatory in the second trimester of pregnancy. Recent data suggest that the diagnosis
of congenital heart disease is feasible in the first trimester (FT) of pregnancy, especially for trained examiners. In this pictorial
essay we aim to illustrate in detail the particularity of the FT heart evaluation: the specific ultrasound techniques including
practical tips for the basic and extended first trimester fetal heart evaluation protocol. The diagnosis is possible by the use of
a variety of ultrasound techniques (B mode, CFM, Bidirectional Doppler, STIC, TUI) using the transabdominal/transvaginal
probes. An abnormal FT heart scan requires second trimester reassessment. Unfortunately, not all cardiac disease can be po-
tentially diagnosed in the first trimester. Based on their own experience the authors recommend that heart examination in the
FT should be attempted in all fetuses, even in low risk pregnancies, using at least the basic Doppler examination protocol.
Keywords: fetal heart; congenital heart defects; ultrasound; spatio-temporal image correlation (STIC); four chamber view

Introduction Carbamazepine, Lithium, Retinoic acid, NSAID), IVF

The obstetrical ultrasound (US) had a fulminant evo-
lution in the latest decades; the number of US visualisable
structures in the first trimester (FT) of pregnancy being
in a continuous growth. Among them, the confirmation
of the normality of the fetal heart (FH) is particularly
important due to an increased prevalence of congenital
heart defects (CHD); 0.8% of the births in the European
Union, representing 30.47% of all congenital defects
[1].
Fetal cardiac evaluation should be mandatory in the
FT in high risk patients: maternal conditions (pregesta-
tional diabetes, phenylketonuria, lupus or Sjögren syn-
drome with positive SSA/SSB autoantibody, exposure to
Received 20.04.2020  Accepted 30.06.2020
Med Ultrason
2021, Vol. 23, No 2, 220-225
Corresponding author: Prof Daniel Mureșan MD, PhD
1st Department of Obstetrics and Gynecology
University of Medicine and Pharmacy
“Iuliu Hațieganu” Cluj Napoca, Romania
E-mail: muresandaniel01@yahoo.com
pregnancy, monochorionic twinning or consanguinity
[2]. Currently the most frequent referral for an extensive
cardiac examination is abnormal ultrasound findings (as-
sociated anomalies, increased nuchal translucency, ab-
normal ductus venosus flow or tricuspid regurgitation in
the FT) or chromosomal anomalies [3]. Heart evaluation
in low-risk population nowadays is not routinely per-
formed, despite the fact that 90% of CHD occurs in low
risk pregnancies [4].
FT FH evaluation can be challenging due to the small
heart dimensions, increased heart frequency, active fetal
movements, unfavorable fetal position or well-represent-
ed adipose panicle in obese patients. The second trimes-
ter (ST) ultrasound detection is the gold standard today,
having a proved efficacy and lowering the rate of perio-
perative  mortality [5]. However, an increasing number
of evidences suggested that the diagnosis of CHD is also
feasible at the end of the FT [6-9].
In this pictorial essay we present in images our ex-
perience related to the possibilities of FT US for the di-
agnostic of CHD together with a short literature review.
 Med Ultrason 2021; 23(2): 220-225 221
Techniques used for FH scanning in the first
trimester

FH examination in the FT is directly linked to the car-

diovascular embryology. A conclusive evaluation of the
heart structures is achievable at the end of the FT (11-13
gestational weeks- GW + 6 days) [7-9].
The evaluation starts in B mode with the upper ab-
dominal section, an oblique plan through the liver at the
level of the umbilical vein, stomach and posteriorly the
spine (fig 1a). By a slight cranial shift, the four-chamber
view (4CV) can be easily obtained (fig 1b). On 4CV the
following structures can be examined: heart position, Fig 1. B-mode evaluation of the heart in the first trimester:
axis and dimensions, morphology and dimensions of the a) upper abdominal transverse section; b) four-chamber view.
two atria and ventricles, ventricular and interatrial sep-
tum integrity, atrioventricular valves presence and move- zation of great vessels enhancing CHD detection rates
ment. By cranial movement and slight rotations, the right [7]. The simplified Doppler evaluation protocol includes
and left ventricular outflow tract can be obtained. two plans (basic evaluation): the 4CV and the three ves-
Color Doppler examination (color flow mode-CFM sels and trachea view (3VT) (fig 2a,b). In good scanning
and/or directional power Doppler) improves the visuali- conditions, the ductal arch and aortic arch are visible on

Fig 2. Examination of a normal fetal heart in the first trimester using the CFM transabdominal probe: a) four-chamber view – normal
diastolic flow from both right and left atrium into right and respectively left ventricle; b) “V” sign – aortic arch and pulmonary arch;
c) parasagittal view of the aortic and ductal arch

Fig 3. 3D technique for the evaluation of the fetal heart in the first trimester: a) spatiotemporal image correlation (STIC) acquisi-
tion with 3D rendered image – diastolic ventricular inflow and systolic aortic and pulmonary flow can be seen on the same image;
b) tomographic ultrasound imaging in a STIC acquisition.
222 Ioana Cristina Rotar et al First trimester fetal heart evaluation. A pictorial essay

parasagittal views (fig 2c). The visualization of systemic

venous drainage, more difficult, should be attempted in
the extending scanning protocol.
The 3D/4D techniques ameliorates heart structures
visualization. Moreover, it allows the analysis of the
cardiac cycle. The spatiotemporal image correlation soft-
ware (STIC) improves the accuracy of the evaluation of
the FH in the FT providing a better resolution (fig 3a)
[9,10]. Furthermore, the use of tomographic ultrasound
imaging (TUI) displays several cross-sectional images
in three orthogonal planes and at specific distances from
4CV (fig 3b) [10].
For screening purposes, the transabdominal (TA) ap-
proach is preferred. The transvaginal examination (TV)
is particularly useful in obese patients or in retroverted
uterus (fig 4).
The complete evaluation of FH in the FT should
include evaluation of the cardiac situs, cardiac connec-
tions, atrioventricular junction, right-left side symmetry
and septal-aortic continuity [10]. The visualization and
normality of all of the above-mentioned structures allows
an early certification of a normal heart in the FT. In some
cases, it is difficult to obtain all these plans by 2 D ultra-
sound. Instead, the simplified protocol using together 2D,
CFM and STIC with two planes (4CV, 3CV) is feasible
for the routine use in the FT [7-9].
Abnormal heart in the first trimester
Unfortunately, not all CHD can be diagnosed in the
FT of pregnancy; exceptions – evolving (eg: pulmonary
stenosis, aortic coarctation) or late onset cardiac diseases
(eg. fibroelastosis). Based on the literature [7-12] and on
our own experience, CHD in the FT can be divided in easy
detectable and difficult or impossible to detect (table I).
Fig 4. Transvaginal four-chamber view section at 12 weeks
of gestation: a) B mode- four-chamber view; b) CFM – four-
chamber view
In the following sections, we aim to exemplify that an
ultrasound diagnosis of CHD in the FT can be sustained.
In most of the cases, a FT CHD suspicion was reevalu-
ated in second trimester when the final prenatal diagnosis
was established. The earlier the diagnosis, the better the
prognosis [5,12].
Dextrocardia
In the figure 5a the stomach is located on the left
side of the abdomen. At the level of 4CV the apex of the
heart is pointing to the right side of the fetus, a condi-
tion known as dextrocardia (fig 5b). The 4CV in this case
looks completely normal. The outflow tracts have also a
normal appearance (fig 5c,d).
Ventricular septal defects (VSD) are the second most
common CHD. In figure 6 the B mode (a) revealed an
interruption, anechoic area, in the cranial part of the ven-
tricular septum. The CFM (b) shows a clear flow through
the septum during the ventricular systole, probably an in-

Table I. Classification of congenital heart defects based on first trimester detectability

Easy detectable Difficult or impossible to detect
– hypoplastic hearts (left or right – tricuspid atresia) – tetralogy of Fallot without severe pulmonary stenosis
– single ventricle – partial AVSDs
– isolated dextrocardia – aortic coarctation
– heterotaxy syndromes – interrupted aortic arch
– common arterial trunk – obstructive disease of semilunar valves (aortic and pulmonary stenosis)
– double outlet right ventricle – corrected great arteries transposition
– double inlet ventricle – ventricular septal defects
– uncorrected great arteries transpositions – venous abnormalities:
– complete atrioventricular septal defects • persistent left superior vena cava
– pulmonary atresia with intact septum • interrupted inferior vena cava
– right-sided aortic arches abnormalities • total/partial anomalous pulmonary venous return
– evolving diseases:
• tumors
• Ebstein diseasew
• fibroelastosis
• double aortic arch with the left arch dominance
 Med Ultrason 2021; 23(2): 220-225 223

Fig 5. Dextrocardia CFM mode: a) upper abdomen section – stomach on the lef, umbilical vein (blue) in the middle of the image;
b) four-chamber view – normal diastolic flow in atria and ventricles with the apex of the heart pointinig to the right; c) outflow tracts;
d) 3D rendering acquisition of the outflow tracts

let VSD. VSD in the first trimester are difficult to detect severe stenosis of the mitral and/or aortic valve or aor-
due to the small dimensions of the defect [12]. They are tic coarctation with no/little communication between the
rarely isolated, more frequently being encountered in the left ventricle and the aorta [12]. Less frequent the right
case of complex cardiac malformations.
Atrioventricular septal defect (AVSD) is a CHD de-
termined by an insufficient septation of the heart associ-
ated with anomalies of the AV valves. In figure 7 a com-
plete form of AV valve is presented: single ventricular
inflow though the unique atrioventricular valve towards
both ventricles. The genetic evaluation of fetuses with
AVSD is mandatory; an association with Down syn-
drome or trisomy 18 is frequently encountered especially
when a high nuchal translucency is detected.
An early morphologic evaluation of the ventricles in
the FT is possible. In figure 8a a ventricular dispropor-
tion can be seen in 4CV: a smaller left ventricle com-
pared to a normal right ventricle. In this case, the sec-
ond trimester follow-up revealed an unbalanced partial
atrioventricular septal defect with aortic arch hypoplasia. Fig 7. Atrioventricular septal defect: spatiotemporal image
Usually the hypoplastic left heart syndrome is due to the correlation acquisition CFM diastolic flow, in four-chamber
underdevelopment of the left ventricle secondary to the view and rendered image of the flow

Fig 8. Ventricular disproportion: a) four-chamber view CFM –

Fig 6. Ventricular septal defects (vsd): a) four-chamber view – small left ventricle, normal right ventricle; b) axial grayscale
B mode; b) four-chamber view – CFM. RV, right ventricle; LV, four-chamber view – apex and small right ventricle (RV), large
left ventricle. left ventricle (LV).
224 Ioana Cristina Rotar et al First trimester fetal heart evaluation. A pictorial essay

Fig 9. Abnormal three-vessel view in a fetus with complex con-
genital heart defects with hypoplastic ductal arch: a) B mode;
b) CFM. SVC, superior vena cava; DescAo, descending aorta;
Ao arch, aortic arch.
ventricle can be hypoplastic in the case of tricuspid atre-
sia (fig 8b) or pulmonary stenosis. In the B mode at the
4CV the right ventricle is obviously smaller than the left
ventricle.
The assessment of the great vessels is an essential
step. Anomalies of the great vessels can potentially be
diagnosed in the FT of pregnancy. In figure 9, a complex
case of CHD is demonstrated – a fetus with severe tri-
cuspid atresia and small right ventricle. In the three-ves-
sel view (3VV) only two vessels are seen: superior vena
cava and the aorta, the pulmonary artery not being visible
nor in B mode (fig 9a), nor in CFM (fig 9b). The 3VV
was essential for the diagnosis.
Aberrant right subclavian artery (ARSA) represent
an anatomic variant when the subclavian artery arises di-
rectly from the aortic arch instead of the brachiocephalic
trunk, more frequently encountered in Down syndrome
compared to euploid fetuses (fig 10) [12].
Color Doppler allows a good evaluation of the ven-
triculoarterial connections, essential for the diagnosis of
transposition of great vessel (fig 11).
The performance of ultrasound scan between
11-14 WA
The visualization rates of fetal cardiac structures are
higher towards the end of FT. Accordingly, CHD detec-
tion rates varies from 20% at 11 WG up to 92% at 13
WG [13].
In Quarello et al study, the visualization rate of 4CV
was 86% but 3VT was evaluated only in 79% of fetuses
[8]. The visualization rates of 4CV was not influenced
by the gestational age, but 3VT view was better seen

