Professional Documents
Culture Documents
ACR TI-RADS: Pitfalls, Solutions, and Future Directions
ACR TI-RADS: Pitfalls, Solutions, and Future Directions
org
2040
Neurologic/Head and Neck Imaging
Nodule
Characteristic US Features with Point Values
Composition Cystic = 0, spongiform = 0, mixed solid-cystic = 1, solid = 2, indeterminate = 2
Echogenicity Anechoic = 0, isoechoic = 1, hyperechoic = 1, indeterminate = 1, hypoechoic = 2, very hy-
poechoic = 3
Shape Wider than tall = 0, taller than wide = 3
Margin Smooth = 0, ill defined = 0, irregular = 2, lobulated = 2, extrathyroidal extension = 3
Echogenic foci None = 0, comet-tail artifact = 0, macrocalcifications = 1, peripheral or rim = 2, punctate = 3
Note.—Points are totaled by adding single selections from the composition, echogenicity, shape, and margin
categories to selections from the echogenic foci category, which allows multiple selections.
the lowest rate of unnecessary thyroid nodule First, still images are obtained in the trans-
FNAB at 25.3% (31). verse and sagittal orientations. Next, US cine
clips of each lobe and the isthmus are obtained in
Challenges, Pitfalls, and Solutions the transverse and sagittal planes.
Implementing new departmental or institutional Thyroid nodules smaller than 5 mm should
guidelines for assessment of a specific condi- generally be ignored. In nodules larger than
tion or organ system presents many challenges. 5 mm, still images and cine clips should be ob-
Practice-wide adoption of ACR TI-RADS can tained in transverse and sagittal planes, depend-
present issues related to education, workflow, and ing on the relevant US features (Table 2). If mul-
adherence to interpretation and reporting stan- tiple nodules are found, then the four nodules
dards. We highlight several potential pitfalls and that are the largest or most suspicious should
challenges related to ACR TI-RADS with some undergo imaging and be cataloged. Each nodule
proposed solutions (Table 4). is measured in anterior-posterior, transverse, and
craniocaudal dimensions.
Imaging, Workflow, Before starting the program, it is crucial to edu-
and Structured Reporting cate and train sonographers to recognize relevant
When managing multiple sites, establishing a uni- TI-RADS US features. It is also important to de-
fied sonographer protocol is an essential first step velop a worksheet or equivalent annotation system
to achieve consistency of service. to assist the sonographer in documenting nodules
2044 November-December 2019 radiographics.rsna.org
and other information that should be brought to the thyroid gland, a comprehensive description
the radiologist’s attention. An illustration of the of up to four nodules to index their size, location,
thyroid gland could be used to mark the loca- core US features, numeric score and TI-RADS
tion of a nodule, with adjacent spaces in which category, any change from a prior examination (if
to record measurements and TI-RADS–specific applicable), and a final recommendation (Fig 2).
features (Fig 1). Associated imaging findings such
as lymph nodes can be similarly illustrated. Information Included in the Report
Structured reporting has been shown to in- Radiologists are faced with several challenges when
crease report homogeneity and language unifor- generating a meaningful structured report of a
mity, standardize conclusions and recommenda- thyroid US examination, especially when there are
tions, and lead to better communication between several nodules. When multiple nodules are pres-
radiologists and clinicians (32). Three levels of ent, the sonographer should identify the four most
structured reporting have been described. The concerning nodules and measure them. Cine clips
first is use of headings such as “Indications,” of the entire thyroid gland should be obtained to
“Comparison,” “Findings,” and “Impression.” demonstrate the nodules. The radiologist should
The second level adds organ systems as subhead- also identify the four most concerning nodules and
ings. This is also known as itemized reporting. describe them using the standard lexicon.