Fig 10. Aberrant right subclavian artery: a) retrotracheal route of the right subclavian artery; b) orthogonal acquisition with 3D
rendering

Fig 11. Transposition of the great vessels: a) aorta and pulmonary artery have a parallel route at the emergence from the ventricles.
There is not the normal crossing; b) 3D acquisition and rendered image shows that the artery that arise from the left ventricle has a
typical pulmonary bifurcation, confirming the diagnostic of transposition.

when the CRL was >75 mm [8]. The use of 4CV section
alone had a sensitivity of only 45.71% compared to the
4CV+3VT - 88.57% for the detection of CHD [7].
The use of transvaginal probes provides a higher
resolution particularly suitable for obese women or in
the case of a retroverted uterus. Indeed, in 58%-62% of
cases the use of a 5-9 MHz TV probe obtains a satis-
factory assessment of the FH structures, the visualization
rates increasing linearly with the gestational age [14]. In
a cross-sectional study that compared TA versus TV US,
the latter was usually superior before 14 GWs [14].
The early detection rates of major CHD depend not
only on the protocol used, but also on the examiner skills;
ranging in an unselected population between 2.3-56%
[15]. In contrast, in high-risk populations, cardiac exam-
ination reached a sensitivity of 78.5-90% with a 74.5%
in concordance with the second trimester diagnosis [15].
The use of STIC ameliorates the intra- and interob-
server agreement [9,11].
Conclusions
The capacity to achieve good quality cardiac ultra-
sound images, important for the confirmation of the nor-
mality of fetal heart in the FT, is feasible and depends
not only on ultrasound probes and software, but, more
important, on the sonographer’s experience, commitment
and awareness. New techniques of ultrasound – Doppler
and 3/4D methods – greatly improve the diagnostic capa-
bilities. The heart is a developing organ; therefore, even
after a successfully complete heart scan in the FT, the
heart should be reassessed in every trimester due the po-
tential evolution of the cardiac condition.
Conflict of interest: none
References
1. European Commission. Prevalence charts and tables. Avail-
able at: https://eu-rd-platform.jrc.ec.europa.eu/eurocat/
eurocat-data/prevalence_en. Accessed on 12.04.2020.
2. Donofrio MT, Moon-Grady AJ, Hornberger LK, et al. Di-
agnosis and treatment of fetal cardiac disease: a scientific
statement from the American Heart Association. Circula-
tion 2014;129:2183–2242.
3. AIUM Practice Parameter for the Performance of Fetal
Echocardiography. J Ultrasound Med 2020;39:E5-E16.
Med Ultrason 2021; 23(2): 220-225 225
4. Bishop KC, Kuller JA, Boyd BK, Rhee EH, Miller S, Bark-
er P. Ultrasound examination of the fetal heart. Obstet Gy-
necol Surv 2017;72:54–61.
5. Li YF, Zhou KY, Fang J, Wang C, Hua YM, Mu DZ. Ef-
ficacy of prenatal diagnosis of major  congenital heart
disease  on  perinatal  management and perioperative  mor-
tality: a meta-analysis. World J Pediatr 2016;12:298-
307.
6. García Fernández S, Arenas Ramirez J, Otero Chouza MT,
Rodriguez-Vijande Alonso B, Llaneza Coto AP. Early fe-
tal ultrasound screening for major congenital heart de-
fects without Doppler. Eur J Obstet Gynecol Reprod Biol
2019;233:93–97.
7. Wiechec M, Knafel A, Nocun A. Prenatal detection of
congenital heart defects at the 11- to 13-week scan using
a simple color Doppler protocol including the 4-cham-
ber and 3-vessel and trachea views. J Ultrasound Med
2015;34:585–594.
8. Quarello E, Lafouge A, Fries N, Salomon LJ, CFEF. Basic
heart examination: feasibility study of first-trimester sys-
tematic simplified fetal echocardiography. Ultrasound Ob-
stet Gynecol 2017;49:224–230.
9. Tudorache S, Cara M, Iliescu DG, Novac L, Cernea N. First
trimester two- and four-dimensional cardiac scan: intra-
and interobserver agreement, comparison between methods
and benefits of color Doppler technique. Ultrasound Obstet
Gynecol 2013;42:659-668. 
10. Hernandez-Andrade E, Patwardhan M, Cruz-Lemini M,
Luewan S. Early Evaluation of the Fetal Heart. Fetal Diagn
Ther 2017;42:161-173.
11. 11.Turan S, Turan OM, Ty-Torredes K, Harman CR, Bas-
chat AA. Standardization of the first-trimester fetal cardiac
examination using spatiotemporal image correlation with
tomographic ultrasound and color Doppler imaging. Ultra-
sound Obstet Gynecol 2009;33:652–656.
12. Abuhamad AZ, Chaoui R. A Practical Guide to Fetal Echo-
cardiography. Normal and abnormal hearts. Third edition.
LWW, 2015.
13. Rasiah SV, Publicover M, Ewer AK, Khan KS, Kilby
MD, Zamora J. A systematic review of the accuracy of
first-trimester ultrasound examination for detecting ma-
jor congenital heart disease. Ultrasound Obstet Gynecol
2006;28:110–116.
14. Vimpelli T, Huhtala H, Acharya G. Fetal echocardiography
during routine first-trimester screening: a feasibility study
in an unselected population. Prenat Diagn 2006;26:475-
482.
15. Clur SA, Bilardo CM. Early detection of fetal cardiac ab-
normalities: how effective is it and how should we manage
these patients? Prenat Diagn 2014;34:1235-1245.
Continuing education Med Ultrason 2021, Vol. 23, no. 2, 226-230
DOI: 10.11152/mu-2873

The utility of ultrasound in the diagnostic evaluation of the posterior

ankle joint
Wojciech Konarski, Tomasz Poboży

Department of Orthopedic Surgery, Ciechanów Hospital, Ciechanów, Poland

Abstract
Sprains are the most common injury of the ankle joint and the most common traumatic injury of the musculoskeletal
system. Ultrasound (US) examination of the posterior ankle joint is a challenge for the examiner. This paper focuses on this
difficult area and provides guidance on how to effectively perform US examination of the posterior ankle.
Keywords: ultrasound; posterior talofibular ligament; posterior tibiotalar ligament; posterior talocalcaneal ligament; flexor
hallucis longus tendon

Introduction is key to obtaining good quality images and to facilitate

accurate assessment of these structures.
Sprains are the most common ankle injuries and
one of the most common traumatic injuries of the hu- Anatomy of the posterior ankle joint
man musculoskeletal system, both related and unrelated
to sport [1-3]. Various ligaments may be damaged as a When visualizing the posterior ankle point, the pos-
result of traumatic injury to the ankle. This paper aims terior part of the tibiotarsal and subtalar joints are vis-
to explore the use of diagnostic ultrasound (US) of the ible deep within the joint, particularly the bony contours
posterior ankle joint, in particular the posterior talofibu- (fig 1).
lar ligament and the posterior part of the deltoid ligament
(the posterior tibiotalar ligament).
The anterior talofibular ligament is the structure most
often damaged in ankle injury [4,5]. Injury of the pos-
terior talofibular ligament is significantly less common
and is therefore considered of less clinical importance.
However, due to the deep location of this ligament, it is
more difficult to assess the condition both via physical
examination and via imaging, particularly US examina-
tion [6,7]. An understanding of the anatomy of this area

Received 09.10.2020  Accepted 30.12.2020

Med Ultrason
2021, Vol. 23, No 2, 226-230
Corresponding author: Wojciech Konarski, MD, PhD
Surgery Clinic, Department of Orthopaedic
Surgery, Ciechanów Hospital,
2 Powstańców Wielkopolskich Street,
06-400 Ciechanów, Poland
E-mail: wkonarski@poczta.onet.pl
Phone: +48 502 110 863
Fig 1. Anatomy of the posterior ankle joint. The posterior pro-
cess of the talus: the medial (1) and the lateral (2) tubercles.
Pink – flexor hallucis longus tendon, green – posterior talofibu-
lar ligament), blue – posterior tibiofibular ligament, red – poste-
rior talocalcaneal ligament, yellow – deltoid ligament (posterior
tibiotalar part).

Fig 2. a) US of the distal part of the Achilles tendon, using
a longitudinal scan; b) patient and probe positioning. ACH –
Achilles tendon; K – Kager fad pad; star – retrocalcaneal bursa.
The posterior process of the talus has two tubercles:
the medial and the lateral, and accurate imaging of the tu-
bercles is crucial to visualization of the ligaments of the
posterior ankle joint. The posterior talofibular ligament
attaches to the lateral tubercle, and the posterior part of
the deltoid ligament attaches to the medial tubercle. As
with most joints, the ligaments provide direct reinforce-
ment of the joint capsule and form the deepest layer of
soft tissue, located adjacent to the articular surface of the
bones forming the joint. Another important anatomical
structure, the flexor hallucis longus tendon, is located
in the groove between the two tubercles of the posterior
process of the talus, and is important to visualize in the
context of diagnostic imaging following ankle injury. Su-
perficial to the joint capsule and ligamentous structures is
the connective tissue of the Kager fad pad and more su-
perficial to this is the Achilles tendon. The retrocalcaneal
bursa is located between the distal part of the Achilles
tendon, the upper surface of the calcaneus and the con-
nective tissue of the Kager fad pad (fig 2).
Fig 4. a) Posterior talofibular ligament (PTFL), longitudinal
view and b) probe position. MT – medial tubercle of the pos-
terior process of the talus; LT – lateral tubercle of the posterior
process of the talus; star - groove between the two tubercles of
the posterior process of the talus (flexor hallucis longus tendon
space); arrows – posterior talofibular ligament (PTFL).
Med Ultrason 2021; 23(2): 226-230 227
Fig 3. US of the middle part of the Achilles tendon, using a
longitudinal probe position. Star – posterior talocalcaneal liga-
ment; arrow – posterior tibiotalar ligament.
US of the posterior ankle joint
There is uncertainty regarding the most convenient
position to perform US of the posterior ankle. The pos-
terior ankle is most accessible when the patient is laying
in a prone position with the knee slightly bent and the
ankle joint held at an angle of 90º with the weight of the
foot resting on the toes. However, in the case of recent
injury, this position may cause considerable pain, so it is
often necessary to adapt to the situation and instead ex-
amine the patient in a supine position with the knee bent
to approximately 90º and the foot resting on the floor.
This alternative position is often sufficient to adequately
evaluate the ligaments.
In the first part of the examination, the transducer is
placed along the long axis of the limb over the central
part of the Achilles tendon. One should remember at the
beginning of the examination to set an appropriate depth
and focus and to choose an appropriate frequency. A mul-
ti-frequency probe with a range of 3–12 MHz, with the
frequency set to the upper range is recommended. The
goal is to visualize the outline of the bony structures and
Fig 5. Posterior tibiotalar ligament (part of the deltoid liga-
ment), longitudinal scan; b) patient and probe positioning.
MM – malleolus mediualis; MT – medial tubercle of the pos-
terior process of the talus; LT – lateral tubercle of the posterior
process of the talus; star – flexor hallucis longus tendon; arrows
– posterior tibiotallar ligament.
228 Wojciech Konarski et al The utility of ultrasound in the diagnostic evaluation of the posterior ankle joint

Fig 6. a) Flexor hallucis longus (FHL), transverse scan; b) pa- Fig 7. a) Flexor hallucis longus (FHL), longitudinal section;
tient and probe positioning. MT – medial tubercle of the pos- b) patient and probe positioning.
terior process of the talus; LT – lateral tubercle of the posterior
process of the talus; star – flexor hallucis longus tendon.

joint margins, presence or absence of exudate, as well as

the presence and size of the os trigonum bone or hyper-
trophied posterior talus process (fig 3).
While the US transducer is positioned along the long
axis of the limb, it can be moved medially and laterally
whilst remaining in the longitudinal plane, to assess the
whole posterior ankle area. This method allows complete
visualization of the flexor hallucis longus tendon. This
tendon is often easier to initially locate using cross-sec-
tional imaging, where the tendon is visible between the
tubercles of the posterior process of the talus.
With the transducer held in the transverse plane, the Fig 8. a) Posterior talocalcaneal ligament, longitudinal scan;
b) patient and probe positioning (arrows – posterior talocalca-
posterior talofibular ligament and the posterior tibiotalar neal ligament).
ligament can be assessed. To fully assess the posterior
talofibular ligament, the transducer is placed transversely
on the posterolateral aspect of the joint where the poste-
rior surface of the distal end of the fibula and the lateral
tubercle of the posterior process of the talus are visible.
The ligament appears as the deepest layer of soft tissue,
with typical fibrous echostructure adjacent to the bone
surface (fig 4).
To assess the posterior tibiotalar ligament (the poste-
rior part of the deltoid ligament) the transducer is placed
transversely on the posteromedial aspect of the ankle,
where the posterior part of the medial ankle and the
medial tubercle of the posterior process of the talus are
Fig 9. Posterior tibiofibular ligament (posterior part of the tibi-
visible. The posterior tibiotalar ligament appears as the ofibular syndesmosis); b) patient and probe positioning (arrows
deepest layer of soft tissue located adjacent to the bone – posterior tibiofibular ligament).
surface (fig 5).
With the transducer in the transverse plane the tendon in cases of serious injuries. To evaluate this ligament the
of the flexor hallucis longus is visible in the groove be- transducer is held in the longitudinal plane, proximal to
tween the tubercles of the posterior process of the talus the superior margin of the calcaneus. The ligament ap-
(fig 6). This can be further assessed in the longitudinal pears as a fibrous tissue layer connecting the talus and
plane (fig 7). calcaneus (fig 8).
The posterior talocalcaneal ligament is not of signifi- The posterior tibiofibular ligament is located on the
cant importance as injury is rare and only tends to occur posterior surface of the tibiofibular syndesmosis (fig 9).

To visualize this ligament, the transducer is held in the
longitudinal plane slightly proximal to the tibiotalar
joint. The ligament is present as a fibrous structure be-
tween the posterior tibial process (the most prominent
part of the posterior tibia) and the fibula. The tibiofibular
syndesmosis should be evaluated from both the posterior
and anterior sides, with appropriate comparisons made to
the opposite limb.
Pathology of the posterior ankle joint
US imaging of the ankle joint is often indicated for
patients with pain persisting for 4–6 weeks following an
injury [8]. However, performing US examination sooner
following injury may have a significant impact on the
manner and timing of definitive treatment. For some pa-
tients, it may reduce the need for unnecessary immobi-
lization and, for others, it may identify damage that re-
quires timely surgical intervention. Typically, ligament
damage is assessed in terms of severity using a three-
point grading scale [9].
Grade I injuries occur when the ligament is stretched,
but remains intact. Typical symptoms are of local tender-
ness, with swelling visible on US as a symptom thickening
and reduction in echogenicity. In grade II injuries, there
is evidence of a partial rupture of the ligament. In recent
injuries and up to a few days after the injury, continuity
of the ligament may appear preserved, or may display
an obvious defect. Grade II injuries are often diagnosed
after 2–3 weeks, when scar tissue has started to form.
This appears on US imaging as thickening and reduced
echogenicity of the ligament, with the thickening being
significantly greater than in the case of grade I damage.
In grade III injuries, there is complete rupture of the
ligament. Recognition of such damage in the acute set-
ting is straightforward; however, it may be more chal-
lenging in the case of chronic injury. Following grade
III ligament damage in the ankle, a thin, inefficient scar
often forms in the area of the damage, without proper
tension of the damaged ligament, either at rest or during
functional examination.
Images of ligament damage may vary depending on
the location of the damage [8]. Damage along the length
of the ligament is easy to evaluate; however, damage to
ligament attachment points can present greater diagnos-
tic difficulty. Some ligament attachments may appear to
be hypoechogenic, but this does not necessarily indicate
damage and rather may suggest anisotropy of the tis-
sue. This often appears at the attachment of the posterior
tibial ligament to the fibula. In cases where there is di-
agnostic doubt, a comparative examination of the other
limb, provided this is normal, may be of value. In cases
Med Ultrason 2021; 23(2): 226-230 229
Fig 10. a) Posterior talofibular ligament (PTFL) and b) poste-
rior tibiotalar ligament (part of the deltoid ligament) injuries.
where there is damage to ligament attachments in the an-
kle joint, it is important to also consider the possibility of
avulsion fracture and presence of bone fragments.
Damage to ligaments of the ankle joint, as with other
joints, can lead to instability, so functional examination
should not be forgotten. The role of functional examina-
tion is particularly important in the case of damage to
ageing ankle joint ligaments. Joint instability following
ligament injury is graded from I to III depending on the
degree of joint displacement. Grade I instability is de-
fined as abnormal displacement of bony surfaces that
does not exceed 5 millimeters. Grade II instability is
defined as displacement between 5 and 10 millimeters.
Grade III involves displacement of greater than 10 mil-
limeters, more often seen in cruciate ligament or lateral
knee joint injury, but also applicable to some anterior
talofibular or calcaneofibular ligament injuries. With re-
gard to posterior ankle joint ligaments, even with com-
plete ligament rupture, it is unlikely that the bones will
be displaced more than 5 millimeters apart.
The degree of joint instability does not necessar-
ily coincide with the degree of damage to the ligament.
Grade I and most grade II ligament injuries do not result
in measurable instability. Ankle instability is usually the
result of grade III damage. The extent of instability is
also related to the manner in which it is healed, which in
turn is a result of the treatment applied. Clinical practice
has shown that US examination is reliable for assessing
grade I and grade II damage (fig 10); however, MRI ap-
pears to be more effective in the assessment of grade III
damage [10-13].
230 Wojciech Konarski et al The utility of ultrasound in the diagnostic evaluation of the posterior ankle joint
Fig 11. Flexor hallucis longus (FHL) inflammation, transverse
(a) and longitudinal (b) scan (X*X and A – edema around the
tendon and its dimension)
Post-traumatic lesions of the flexor hallucis longus
tendon are rare (fig 11). The most frequently observed
pathology is tenosynovitis (fig 12), often associated with
posterior ankle impingement syndrome, which is caused
by the presence of a large os trigonum bone. The medial
surface of the bone may adhere to the tendon sheath and
as a result of mechanical irritation, cause exudative ten-
osynovitis This condition most often occurs in athletes
such as footballers or ballet dancers and is associated
with frequent flexion of the ankle [11].
Conclusion
US is a valuable modality to assess the structures
within the posterior ankle joint. The main utility is in the
diagnostic imaging of the posterior talofibular ligament
and the posterior portion of the deltoid ligament follow-
ing ankle injury. US is also useful in the evaluation of
posterior ankle impingement syndrome caused by the
presence of a hypertrophied os trigonum bone, which can
cause an exudative flexor hallucis longus tenosynovitis.
US also provides relatively low cost diagnostic imaging
following ankle trauma in real time.
References
1. Polzer H, Kanz KG, Prall WC, et al. Diagnosis and treat-
ment of acute ankle injuries: development of an evidence-
based algorithm. Orthop Rev (Pavia) 2012;4:e5.
2. Boruta PM, Bishop JO, Braly WG, Tullos HS. Acute Lat-
eral Ankle Ligament Injuries: A Literature Review. Foot
Ankle 1990;11:107-113.
3. Fong DT, Man CY, Yung PS, Cheung SY, Chan KM. Sport-
related ankle injuries attending an accident and emergency
department. Injury 2008;39:1222-1227.
Fig 12. Flexor hallucis longus (FHL) tenosynovitis, transverse
scan (a) due to the presence of os trigonum (arrow) (b).
4. Fursdon T, Platt S. The incidence and significance of pos-
terior talofibular ligament injury on magnetic resonance im-
aging. Orthop Proc 2018;94B SUPP_XXII:56.
5. Kumai T, Takakura Y, Rufai A, Milz S, Benjamin M. The
functional anatomy of the human anterior talofibular liga-
ment in relation to ankle sprains. J Anat 2002;200:457-
465.
6. Arthurs G, Nicholls B. Ultrasound in anesthesia, critical
care and pain management. Cambridge University Press
2016.
7. Park JW, Lee SJ, Choo HJ, Kim SK, Gwak HC, Lee
SM. Ultrasonography of the ankle joint. Ultrasonography
2017;36:321-335.
8. Fessell DP, Vanderschueren GM, Jacobson JA, et al. US of
the ankle: technique, anatomy, and diagnosis of pathologic
conditions. Radiographics 1998;18:325-340.
9. George J, Jaafar Z, Hairi IR, Hussein KH. The correlation
between dynamic ultrasound evaluation and clinical laxity
grading of ATFL and CFL tears among athletes. J Sports
Med Phys Fitness 2020;60:749-757.
10. Margetic P, Salaj M, Lubina IZ. The Value of Ultrasound
in Acute Ankle Injury: Comparison With MR. Eur J Trauma
Emerg Surg 2009;35:141-146.
11. Yasui Y, Hannon CP, Hurley E, Kennedy JG. Posterior
ankle impingement syndrome: A systematic four-stage ap-
proach. World J Orthop 2016;7:657-663.
12. Bianchi S, Martinoli C, Gaignot C, De Gautard R, Meyer
JM. Ultrasound of the Ankle: Anatomy of the Tendons,
Bursae, and Ligaments. Semin Musculoskelet Radiol
2005;9:243-259.
13. Khoury V, Guillin R, Dhanju J, Cardinal É. Ultrasound of
Ankle and Foot: Overuse and Sports Injuries. Semin Mus-
culoskelet Radiol 2007;11:149-161.
Case report Med Ultrason 2021, Vol. 23, no. 2, 231-234
DOI: 10.11152/mu-2325

The role of multimodal imaging in the diagnosis of an asymptomatic

patient with congenital anomaly
Raluca Tomoaia1,2, Adrian Molnar1,2, Ruxandra Ștefana Beyer1, Alexandra Dădârlat-Pop1,2,
Florina Frîngu2,3, Diana Gurzău2,3, Gelu Simu2,3, Ioan Alexandru Minciună3, Bogdan
Caloian2,3, Dumitru Zdrenghea2,3, Dana Pop2,3

1“Niculae Stancioiu” Heart Institute, 2“Iuliu Haţieganu” University of Medicine and Pharmacy, 3Clinical Rehabilitation
Hospital, Cardiology Department, Cluj-Napoca, România

Abstract
Anomalous left coronary artery from the pulmonary artery (ALCAPA) syndrome is a rare congenital coronary anomaly,
which can cause potentially fatal complications, such as heart failure, myocardial infarction and sudden cardiac death. Only a
few patients left untreated survive to adulthood. We highlight the importance of multimodal imaging in the diagnosis of AL-
CAPA syndrome in a young asymptomatic female patient with inducible ischemia on exercise. The patient was successfully
treated with surgery.
Keywords: multimodal imaging; ALCAPA; coronary angiography; echocardiography