The third level adds standardized terminology Careful comparison with any prior images helps
and language (33,34). TI-RADS reports are best determine if there has been threshold growth,
suited for the third level, as standardized terms which is defined as enlargement by 20% and 2
are available to describe thyroid nodules. mm in two dimensions or a 50% increase in the
A TI-RADS–specific structured report can be volume of the nodule. If there is evidence that the
designed by using dictation software that can cre- TR category has increased but the nodule still
ate templates with customizable fields. This enables does not warrant biopsy because of size criteria,
input of various types of standardized text and then follow-up in 1 year is needed, and the prior
numeric values. Features such as pick lists allow surveillance recommendations are superseded.
selection of designated options in a field, which A major component of TI-RADS is the detailed
improves adherence to terminology. On the basis lexicon that provides definitions for US features.
of ACR recommendations, default components Various descriptive terms that were once com-
for such a report include a general assessment of monly used to describe imaging features are no
RG • Volume 39 Number 7 Tappouni et al 2045
Figure 3. Multiple spongiform nodules in the same patient. US images show multiple similar-appearing low-risk nodules. Only the
largest one needs to be characterized in the structured report.
Figure 5. Three mildly to moderately suspicious thyroid nodules in the same patient. (a) US image shows a TR4 nodule with macro-
calcification in the isthmus. (b) US image demonstrates a TR3 nodule measuring 2.6 cm. (c) US image shows a TR3 nodule measur-
ing 3.1 cm. All three nodules meet FNAB criteria, but biopsy should be recommended only for the two largest and most suspicious
nodules, which are the nodules in a and c in this case.
If the nodule does not have peripheral calci- technique could depict imaging features in an
fication or cystic or spongiform morphology, a inaccurate or misleading fashion. It is mandatory
full feature assessment of the nodule should be that all personnel be taught the TI-RADS lexicon
performed. and trained how to perform all relevant imaging
By following this algorithmic approach, the examinations, including how to acquire cine clips.
radiologist can quickly differentiate suspicious Inconsistency in reporting imaging features can
from benign nodules and focus effort on nodules undermine the scoring system.
that must be indexed. Fortunately, agreement on the overall TI-
Radiologist agreement for nodule features RADS category is generally more consistent
ranges from fair to moderate (35–37). A spongi- across readers than agreement on individual
form nodule is defined as having interleaved cys- features, and agreement for recommendation
tic spaces in more than 50% of the nodule. This to pursue biopsy is even higher (35,37). This is
type of nodule can be challenging to interpret a strength of TI-RADS, as maintaining reason-
(Fig 7) because of the specificity of the lexicon able recommendations for treatment is an im-
definition. portant factor from the perspective of patient-
Other features can present problems during centered care.
image interpretation. It is sometimes difficult to To reduce interreader variability, radiologists
differentiate punctate echogenic foci that represent should be aware of their own risk preference,
microcalcifications from the comet-tail artifact that conduct regular peer learning meetings to discuss
indicates inspissated colloid (Fig 8). It is debated discrepant cases, and obtain second opinion con-
whether to add points for echogenicities along the sultations when there is uncertainty.
walls of cystic spaces. Also, it can be troublesome
to obtain an accurate measurement of a nodule in Implementation and Quality Measures
a patient with a background multinodular gland or There are several opportunities to introduce
a significantly heterogeneous gland (Fig 9). quality measures in thyroid imaging. A standard
Interpretation could be affected by operator lexicon and structured reporting system are avail-
dependence. Sonographer inexperience and poor able to convey relative risk of malignancy, and
RG • Volume 39 Number 7 Tappouni et al 2047
Figure 6. Flowchart shows an algorithmic approach to interpreting and classifying a thyroid nodule by using ACR TI-RADS features.
While all components should be reported as required, useful checkpoints include evaluation of composition and peripheral calcifica-
tion. A nodule with these characteristics may be placed immediately into higher-suspicion categories. This is an efficient way to stratify
indexed nodules.