Introduction all associated with an increased morbidity [1]. We report

Anomalous left coronary artery from the pulmonary
artery (ALCAPA) syndrome is a rare congenital anoma-
ly, affecting 1 of 300.000 newborns. Approximately 90%
of patients die during the first year of life due to myocar-
dial infarction and heart failure. The rest of the cases (18-
25%) develop sufficient myocardial collaterals between
the anomalous left coronary artery and the normal right
coronary artery and manage to survive to adulthood. This
compensatory mechanism leads to a left to right shunt to
the pulmonary artery, with coronary steal phenomenon,
resulting in the development of myocardial ischemia,
arrhythmias and heart failure later in life. Treatment of
ALCAPA syndrome includes several surgical techniques,
Received 28.11.2019  Accepted 02.02.2020
Med Ultrason
2021, Vol. 23, No 2, 231-234
Corresponding author: Adrian Molnar MD, PhD
“Niculae Stancioiu” Heart Institute
19-21, Moților street,
400001, Cluj-Napoca, Romania
Phone: +40264597256
E-mail: adimolnar45@yahoo.com
the case of an asymptomatic 24-year-old female patient
referred to our center for cardiology evaluation before
joining the national basketball team. She was diagnosed
with ALCAPA syndrome and was managed surgically
with good outcome.
Case report
A 24-year-old previously asymptomatic female pa-
tient was referred to our cardiology department for eval-
uation before joining the national basketball team. She
denied any history of palpitations, angina or dyspnea and
had no family history of sudden cardiac death (SCD).
There were no pathological findings on clinical examina-
tion. Electrocardiogram revealed diffuse ST-T changes.
The exercise stress test showed further ST depression (up
to 5 millimeters), which was predominant in V2-V6.
Routine transthoracic echocardiography (TTE) re-
vealed multiple turbulent flows in the interventricular
septum and in the lateral left ventricular (LV) wall, both
in systole and diastole (fig 1A-D). There was also a signif-
icant thickening of the lateral LV wall, which was not di-
lated and had a preserved global systolic function (fig 1E),
232 Raluca Tomoaia et al The role of multimodal imaging in the diagnosis of an asymptomatic patient with congenital anomaly
Fig 1. TTE showing multiple turbulent flows (arrows) in the
interventricular septum from parasternal long (A) and short (B)
axis. Same findings (arrows) from apical 4 chamber view in the
interventricular septum (C) and in the lateral LV wall (D). Api-
cal 4chamber view revealing significant thickening (arrows)
of the lateral LV wall (E). Bulls-eye showing reduced regional
longitudinal strain at the level of the lateral wall (F).
but with reduced regional longitudinal strain at the level
of the lateral wall (fig 1F). Therefore, the suspicion of
multiple ventricular septal (VSD) defects was raised,
possibly associated with left ventricular noncompacta-
tion (LVNC).
Magnetic resonance imaging (MRI)was afterwards
performed, with visualization of a LV with preserved
ejection fraction and sub-endocardial late gadolinium
enhancement (LGE) at the level of the lateral and ante-
rior septal LV walls. MRI criteria for LVNC were also
positive (Petersen index of 4.4 at the level of the apex,
2.6 at the level of the lateral wall). Furthermore, there
were multiple millimetric lodges within the myocardium,
predominantly at the level of the same LV walls (fig 2).
Thus, a coronary anomaly was suspected and a coronary
computed tomography angiography (CCTA) was consid-
ered necessary as a next step.
CCTA revealed giant coronary arteries (CA) - the left
CA (LCA) measured 10.3 mm, the left anterior descend-
ing CA (LAD) 7.4 mm and the right CA (RCA) 8.8 mm.
There was an anomalous origin of the LCA from the pul-
monary artery (PA), with a normal origin of the RCA and
many myocardial collaterals between the LAD and the
RCA (fig 3).
Subsequently the patient underwent cardiac surgery
with ligation of the proximal segment of the anomalous
LCA and grafting of the left internal mammary artery into
the distal segment of the artery. Two months later, the
follow-up echocardiography revealed a significant reduc-
tion of the collateral vessels (fig 4). The patient decided
to postpone joining the basketball team at that moment.
Discussion
ALCAPA syndrome is a very rare congenital disease
and accounts for 0.25-0.5% of all congenital heart de-
fects [1]. Survival is a rare finding in the adult popula-
tion, depending on the development of inter-coronary
collaterals between the left and the right CA. The left-to-
right shunt leads to coronary steal phenomenon, with low
oxygenation of the lateral left ventricular (LV) wall as a
consequence of preferential flow of the blood to the low-
pressure PA as opposed to the myocardium. This mecha-
nism may cause chronic myocardial ischemia and myo-
cardial infarction, leading to malignant arrhythmias and
SCD [2]. Once ALCAPA is diagnosed, surgery should be
performed immediately [3].
Electrocardiography may raise the suspicion of this
anomaly in a young adult, showing ischemic changes
(most frequently negative T waves in DI and aVL) [4],
which were also present in our patient.
Even though coronary artery angiography was the
gold standard for the diagnosis of ALCAPA syndrome,
noninvasive imaging is sufficient in the modern era [3].
Echocardiography may visualize a dilated RCA, retro-
Fig 2. MRI visualizing multiple millimetric lodges within the
myocardium, predominantly at the level of the septum and lat-
eral LV wall (arrows).

Fig 3. CCTA revealing an anomalous origin of the LCA from the PA with a normal origin of the RCA(A) and many collaterals (red
arrows) between LAD and RCA (B).
grade flow from the LCA to the PA and the collaterals
with systolic and diastolic blood flow [5]. Moreover,
speckle tracking echocardiography reveals reduced lon-
gitudinal and circumferential strain in the regions corre-
sponding to the LCA [2]. CCTA allows excellent spatial
resolution to establish the origin and course of the CA [2].
MRI can be useful due to the benefit of LGE, which in-
dicates fibrosis secondary to chronic ischemia [6]. Some
of these findings were also present in our patient. There
were septal and lateral color flow signals on echocardi-
ography, with reduced regional strain at the level of the
lateral LV wall. The CCTA revealed the abnormal course
of the RCA with anastomoses between the right and left
CA. MRI demonstrated fibrosis at the level of the lateral
and anterior septal walls as a sign of chronic ischemia.
The connection between ALCAPA syndrome and
LVNC can be explained by early embryonic develop-
ment. ALCAPA is the result of abnormal septation of the
conotruncus into the aorta and pulmonary artery or due
to persistence of the pulmonary buds and involution of
Fig 4. Three-months postoperative TEE showing significant
reduction of the collateral vessels.
Med Ultrason 2021; 23(2): 231-234 233
the aortic buds. On the other hand, LVNC is caused by
the arrest of the embryogenesis of the endocardium and
myocardium, with coronary circulation being developed
simultaneously during this process, when intratrabecular
recesses are reduced to capillaries [2]. In our patient, both
echocardiography and MRI showed a thickened lateral
LV wall with positive criteria of LVNC, but with multiple
milimetric lodges inside the myocardium, demonstrating
the connection between the two processes (development
of the CA and compactation of the LV).
If untreated, ALCAPA syndrome has a high mortality
(80-90%) [2]. SCD occurs mainly in young athletes and
basketball players [1]. A literature review of 151 patients
with this pathology found that 14% were asymptomatic
and 62% of those with SCD were asymptomatic before
the diagnosis was established [7].
The particularity of our case consists in the rarity of
survival in patients with ALCAPA syndrome when left
untreated, moreover with the patient being asymptomat-
ic and physically active. Even though ALCAPA is rare
in the adult population, the diagnosis is essential since
early treatment may prevent myocardial damage. Cur-
rent guidelines indicate that such patients may return to
competitive sports 3 months postoperatively, provided
that they remain asymptomatic and an exercise stress test
does not show ischemia or important arrhythmias [3].
In conclusion we underline the importance of im-
aging in the early diagnosis of patients with ALCAPA
syndrome. Furthermore, we highlight the connection be-
tween CA anomalies and LVNC, as these two processes
coincide during early embryogenesis.
References
1. Harmon KG, Asif IM, Maleszewski JJ, et al. Incidence,
cause, and comparative frequency of sudden cardiac death
in national collegiate athletic association athletes: a decade
in review. Circulation 2015;132:10-19.
234 Raluca Tomoaia et al The role of multimodal imaging in the diagnosis of an asymptomatic patient with congenital anomaly
2. Elumalai G, Sujitha AS. “Anomalies origin of left coronary
artery” its embryological basis and clinical significance.
Elixir Embryology 2016;100:43446-43449.
3. Brothers JA, Frommelt MA, Jaquiss RDB, Myerburg RJ,
Fraser CD, Tweddell JS. Expert Consesnsus guidelines:
Anomalous aortic origin of a coronary artery. J Thorac Car-
diovasc Surg 2017;153:1440-1457.
4. Dilawar M, Ahmad Z. Anomalous left coronary artery from
pulmonary artery: Case series and brief review. OJPed
2012;2:77-81.
5. Patel SG, Frommelt MA, Frommelt PC, Kutty S, Cramer
JW. Echocardiographic diagnosis, surgical treatment,
and outcomes of anomalous left coronary artery from the
pulmonary artery. J Am Soc Echocardiogr  2017;30:896-
903.
6. Boutsikou M, Shore D, Li W, et al. Anomalous left coro-
nary artery from the pulmonary artery (ALCAPA) diag-
nosed in adulthood: Varied clinical presentation, thera-
peutic approach and outcome. Int J Cardiol 2018;261:
49-53.
7. Memon MKY, Amanullah M, Atiq M. Anomalous Left
Coronary Artery from Pulmonary Artery: An Important
Cause of Ischemic Mitral Regurgitation in Children. Cureus
2019;11:e4441.
Case report Med Ultrason 2021, Vol. 23, no. 2, 235-237
DOI: 10.11152/mu-2503

A rare cause of biliary obstruction – intraductal neuroendocrine

tumor of the right hepatic biliary duct: a case report
Mirela Danila¹, Roxana Sirli¹, Alina Popescu¹, Nicoleta Iacob², Ana-Maria Ghiuchici¹

¹Department of Gastroenterology and Hepatology, “Victor Babeș ”University of Medicine and Pharmacy, ²Neuromed
Diagnostic Imaging Center Timișoara, Romania

Abstract
Primary biliary tract neuroendocrine tumors (NETs) are extremely rare tumors that account for 0.2-2% of all gastrointes-
tinal neuroendocrine tumors. The typical presentation is with jaundice and other symptoms related to biliary obstruction.We
present a case of right hepatic duct NET in a 27-year-old female patient, asymptomatic, presented for a routine ultrasound
examination that revealed moderate dilatation of the intrahepatic biliary ducts and a 20 mm hyperechoic lesion in the right
hepatic biliary duct. Additional imaging was performed with the presumptive diagnosis of cholangiocarcinoma. After surgery,
the histopathological and immunohistochemical report was conclusive for the diagnosis of G2 well-differentiated NET.
Keywords: neuroendocrine tumor; right hepatic bile duct; biliary obstruction

Introduction sifies neuroendocrine neoplasms based on their Ki67 and

Primary biliary tract neuroendocrine tumors (NETs)
are extremely rare tumors with less than 100 reported
cases in the literature [1,2], accounting for 0.2-2% of all
gastrointestinal neuroendocrine tumors [2,3]. NET is de-
rived from embryonal neural cells called Argentaffin or
Kulchitsky-enterochromaffin cells that have the potential
of secreting serotonin. These cells are found in the high-
est proportion in the small intestine and rarely within the
biliary ducts, explaining the low incidence of extrahepat-
ic biliary neuroendocrine tumors (NET) [2,4-6].
According to the World Health Organization (WHO)
classification system (2010) NETs are defined as neo-
plasms with neuroendocrine differentiation. WHO clas-
Received 10.03.2020  Accepted 21.04.2020
Med Ultrason
2021, Vol. 23, No 2, 235-237
Corresponding author: Mirela Danilă, MD
73/9, Martir Petru Domășneanu street,
300351 Timişoara, Romania
Phone: + 40-723-480.752
Fax: + 40-256-488.003
E-mail: danila.mirela@umft.ro
Mitotic indices. In 2017, the WHO classification was up-
dated based on recent evidence [7-9].
We report a case of primary biliary NET arising from
the right hepatic biliary duct. After an extensive search of
the literature, no references regarding a NET of the right
hepatic biliary duct was found.
Case report
A 27-year-old female patient, asymptomatic, with no
known pathology, presented for the evaluation of a minor
cytolytic syndrome. Clinical examination revealed no he-
patomegaly/splenomegaly, jaundice or other abnormali-
ties. Laboratory tests – slightly increased transaminases:
AST 40 U/L (N<35), ALT 64 U/L (N<35) with negative
markers of viral B or C chronic hepatitis. Diagnostic
workup with an abdominal ultrasound revealed moder-
ate dilatation of the intrahepatic biliary ducts and a 20
mm hyperechoic lesion in the right hepatic biliary duct
(fig 1). Contrast enhanced ultrasound (CEUS) examina-
tion showed a homogenous hyperenhancement pattern in
the arterial phase with intense washout in the late phase,
suggesting malignancy (fig 2).
236 Mirela Danila et al A rare cause of biliary obstruction – intraductal neuroendocrine tumor of the right hepatic biliary duct
Fig 1. Abdominal ultrasound showing moderate dilatation of
the intrahepatic biliary ducts and a slightly hyperechoic lesion,
20 mm in size, in the right hepatic biliary duct.
Additional laboratory tests showed: mild cholestasis
AP 206 U/L (N<129), γ-glutamyl transpeptidase 286 U/L
(N<38). Tumor markers, including carcinogenic embry-
onic antigen (CEA), carbohydrate antigen 19-9 (CA19-9)
and alpha-fetoprotein, were normal.
Subsequently, abdominal MRI and MRCP confirmed
a solid lesion in the right hepatic biliary duct, with dis-
creet hyperintensity in T2 and restricted diffusion. In the
delayed phase, the lesion presented washout, suggestive
of malignancy (fig 3).
The presumptive diagnosis was biliary obstruction
due to localized Klatskin/cholangiocarcinoma tumor
without evidence of metastasis.
The patient was referred to the Department of Sur-
gical Oncology and resection of the common bile duct,
cholecystectomy, Roux-en-y hepatico-jejunostomy re-
construction and locoregional lymphadenectomy was
performed. The histopathological and immunohisto-
chemical report of the resected tumor was conclusive
for the diagnosis of a well-differentiated neuroendo-
crine tumor (NET G2). Immunohistochemical staining
Fig 2. CEUS examination: a) homogenous hyperenhancement pattern in the arterial phase; b) mild washout starting in the portal
vascular phase; c) obvious washout in the late vascular phase.
showed that the tumor cells were strongly positive for
Chromogranin A and Synaptophysin with the Ki-67 in-
dex 3-5%.
After surgery, the patient received chemotherapy and
has remained under oncologic surveillance. There was no
recurrence of the disease at three years after surgery.
Discussion
Neuroendocrine tumors derive from Argentaffin or
Kulchitsky-enterochromaffin cells. The most common
sites of NET occurrence are ileum, appendix, bladder,
prostate, rectum, stomach, bronchi, pancreas, and biliary
tree. The paucity of enterochromaffin cells in the biliary
tree explains the rare incidence of NETs in the biliary
system. Biliary tract NETs are extremely rare and only
one-fifth of these tumors are well-differentiated NETs
[5,10]. Biliary tract NETs are often non-secreting tumors
and larger tumors over 2 cm are associated with aggres-
sive behavior [5].
Published data show that in most cases the presenting
symptom was related to the local invasion of the tumor
[10]. The most common presenting symptom in patients
with biliary tract NETs is jaundice, followed by pain and
other nonspecific symptoms (pruritus, nausea, vomiting,
weight loss) [1,11]. Our patient was asymptomatic prob-
ably due to the small dimensions of the tumor.
Cholangiocarcinoma is the most frequent biliary tract
malignancy. Differential diagnosis between cholangio-
carcinoma and other bile duct tumors, such as NET, is
challenging before surgical resection and histopathologi-
cal exam. In our case, the final diagnosis was also estab-
lished postoperatively.
NETs of the biliary ducts are generally well-differen-
tiated, slow-growing tumors with rare metastatic spread
and have a favorable prognosis with a high-rate survival,
mostly in cases in which curative surgical resection is
possible [11-15]. Our patient also had a well-differentiat-
ed biliary tract NET- G2.

Fig 3. T1- Gadolinium postcontrast sequences, late phase: cor-
onal (a) and axial plane (b).
The literature search did not reveal other cases of
NET isolated in the right hepatic biliary duct, possibly
making our case the first reported case of an intraductal
NET of the right hepatic biliary duct.
In conclusion, biliary tract NETs have as a typical
presentation, jaundice and other symptoms related to bil-
iary obstruction, but patients can also be asymptomatic.
Due to the absence of specific symptoms, a correct pre-
operative diagnosis is rare: diagnosis is usually made
postoperatively, based on the histopathology exam. Radi-
cal tumor surgery is the only available curative approach,
with high survival rates.
References
1. Khan FA, Stevens-Chase A, Chaudhry R, Hashmi A, Edel-
man D, Weaver D. Extrahepatic biliary obstruction second-
ary to neuroendocrine tumor of the common hepatic duct.
Int J Surg Case Rep 2017;30:46-49.
2. Modlin IM, Sandor A. An analysis of 8305 cases of Carci-
noid tumors. Cancer 1997;79:813-829.
3. Lauffer JM., Zhang T, Modlin IM. Review article: current
status of gastrointestinal carcinoids.  Aliment Pharmacol
Therap 1999;13:271-287.
4. Barron-Rodriguez LP, Manivel JC, Mendez-Sanchez N,
Jessurun J. Carcinoid tumors of the common bile duct:
Med Ultrason 2021; 23(2): 235-237 237
evidence for its origin in metaplastic endocrine cells. Am J
Gastroenterol1991;86:1073-1076.
5. Noronha YS, Raza AS. Well-differentiated neuroendocrine
(carcinoid) tumors of the extrahepatic biliary ducts. Arch
Pathol Lab Med 2010;134:1075-1079.
6. Kuwabara H, Uda H. Small cell carcinoma of the gall-
bladder with intestinalmetaplastic epithelium. Pathol Int
1998;48:303-306.
7. Bosman FT, Carneiro F, Hruban RH, Theise ND. WHO
Classification of Tumours of the Digestive System. 4th ed.
Lyon: IARC, 2010.
8. Lloyd RV, Osamura RY, Klöppel G, Rosai J. WHO Clas-
sification of Tumours of Endocrine Organs. 4th ed. Lyon:
IARC, 2017
9. UedaY, Toyama H, Fukumoto T, Ku Y. Prognosis of Pa-
tients with Neuroendocrine Neoplasms of the Pancreas ac-
cording to the World Health Organization 2017 Classifica-
tion. JOP 2017;18 S(3):366-370.
10. Michalopoulos N, Papavramidis TS, Karayannopoulou G,
Pliakos I, Papavramidis ST, Kanellos I. Neuroendocrine
tumors of extrahepatic biliary tract. Pathol Oncol Res
2014;20:765-775.
11. Bhandarwar AH, Shaikh TA, Borisa AD, Palep JH, Patil
AS, Manke AA. Primary neuroendocrine tumor of the left
hepatic duct: a case report with review of the literature.
Case Rep Surg 2012;2012:786432.
12. Nesi G, Lombardi A, Baltignani G, Ficari F, Ru-
bio CA, Tonelli F. Well-differentiated endocrine tu-
mor of the distal common bile duct: a case study
and literature review. Virchows Arch 2006;449:104-
111.
13. Oskuie AE, Valizadeh N. Carcinoid tumor of the common
bile duct misdiagnosed as cholangiocarcinoma. Middle
East J Cancer 2011;2:139-142.
14. Costin AI, Păun I, Păun M, Constantin VD, Vârcuş F. Pri-
mary neuroendocrine tumors–an extremely rare causeof
obstruction of extrahepatic bile ducts: a case report. Rom J
Morphol Embryol 2017;58:641-644.
15. Choi J, Lee KJ, Kim SH, Cho MY. Preoperative diagnosis
of well-differentiated neuroendocrine tumor in common
hepatic duct by brush cytology: A case report. Diagn Cyto-
pathol 2019;47:720-724.
Letter to the Editor Med Ultrason 2021, Vol. 23, no. 2, 238-247

Dynamic air bronchogram and lung hepatization:

ultrasound for early diagnosis of pneumonia

Koya Akutagawa

General Medical Education Center, Kumamoto University Hospital, Kumamoto, Japan

To the Editor,

An 81-year-old woman was admitted to the intensive

care unit for septic shock. Crystalloid resuscitation was
initiated, followed by treatment with broad-spectrum an-
tibiotics immediately after sputum and blood cultures.
As the patient was hemodynamically unstable, she was
intubated and put on mechanical ventilation. On the
first day of admission, auscultation revealed faint coarse
crackles in the right lung. Chest radiography showed
bilateral pleural effusion. Blood test results revealed a
white blood cell (WBC) count of 9.2×103/µL (neutro-
phils, 96.2%) and marked elevation of C-reactive pro-
tein (CRP) levels at 26.3 mg/dL; other results, including
urinary tests, were within normal ranges. The sputum
Gram stain was negative. On the second day of admis-
sion, bedside lung ultrasound revealed bilateral pleural Fig 1. Chest radiographs and ultrasound images. Lung ultra-
sound revealed lung hepatization and dynamic air bronchogram
effusion; right lung “hepatization,” liver-like echogenic- (see video 1, on the journal site).
ity of the consolidated lung (fig 1a) and “dynamic air
bronchogram” (hyperechoic bubbles - sputum in the on chest radiography, and lung aeration improved on ul-
bronchus - moving synchronously with respiration) in the trasound, excluding the right dorsal lower lobe (fig 1b).
right lung. The latter two features strongly suggest pneu- Klebsiella pneumoniae at >100,000 colonies/mL were
monia [1,2]. On a new chest radiography, an increased obtained from sputum culture sampled on the first day
density of the right lung parenchyma reflecting pneumo- of admission. WBC and CRP levels decreased stead-
nia was found. Further blood tests revealed an elevated ily over a few days and CT scan showed no spread of
WBC count (15.7×103/µL) and CRP levels (29.2 mg/dL). infection from the right lung and no other infectious
Positive end-expiratory pressure was raised from 5 cm focus.
H2O to 10 cm H2O, and postural drainage and high-fre- CT is the gold standard for diagnosing pneumonia.
quency chest wall oscillation were initiated. On the third However, it is sometimes difficult to transfer an unsta-
day of admission, right lung density decreased slightly ble patient to the radiology department. Although chest
radiography is the first choice for patients with suspected
Received 02.03.2021  Accepted 18.04.2021 pneumonia, a systematic review showed that chest radi-
Med Ultrason
2021, Vol. 23, No 2, 238-239, DOI: 10.11152/mu-3136,
ography has 54% sensitivity in the diagnosis of pneumo-
Corresponding author: Koya Akutagawa nia compared with 93% for lung ultrasound [3]. Addi-
General Medical Education Center, tionally, CT scans may fail to differentiate pneumonia
Kumamoto University Hospital, from atelectasis, unlike dynamic air bronchogram on re-
1-1-1 Honjo, Chuo Ward, Kumamoto,
860-8556, Japan
al-time ultrasound [2]. Lung ultrasound may enable early
Phone: +8196344211 identification and timely intervention in pneumonia as in
E-mail: dirtyrivermed@gmail.com this case.

References
1. Lichtenstein DA, Lascols N, Mezière G, Gepner A. Ultra-
sound diagnosis of alveolar consolidation in the critically
ill. Intensive Care Med 2004;30:276-281.
2. Lichtenstein D, Mezière G, Seitz J. The dynamic
air bronchogram. A lung ultrasound sign of alveolar
Med Ultrason 2021; 23(2): 238-247 239
consolidation ruling out atelectasis. Chest 2009;135:1421-
1425.
3. Ye X, Xiao H, Chen B, Zhang S. Accuracy of Lung Ul-
trasonography versus Chest Radiography for the Diagnosis
of Adult Community-Acquired Pneumonia: Review of the
Literature and Meta-Analysis. PLoS One 2015;10:e0130
066.

Ultrasound imaging and guided hydro-dissection for injury of the

recurrent motor branch of the median nerve

Ke-Vin Chang1,2, Wei-Ting Wu2, Yi-Chiang Yang3, Levent Özçakar4

1Department of Physical Medicine and Rehabilitation, National Taiwan University Hospital, Bei-Hu Branch and
National Taiwan University College of Medicine, Taipei, Taiwan, 2Center for Regional Anesthesia and Pain Medicine,
Wang-Fang Hospital, Taipei, Medical University, Taipei, Taiwan, 3Physical Medicine and Rehabilitation, Taipei
Veterans General Hospital, Taipei, Taiwan, 4Department of Physical and Rehabilitation Medicine, Hacettepe University
Medical School, Ankara, Turkey

To the Editor

A 30-year-old man had right thumb pain - irrespon-

sive to oral medication and physical therapy - for the
last six months. He had visited a pain physician who had
performed landmark-guided trigger point injections over
the base of his right thumb. Unfortunately, he had been
feeling weakness during thumb opposition since then.
Electrophysiological studies were general. Ultrasound
(US) showed a normal cross-sectional area (<10 mm2)
[1] of his right median nerve at the carpal tunnel inlet.
The transducer was then relocated to the thenar emi-
nence. Compared with the asymptomatic hand, the the-
nar muscle at the painful side appeared thinner and more Fig 1. Ultrasound imaging of the recurrent motor branch (black
and white arrowheads) of the median nerve at the normal (A)
echogenic (fig 1A, B). The recurrent motor branch of and symptomatic (B) sites; ultrasound-guided perineural hy-
the median nerve (RBMN) appeared swollen compared dro-dissection for the swollen nerve (C); schematic drawing
with the contra-lateral side. Under the impression of of the recurrent motor branch (yellow dashed line) of the me-
dian nerve and the position of the transducer (red square) (D).
Arrow, needle; white asterisks, flexor pollicis longus tendon;
Received 31.03.2021  Accepted 18.04.2021 yellow asterisks, main trunk of the median nerve; APB, abduc-
Med Ultrason tor pollicis brevis; OP, opponens pollicis; FPB, flexor pollicis
2021, Vol. 23, No 2, 239-240, DOI: 10.11152/mu-3183, brevis.
Corresponding author: Ke-Vin Chang, MD, PhD
Department of Physical Medicine and thenar muscle atrophy secondary to RBMN neuropathy,
Rehabilitation, National Taiwan University
Hospital, Bei-Hu Branch and National Taiwan
US-guided hydro-dissection with 2 mL 5% dextrose was
University College of Medicine, Taipei, Taiwan performed (fig 1C, Video, on the journal site) four times
E-mail: kvchang011@gmail.com and the patient experienced a gradual recovery of thenar
240 Yi-Hsiang Chiu et al
muscle strength and more than 50% decrease in thumb
pain.
The RBMN provides innervation to the thenar mus-
cles [2]. It mostly emerges from the main trunk of the me-
dian nerve distal to the carpal tunnel outlet and ascends
vertically to the palmar surface through the anterior edge
of the transverse carpal ligament [3]. The RBMN further
curves backward to pierce and innervate the thenar mus-
cles (fig 1D). The RBMN neuropathy is uncommon and
can result from neurogenic tumours, compression of the
thenar muscles during long-distance cycling and iatro-
genic injury. In this case, we speculated that the anteced-
ent trigger point injection might be the culprit of RBMN
neuropathy. Moreover, the subsequent thenar fibrosis
could have led to entrapment of the nerve across its pas-
sage through the palmar fascia. Therefore, the US-guided
perineural hydro-dissection relieved the focal compres-
sion and facilitated the neural recovery [4,5].
Acknowledgement: The current research project was
supported by National Taiwan University Hospital, Bei-
Hu Branch, Ministry of Science and Technology (MOST
Utilization of diagnostic ultrasound in the detection of hip fracture
Yi-Hsiang Chiu, Shaw-Gang Shyu
Department of Physical Medicine and Rehabilitation, National Taiwan University Hospital and National Taiwan
University College of Medicine, Taipei, Taiwan
To the Editor,
A 72-year-old woman with history of rectal ad-
enocarcinoma and brain metastasis (but no documented
bone metastasis) had in the recent medical history left
proximal femoral vein thrombosis treated with an anti-
coagulation agent. One month before visiting the clinic
she had sudden onset of a severe right hip pain when she
Received 30.03.2021  Accepted 18.04.2021
Med Ultrason
2021, Vol. 23, No 2, 240-241, DOI: 10.11152/mu-3180,
Corresponding author: Dr. Shaw-Gang Shyu
Department of Physical Medicine and
Rehabilitation, National Taiwan University
Hospital, 7 Zhongshan South Road,
Zhongzheng District, 100 Taipei City, Taiwan
Phone: 886-2-23123456-67587
E-mail: taotaohughs@gmail.com
Utilization of diagnostic ultrasound in the detection of hip fracture
106-2314-B-002-180-MY3 and 109-2314-B-002-114-
MY3), and Taiwan Society of Ultrasound in Medicine.
References
1. Chen IJ, Chang KV, Lou YM, Wu WT, Özçakar L. Can ul-
trasound imaging be used for the diagnosis of carpal tunnel
syndrome in diabetic patients? A systemic review and net-
work meta-analysis. J Neurol 2020;267:1887-1895.
2. Riegler G, Pivec C, Platzgummer H, et al. High-resolution
ultrasound visualization of the recurrent motor branch of
the median nerve: normal and first pathological findings.
Eur Radiol 2017;27:2941-2949.
3. Smith J, Barnes DE, Barnes KJ, et al. Sonographic visual-
ization of thenar motor branch of the median nerve: A ca-
daveric validation study. PM R 2017;9:159-169.
4. Lin CP, Chang KV, Huang YK, Wu WT, Özçakar L. Regen-
erative injections including 5% dextrose and platelet-rich
plasma for the treatment of carpal tunnel syndrome: A sys-
tematic review and network meta-analysis. Pharmaceuti-
cals (Basel) 2020;13:49.
5. Chang KV, Wu WT, Özçakar L. Ultrasound imaging and
guidance in peripheral nerve entrapment: hydrodissection
highlighted. Pain Manag 2020;10:97-106.
took a walk. She reported no falling episode at the time.
Initially, she could walk with a walker but, during the
following days, her hip pain progressed and finally she
became wheelchair-bound.
On examination, she was intolerant to both the pas-
sive and active range of motion test due to severe right
hip pain. The pain worsened on weight bearing and im-
proved when resting supine. Ultrasound evaluation dem-
onstrated discontinued cortex of the right femoral neck
and a hypervascular hyperechoic amorphous soft tissue
was noted just over the cortical cleft (fig 1a, b). These
findings were consistent with right hip fracture at the
femoral neck with callus formation. The hip plain film
disclosed displaced hip fracture of the right femoral neck,
Garden type IV (fig 1c). Therefore, she underwent bipo-
lar hemiarthroplasty of the right hip. Pathological exam
disclosed osteoporosis with marrow atrophy and no evi-
dence of malignancy.

Fig 1. Coronal view of the right femoral ultrasound shows dis-
continued bony cortex with penetrating ultrasound beam (white
arrow) at the right femoral neck. A hyperechoic amorphous soft
tissue (black arrow) is also noted; b) Doppler mode showed
hypervascularity of the amorphous soft tissue; c) the right
hip x-ray reveals right femoral neck fracture, Garden type IV
(white arrowhead).
Hip fracture is a debilitating health issue that usually
results in a decreased quality of life and marked mobil-
ity and mortality. It deserves special attention when the
patient has disabling hip pain or the problem of weight
bearing. Although the patient usually has a functional de-
cline after a fracture reduction and fixation operation, the
When meniscus ‘tears’ make the Baker’s cyst ‘cry’: a story on knee
ultrasound
Carmelo Pirri1, Carla Stecco2, Nina Pirri3, Raffaele De Caro2, Levent Özçakar4
1Department of Neurosciences, Institute of Human Anatomy, University of Padova, Padova, Italy, 2School of Medicine
and Surgery, University of Messina, Messina, Italy, 3Department of Physical and Rehabilitation Medicine, Hacettepe
University Medical School, Ankara, Turkey
Received 28.12.2020  Accepted 03.01.2021
Med Ultrason
2021, Vol. 23, No 2, 241-242, DOI: 10.11152/mu-3193,
Corresponding author: Carmelo Pirri
Department of Neurosciences, Institute of
Human Anatomy, University of Padova,
Via Gabelli 67, 35121, Padova, Italy
E-mail: carmelop87@hotmail.it
Med Ultrason 2021; 23(2): 238-247 241
overall mortality rate of conservative treatment is four
times as high in one year [1]. Decision of the surgical
method depends on the likelihood of blood supply resto-
ration. Internal fixation is indicated in the femoral neck
fracture with age less than 60 or non-displaced fracture,
while arthroplasty is indicated in displaced fracture for
those aged more than 60 [2].
Ultrasound is a convenient image modality that al-
lows early diagnosis of the hip fracture, especially when
occult fracture is undetectable in the x-ray. A previous
study disclosed that compared with magnetic resonance
imaging, sensitivity and specificity of ultrasound for hip
fracture was 100% and 65% respectively [3]. Classical
ultrasound findings include fracture line with cortical dis-
continuity, peritrochanteric hypoechoic hematoma or flu-
id collection, soft tissue swelling and callus formation [3].
This case highlights the importance of the detection
of a potential hip fracture in a patient with negative high
energy trauma or fall history. In view of the high cost-
effectiveness and accessibility, we suggest the physician
should become familiar with the clinical picture and ul-
trasound findings of the hip fracture.
References
1. Tay E. Hip fractures in the elderly: operative versus non-
operative management. Singapore Med J 2016;57:178-181.
2. Bhandari M, Swiontkowski M. Management of acute hip
fracture. N Engl J Med 2017;377:2053-2062.
3. Safran O, Goldman V, Applbaum Y, et al., Posttraumatic
painful hip: sonography as a screening test for occult hip
fractures. J Ultrasound Med 2009;28:1447-1452.
To the Editor,
A 60-year-old female patient was seen because of
pain and functional limitation in the right knee for the last
two months. The pain was worse at night and when run-
ning. She declared that she had medial meniscus and an-
terior cruciate ligament tears for about five years and that
242 Carmelo Pirri et al When meniscus ‘tears’ make the Baker’s cyst ‘cry’: a story on knee ultrasound

previous physiotherapy had been only partially effective.

The medical history was otherwise noncontributory. Her
physical examination revealed painful and limited right
knee movements especially during flexion. There was no
joint swelling and instability tests were negative. Ultra-
sound (US) examination of the knee was performed in
accordance with the EURO-MUSCULUS/USPRM basic
scanning protocol [1]. The meniscal tear as well as the
multilobular Baker’s cyst (area: 2.94 cm2) were clearly
visualized (fig 1) whereby the latter also was tender to
sono-palpation (Video 1, on the journal site). Power Dop-
pler imaging was unremarkable. While the patient lay in a
prone position, dynamic/transverse US imaging was fur-
ther carried out during knee flexion i.e., in order to better
understand the movement and path of the fluid inside the
cyst/knee (Video 2, on the journal site). Based on the US
findings, three sessions of manual therapy (Fascial Ma-
nipulation®) and physical therapy were prescribed - for
maintaining knee flexibility and reducing fluid collection
and pain alike.
Baker’s cysts are commonly found in association
with intra-articular knee disorders, such as osteoarthri-
tis, meniscus or cruciate ligament tears, chondral lesions
and inflammatory arthritis [2]. Based on cadaveric stud-
ies, a valvular opening of the posterior capsule, on the
medial side and deep to the medial head of the gastroc-
nemius, is present in up to 40% to 54% of healthy adult
knees [3]. This valvular opening allows flow during knee
flexion, but (due to the tension between the semimem-
branosus muscle and the medial head of gastrocnemius
muscle) it is compressed/closed during knee extension.
In this sense, it is noteworthy that sono-palpation and the
dynamic US imaging of the cysts should be evaluated
similar to the clinical examination [4] i.e., to uncover the
Fig 1. (A) Sono-inspection of the multi-lobulated Baker’s cyst
(area: 2.94 cm2); (B) Sono-inspection of the medial meniscal
(m) tear (arrowhead). GM, gastrocnemius medialis muscle;
*, osteophytes
exact pain generator as well as to confirm the fluid path
during normal (daily life) movements of the joint. In this
way not only can a prompt diagnosis of the patient be
established (with better understanding of the symptoma-
tology and tissue/fluid biomechanics), but also targeted
interventions can be planned accordingly.
References
1. Özçakar L, Kara M, Chang KV, et al. EURO-MUSCULUS/
USPRM. Basic scanning protocols for knee. Eur J Phys Re-
habil Med 2015;51:641-646.
2. Serrano S, Ferreira JB, Özçakar L. When “sono-palpation”
becomes “sono-explosion”: The Baker’s cyst report. Am J
Phys Med Rehabil 2020;99:e125.
3. Frush TJ, Noyes FR. Baker’s Cyst: Diagnostic and surgical
considerations. Sports Health 2015;7:359-365.
4. Pirri C, Stecco C, Fede C, De Caro R, Özçakar L. Dy-
namic ultrasound examination of the paratenon and fascia
in chronic achilles tendinopathy. Am J Phys Med Rehabil
2021;100:e75. 
 Med Ultrason 2021; 23(2): 238-247 243

Controversies in the management of bowel obstruction in pregnant

woman

Florina Popa1, Pierre Bernard2, Elia Georgescu3

1Department of General Surgery, “Dunarea de Jos” University of Galati, Faculty of Medicine and Pharmacy, Galati,
Romania, 2Service de Chirurgie Digestive, Unité 23, Chirurgie viscérale, Centre Hospitalier de Mâcon, France,
3Department of Morphological and Functional Sciences, “Dunarea de Jos” University of Galati, Faculty of Medicine

and Pharmacy, Galati, Romania

To the Editor,

A 44-year-old female patient with a 13 week pregnan-

cy in evolution, presented to the emergency department
with severe and persistent epigastric pain associated
with gastrointestinal reflux. Surgical history mentioned
a pelvic endometriosis operated in 2010 and rectal resec-
tion and caesarean in 2014. The patient received in vitro
fertilization in October 2019. An abdominal ultrasound
was performed showing dilation of the colon, stercoral
stasis and free fluid in the pocket of Morisson (fig 1). The
computed tomography (CT) scan confirmed the bowel
obstruction. Laparoscopic approach was initiated and
conversion to laparotomy was done due to the high risk
of small bowel perforation. The postoperative evolution
was complicated by pregnancy loss, urinary retention,
paralytic ileus and abdominal wall abscess. After several
months of medical care, the outcome was clinically fa-
vorable and the patient recovered.
Even if there is no clear-cut protocol of investigations
for pregnant women [1], intestinal ultrasound is recom-
mended to be the first-line imaging modality in preg-
nancy for the assessment of luminal disease activity in
both colonic and small bowel diseases [2]. As in our case,
the first imaging line she had was abdominal ultrasound,
followed by a CT scan. The hospital was not equipped
with MRI at that time. After long discussions with the
gynecologist a decision was made to consent the patient
Received 22.03.2021  Accepted 25.04.2021
Med Ultrason
2021, Vol. 23, No 2, 243, DOI: 10.11152/mu-3165,
Corresponding author: Florina Popa
Department of General Surgery,
“Dunarea de Jos” University of Galati,
Faculty of Medicine and Pharmacy,
Galati, Romania
E-mail: Florina.Popa@ugal.ro
Fig 1. Abdominal ultrasound showing dilation of the colon,
with presence of stercoral stasis  and  free fluid  in  Morisson
space in longitudinal (a) and transvers scan (b)
for surgery. Sometimes current practice is different from
guidelines and controversies remain. We mention some
of the fetal mortality risk factors that were present in our
case: medical and surgical past history of the patient,
stress triggered due to multiple medical maneuvers, ex-
posure to radiation imaging, the risk of general anesthe-
sia, the risk factors of the surgery involving the manipu-
lation of the uterus.
Some of these factors could explain the complica-
tions occurred postoperatively. Bowel obstruction in
pregnant women is challenging to diagnose and difficult
to manage and, in spite of multiple treatment modalities
the patient was submitted, with the intention of reducing
the mortality risk factors, she had a bad outcome losing
her pregnancy.
References
1. Mukherjee R, Samanta S. Surgical emergencies in pregnan-
cy in the era of modern diagnostics and treatment. Taiwan J
Obstet Gynecol 2019;58:177-182.
2. Flanagan E, Bell S. Abdominal Imaging in pregnancy (ma-
ternal and foetal risks). Best Pract Res Clin Gastroenterol
2020;44-45:101664.
244 Jiangfeng Wu, Xiaoshan Hu et al An interesting dynamic ultrasound  finding of pharyngoesophageal diverticulum

An interesting dynamic ultrasound  finding of pharyngoesophageal

diverticulum: technical advice
Jiangfeng Wu1*, Xiaoshan Hu2*, Xiaoyun Wang3
* the authors share the first authorship

1Department of Ultrasound, 2Department of Radiology, 3Department of Nephrology, The Affiliated Dongyang Hospi-
tal of Wenzhou Medical University, Dongyang, Zhejiang, China

Fig 1. The lesion is abutting posterior aspect of the left lobe of the thyroid and seems connected with the esophagus (arrow); b) No
apparent blood flow signal in or around the lesion; c) Ultrasound imaging reveals a gas-filled lumen projecting to the left lobe of the
thyroid; d) CT reveals a low-density lesion with a clear contour and partial gas-fluid filled (arrow); e) Barium swallow demonstrates
a barium-filled pouch projecting from the esophagus.

To the Editor, thermore, a 2×1.9 cm hypoechoic, heterogeneous lesion,

A 60-year-old male with mild dysphagia and nausea
for three months was admitted to our hospital. His physi-
cal examination was within normal limits. Esophagogas-
troscopy was carried out and chronic superficial gastritis
was found.
Neck ultrasound was performed. Several heterog-
enous hypoechoic nodules with clear boundary, oval in
shape in bilateral lobes of the thyroid were found. Fur-
Received 02.04.2021  Accepted 18.04.2021
Med Ultrason
2021, Vol. 23, No 2, 244-245, DOI: 10.11152/mu-3187,
Corresponding author: Jiangfeng Wu
Department of Ultrasound, The Affiliated
Dongyang Hospital of Wenzhou Medical
University, 60 Wuning West Road,
Dongyang 322100, Zhejiang, China
E-mail: wjfhospital@163.com
Phone: 18257937213
with clear boundary, internal strong echo was identified
abutting the posterior upper aspect of the left lobe of the
thyroid gland. At first, we thought it was a thyroid nod-
ule with internal calcification. After careful scanning, we
found that the lesion seemed connected with the esopha-
gus (fig 1a). There was no apparent blood flow signal in
or around the lesion (fig 1b). The patient was asked to
swallow, but no significant changes in the lesion were
observed. Likewise, the shape and internal echo of the
lesion remained unchanged when compressed with the
probe. The patient was asked to drink water, to detect
whether fluid entered the lesion, and ultrasound imaging
revealed a gas-filled lumen projecting to the left lobe of
the thyroid (fig 1c).
A pharyngoesophageal diverticulum was suspected.
Computed tomography revealed a 2.3×1.6 cm low-
density lesion with a clear contour and partial gas-fluid
filled in the posterior aspect of the left lobe of the thy-

roid (fig 1d). Finally, the diagnosis was confirmed by an
esophageal barium swallow, which demonstrated a bari-
um-filled pouch projecting from the esophagus and abut-
ting posterior upper aspect of the left lobe of the thyroid
(fig 1e).
Pharyngoesophageal diverticulum is a relatively rare
disorder with the characters of a sac-like pouch or a dilat-
ed lumen resulting from the esophagus. The diagnosis of
pharyngoesophageal diverticulum has mostly been made
on the basis of a barium swallow test or endoscopy [1,2].
In our case, there were no obvious changes found in the
lesion when compressed with the probe or when the pa-
tient was asked to perform the act of swallowing, which
Comment to: Effectiveness of contrast-enhanced ultrasound
for detecting the staging and grading of bladder cancer:
a systematic review and meta-analysis
Jiangfeng Wu1*, Xiaoshan Hu2*, Xiaoyun Wang3
* the authors share the first authorship
1Department of Ultrasound, 2Department of Radiology, 3Department of Nephrology, The Affiliated Dongyang Hospi-
tal of Wenzhou Medical University, Zhejiang, China
To the Editor,
We read with great interest the manuscript of Ge et al
[1]. We strongly agree with the authors about the impor-
tance of contrast-enhanced ultrasound in the diagnosis of
bladder cancer, but we would like to pay attention to sev-
eral important missing aspects in the article.
We reveal some different results regarding the quality
assessment of the included studies in the article [1]. Con-
cerning the reference standard domain, 3 studies [2-4]
should be considered as an unknown risk of bias because
Received 10.03.2021  Accepted 18.04.2021
Med Ultrason
2021, Vol. 23, No 2, 245-246, DOI: 10.11152/mu-3146,
Corresponding author: Jiangfeng Wu
Department of Ultrasound, The Affiliated
Dongyang Hospital of Wenzhou Medical
University, 60 Wuning West Road,
Dongyang 322100, Zhejiang, China
E-mail: wjfhospital@163.com
Phone: 18257937213
Med Ultrason 2021; 23(2): 238-247 245
might be due to the narrow junction between the esoph-
agus and the diverticulum. So, we advise that drinking
water may be helpful in differentiating pharyngoesopha-
geal diverticulum from other lesions when no obvious
changes are found after swallowing or compression.
References
1. Shao Y, Zhou P, Zhao Y. Ultrasonographic findings of phar-
yngoesophageal diverticulum: two case reports and review
of literature. J Med Ultrason (2001) 2015;42:553-557. 
2. Chen HC, Chang KM, Su WK. Incidental pharyngoesopha-
geal diverticulum mistaken for a thyroid nodule: Report of
two cases. Diagn Cytopathol 2019;47:503-506.
the blinded status of the index test was not definitely
reported. Furthermore, concerning the flow and timing
domain, one study [4] should be considered as having an
unknown risk as it did not definitely report the interval
time between the index test and the reference standard.
The authors found that, after the test, the positive
likelihood ratio increased from 20% to 70%, while the
negative likelihood ratio decreased to 2%. We think the
interpretation is not appropriate. The rational interpreta-
tion is that patients with a probability of 20% of develop-
ing the disease show a probability of 70% of the disease
when a positive result of the test, while a probability of
2% of the disease when it is a negative result.
Ge et al [1] revealed that the pooled sensitivity
(I2=62.02%, p=0.03>0.01) and specificity (I2=45.69%,
p=0.12>0.01) indicated no significant heterogeneity. But
as described in the statistical analysis that if the outcome
was I2>0.5 or p<0.1 a random-effects model was selected
to indicated the heterogeneity’s result. Therefore, the
p>0.01 indicating no significant heterogeneity was in-
246 Xin-Yue Ge et al Author’s response

consistent with that described in the statistical analysis.

So, there is a moderate heterogeneity of the sensitivity
based on I2=62.02% and p=0.03<0.1.
Figure 5 of the study [1] showed that the study by
Gupta et al [5] was an outlier. Therefore, subgroup analy-
sis excluding the outlier was performed, which implied
the heterogeneity was acceptable (sensitivity: I2=25.36%,
p=0.26>0.1; specificity: I2=0.00%, p=0.90>0.1) (fig 1).
Hence, we believe that this study might be also a source
of heterogeneity.

References
1. Ge X, Lan ZK, Chen J, Zhu SY. Effectiveness of contrast-
enhanced ultrasound for detecting the staging and grading
of bladder cancer: a systematic review and meta-analysis.
Med Ultrason 2021;18;23:29-35.
2. Caruso G, Salvaggio G, Campisi A, et al. Bladder tumor
staging: comparison of contrast-enhanced and gray-scale
ultrasound. AJR Am J Roentgenol 2010;194:151-156.  70889.
3. Drudi FM, Di Leo N, Malpassini F, Antonini F, Corongiu
E, Iori F. CEUS in the differentiation between low and
high-grade bladder carcinoma. J Ultrasound 2012;15:247-
251.
Fig 1. Forest plot of CEUS diagnosing bladder cancer, showing
sensitivity and specificity
4. Li Q, Tang J, He E, Li Y, Zhou Y, Wang B. Differentiation
between high- and low-grade urothelial carcinomas using
contrast enhanced ultrasound. Oncotarget 2017;8:70883-
 
5. Gupta VG, Kumar S, Singh SK, Lal A, Kakkar N. Con-
trast enhanced ultrasound in urothelial carcinoma of urinary
bladder: An underutilized staging and grading modality.
Cent European J Urol 2016;69:360-365.

Author’s response

Effectiveness of contrast-enhanced ultrasound for detecting

the staging and grading of bladder cancer:
a systematic review and meta-analysis

Xin-Yue Ge1,2, Zhong-Kai Lan3, Jing Chen2, Shang-Yong Zhu1

1Department of Medical Ultrasound, First Affiliated Hospital of Guangxi Medical University, Nanning, 2Department
of Medical Ultrasound, Fourth Affiliated Hospital of Guangxi Medical University, Liuzhou, 3Department of Medical
Ultrasound, Liuzhou People’s Hospital affiliated to Guangxi Medical University, Liuzhou, Guangxi, China

Dear Editor,
We appreciate Wu et al’s interest in our review and
we have read with interest the comments.
Before the literature search, we limited the inclusion
criteria. The reference standards were considered to be
cystoscopy and/or transurethral resection of bladder tu-
mors and, concerning the index test domain, the readers
were blinded to the final diagnoses of the patients, there-
fore, 3 studies [1-3] were considered low risk. As for the
Li et al study [3], in another study by the same group [4]
published in 2012, the interval time between the index
test and the reference standard was mentioned and we
took it in consideration.
Regarding probability, our interpretation is as fol-
lows: the pre-test probability is the probability of a blad-
der cancer (BC) being detected without taking contrast-
enhanced ultrasound (CEUS) into account. The post-test

probability takes into account the results of CEUS. When
detection of BC was based on a CEUS-positive result,
there was a 70% “post-test” probability of detecting a
subsequent BC. With a negative CEUS, the “post-test”
probability of detecting BC dropped to 2%. Consistent
with other diagnostic meta-analysis articles [5,6].
We believe that moderate heterogeneity exists, so we
stated in the discussion that the P-SEN and PLR− with
95%CI forest plots displayed moderate heterogene-
ity and we made corresponding explanations and analy-
ses.
We agree with your valuable comments on the Gupta
et al study [8]. We included this article in analysis be-
cause in this study CEUS was a good alternative for pre-
operatively T staging and grading of BC.
References
1. Caruso G, Salvaggio G, Campisi A, et al. Bladder tumor
staging: comparison of contrast-enhanced and gray-scale
ultrasound. AJR Am J Roentgenol 2010;194(1):151-156.  Cent European J Urol 2016;69:360-365.
Med Ultrason 2021; 23(2): 238-247 247
2. Drudi FM, Di Leo N, Malpassini F, Antonini F, Corongiu E,
Iori F. CEUS in the differentiation between low and high-grade
bladder carcinoma. J Ultrasound 2012;24;15(4):247-251.
3. Li QY, Tang J, He EH, Li YM, Zhou Y, Wang BJ. Dif-
ferentiation between high- and low-grade urothelial car-
cinomas using contrast enhanced ultrasound. Oncotarget
2017;10;8(41):70883-70889. 
4. Li QY, Tang J, He EH, et al. Clinical utility of three-dimen-
sional contrast-enhanced ultrasound in the differentiation
between noninvasive and invasive neoplasms of urinary
bladder. Eur J Radiol 2012; 81:2936-2942.
5. Sang L, Wang XM, Xu DY, Cai YF. Accuracy of shear
wave elastography for the diagnosis of prostate cancer: A
meta-analysis. Sci Rep, 2017; 7(1): 1949.
6. Yang Y, Zhao XX, Shi JW, Huang Y. Value of shear wave
elastography for diagnosis of primary prostate cancer: a
systematic review and meta-analysis. Med Ultrason 2019,
21(4): 382-388.
7. Gupta VG, Kumar S, Singh SK, Lal A, Kakkar N. Con-
trast enhanced ultrasound in urothelial carcinoma of urinary
bladder: An underutilized staging and grading modality.
 Guidelines for Authors
Medical Ultrasonography is the official publication of the
Romanian Society for Ultrasonography in Medicine and Biol-
ogy (SRUMB).
The journal aims to promote ultrasound diagnosis by
publishing papers that deal with fundamental and clinical
research, scientific reviews, clinical case reports, progress in
ultrasound physics or in the field of medical technology and
equipment, as well as educational papers, special reports and
letters to the editor.
The journal is published quarterly and papers are accepted
for publication in English language.
The official title abbreviation is Med Ultrason. It should
be used in citations, footnotes, legends for figures and all bib-
liographic references.
1. Copyright
Submitting a scientific paper to Medical Ultrasonography
for publishing is subject to compliance with the following
statements:
• the paper is original and has not been published in other
journals or books;
• the paper has not been sent or is not under consideration for
publication elsewhere;
• all authors agree upon publication of the paper.
The statements can be downloaded from the journal web-
site.
All these statements should be included in a formal dec-
laration signed by all the authors. The filled in, signed and
scanned images of these two forms have to be uploaded at the
last step of the submission process (Step 4. Uploading Sup-
plementary Files).
In cases where the paper is accepted for publication, copy-
right will be transferred to Medical Ultrasonography and the
“Iuliu Hatieganu” Medical Publishing House. Authors must
agree to undertake all responsibility for the scientific content
and originality of the paper; Medical Ultrasonography will
take no responsibility whatsoever in this respect.
2. Preparing the manuscript
2.1 General
a. Style & language
Manuscripts should be prepared according to the style of
the journal. They should be written in concise and grammati-
cally correct US English. Authors should ask for assistance if
not writing in their native language.
Papers not conforming to the style of the journal, incom-
prehensible, written in inappropriate English, or not submit-
ted strictly according to the journal guidelines will be re-
turned to the authors for revision, without peer reviewing.
If the manuscript is not complying to accepted standards
of English usage, the authors may be required to bear the cost
of English supervision/ translation.
b. Text and page layout
The papers submitted for publication must be write in
Microsoft Word program, in 12p Times New Roman font, 1.5
line spacing. Page size should be A4, margins must be normal
(2.54 mm in top, bottom, right and left). Page numbering must
begin with the title page. No numbering of the lines is needed.
c. Text length, abbreviations and measurement units
The length of the manuscripts: maximum 5000 words
for original paper, 6000 for reviews, 2000 for pictorial essay,
1500 for case report, 500 for letter to editor (including the
abstract, keywords, references, legend and tables). For letter
to editor only one imagine or table is accepted.
The editors reserve the right of condensing any paper sub-
mitted.
Use only standard abbreviations. Avoid abbreviations in
the title and abstract. Explain every abbreviation before to use
it.
Measurements of length, height, weight, and volume
should be reported in metric units. All hematological and clin-
ical chemistry measurements should be reported in the metric
system in terms of the International System of Units (SI).
d. Images and tables
Imagines (figures) and tables should be grouped in a dis-
tinct section. They must be numbered according to the order in
which they appear in the text- the imagines (figures) in Arabic
numeral (ex. fig 1) and the tables in Roman numeral (ex. tab I).
Do not send them in PDF format!
The captions for figures (images) must be typed on a sepa-
rate page entitled “Legend for figures”. Each table must have
a title. Images or tables should not appear in the text; the de-
sired location for insertion should be indicated by means of a
paragraph, such as: (location for figure no….) or (location for
table no…)
Only high quality images will be accepted for publica-
tion. File formats: BMP or TIFF for images, 300 dpi, image
width 8 cm (single column) or 16.7 cm (double column), for
color images color mode should be CMYK.
The tables (design: Table Grid) are accepted as word doc-
uments included at the end of the main document (see below).
In cases where reproduction of previously published im-
ages is intended, it is necessary to attach the written consent of
the author and of the publishing house where it was formerly
published. All prospective or experimental papers involving
human subjects or experimental animals must include the
agreement granted by the medical ethics commission of the
institution where the research was conducted.
If the manuscript reports medical research involving hu-
man subjects, authors must include a statement confirming
that informed consent was obtained from all subjects, accord-
ing to the World Medical Association Declaration of Helsinki,
revised in 2000, Edinburgh.
3. Structure of the submission files
The manuscript has to be submitted in this form:
 a) in step 2 of submission process, the word document (the
original file) that is uploaded has to include the following: title
of the manuscript, abstract and keywords, main manuscript,
references, tables, legend. Do not insert here the name of the
authors and affiliation in order to ensure a blind review.
b) in step 4 of submission process the supplementary files
that must be uploaded are: title page, images, submission let-
ter, declaration of conflict of interest.
Title page  includes: title of the paper, full names of the
authors, department and institution(s) where the study was
conducted, postal code, city, district, phone and/or fax number
and/or e-mail address for contacting the first author and cor-
responding author, full postal address for correspondence and
ordering reprints.
Abstract  (on a separate page) preceding the text body.
In the case of original papers, abstracts should not exceed
250 words and should have the following structure: 1) aims;
2) material and methods; 3) results; 4) conclusions. Abstracts
for literature reviews and educational papers should not
exceed 200 words. For case reports, the abstract must not
exceed 120 words and must underline the following: 1) pur-
pose of the presentation; 2) peculiarities of the case; ranking
of the issues approached within the general knowledge of the
respective condition.
Three to five keywords must be selected for every paper
from the Index Medicus (http://www.ncbi.nlm.nih.gov/sites/
entrez?db=mesh); the key words should be inserted after the
abstract and separated by semi-colon (term1; term2; term3).
The main document has to be structured as follows:
Introduction  – should define the topic of the paper and
present the status of current knowledge in the field.
Material and methods  – should describe the equipment
employed, the group of patients studied and the methodology.
We recommend specification of the type of ultrasound
equipment employed. The statistical analysis methodology
used must also be described.
Results  – should present the obtained data, in a concise
manner, preferably in tables and diagrams.
Discussions – should present the interpretation of the re-
searcher’s results from the perspective of relevant literature
data.
Conclusions of the paper must be clearly stated in the end.
Acknowledgements – should be made only to those who
have made a substantial contribution to the study
References must include only papers that are quoted in the
text and that have been published. References must be num-
bered in Arabic numerals in the order in which they appear
in the text (where they should be inserted between square
brackets [ ]) and listed in numerical order. Titles of medical
journals must be abbreviated according to the Index Medicus.
All authors must be quoted for an article, if they are up to
six. Over seven authors, only the first three will be quoted,
followed by the “et al” indication. References should be listed
according to the following format (examples):
a) Article:
• Marks WM, Filly RA, Callen PW. Real-time evaluation of
pleural lesions: new observations regarding the probability
of obtaining free fluid. Radiology 1982;142:163-164.
b) Papers published only with DOI numbers:
• Guerriero S, Alcazar JL, Pascual MA, Ajossa S, Olartecoe-
chea B, Hereter L. The pre-operative diagnosis of metastatic
ovarian tumors is related to the origin of the primary tumor.
Ultrasound Obstet Gynecol 2011, doi: 10.1002/uog.10120.
c) Book:
• Talano JV, Gardin JM. Textbook of two dimensional echo-
cardiography. London: Gruene & Stratton, 1983.
d) Book chapter:
• Brooks M. The Liver. In: Goldberg BB, Pettersson H (eds).
Ultrasonography. Oslo, The Nicer Year Book 1996:55-82.
Tables should be added at the end of the main document,
in the same word document.
4. Editorial policy
Medical Ultrasonography promotes evaluation of all the
scientific papers by independent reviewers. The editor-in-
chief or one of the editors evaluates each manuscript and, in
1-3 weeks from reception, decides upon their priority level
(sent to review, rejected without being sent for review or re-
turned to authors with suggestions for improvement before
submitting to review). The editors reserve the right to request
any changes they may consider appropriate, in the title, struc-
ture or body of the paper.
The papers are submitted to two blind reviewers with
expertise in ultrasonography. Based on the reviewers’ recom-
mendations, the editors decide whether a paper is published
or not. In case of marked discrepancy between the decisions
of the two reviewers, the editor may send the manuscript to
another arbitrator for additional comments and a recommend-
ed decision. The full decisional process may last 6-8 weeks.
Failure of the authors to comply with the editorial revision
requests may induce publication rejection.
A 150 euro processing fee (100 euro for the first author if
SRUMB member) is charged for each accepted paper starting
from 1st March 2021. The fee must be paid ahead of publication,
at the time when the authors give the authorization for publi-
cation of the final proof. Instructions for online payment are
available on the journal site. No free reprints of the published
paper are supplied. No additional reprints may be ordered.
5. Submitting manuscripts for publication
To submit a manuscript for publication, one author
should register on our website, and follow the five steps of the
submission process. The journal does not accept email or
CD submissions of articles. The editorial office will send a
confirmation e-mail to the correspondence address.
Remark: Medical Ultrasonography cannot be held respon-
sible for losing or damaging the files delivered through the
Internet.