Figure 7. Challenges of evaluation of spongiform nodules. (a) US image shows a smaller
superficial nodule with characteristic tiny spaces and a large comet-tail artifact caused by col-
loid content. The larger nodule also has spongiform imaging findings in its upper portion, but
since this constitutes less than 50% of its composition, it is considered to be solid. (b) US image
shows a spongiform nodule in a different patient. This nodule demonstrates both stereotypical
features with at least 50% involvement of the nodule.
radiologists are involved in the performance of measure 406 (“Appropriate Follow-up Imaging for
US-guided biopsies. Incidental Thyroid Nodules in Patients”) is applied
Currently, there are two existing national quality to incidental thyroid nodules identified at CT and
measures that can be applied to thyroid nodules. MRI. MIPS measure 265 (“Biopsy Follow-up”)
Merit-based Incentive Payment System (MIPS) ensures that the performing physician has reviewed
2048 November-December 2019 radiographics.rsna.org
Figure 9. Difficult interpretation of nodules. (a) US image shows a heterogeneous nodule that appears to span the
length of the thyroid lobe. (b) US image demonstrates improved conspicuity of an interface between two abutting
nodules. This image was obtained after the sonographer made subtle adjustments to the insonation angle and other
technical parameters, which revealed imaging features that had been obscured slightly by a macrocalcification.
the biopsy results and communicated them to the adequate for diagnosis. Quality measures re-
referring physician and patient. There are oppor- flecting diagnostic yield (or nondiagnostic rate),
tunities for new quality measures in the areas of false-positive rate, and false-negative rate can be
diagnostic accuracy, appropriateness, efficiency, compared across providers who perform these
and patient-centered care. procedures to ensure optimal procedural tech-
nique and avoid unnecessary repeat biopsies.
Diagnostic Accuracy.—Since there are data
regarding the expected rates of malignancy for Patient-centered Care.—One opportunity to
TR1–TR5 nodules, measures such as positive improve patient-centered care involves measuring
predictive value (PPV) for TR3–TR5 nodules and and managing patient anxiety and pain related to
negative predictive value for TR2 nodules can be US-guided biopsies. Another opportunity is the
benchmarked to national data. PPV measures ex- creation of patient-friendly documents that help
ist in the ACR Lung Cancer Screening Registry patients understand the results of their thyroid US,
and CT Colonography Registry (38). provide guidance on how they can get more in-
formation, and what the next steps may be. These
Appropriateness.—Given that multiple guide- types of activities allow radiologists to be more
lines are available to help stratify thyroid nodules engaged with patients, assist with shared decision
by risk, quality measures for biopsy and follow- making, and address issues that are important to
up imaging can be based on established guide- patients and their family members (40).
lines (39). The existing MIPS measures can be It may be challenging to incorporate ACR
updated to reflect current recommendations that TI-RADS into quality measures at the local and
are based on nodule morphology and patient age national levels. Multiple practice guidelines, as
as well as nodule size. described earlier in this article, can lead to differ-
ent criteria for biopsy and lack of reconciliation
Efficiency.—Thyroid biopsy techniques vary with regards to biopsy and imaging follow-up.
among different practitioners. There is an incon- Thyroid US may be performed and its results
sistent use of resources such as on-site cytopa- interpreted not by radiologists, but by endocri-
thologists who determine if biopsy samples are nologists who use ATA guidelines.
RG • Volume 39 Number 7 Tappouni et al 2049
Figure 10. Follicular neoplasm. (a) Axial PET/CT image depicts a hypermetabolic nodule in the left thyroid
lobe. (b) US image reveals a corresponding TR4 nodule in the left lobe, which was biopsied and shown to
be a follicular neoplasm.
US Evaluation Guided
by FDG-PET Activity
Diffuse thyroid gland uptake can be seen in
thyroiditis or Graves disease. Focal uptake, which
is seen mostly in thyroid nodules, can be seen in
benign and malignant conditions (41). Incidental
Figure 11. TR5 nodule in a patient with a his- hypermetabolic thyroid nodules can be a relatively
tory of papillary cancer. US image obtained for common finding at fluorine 18 fluorodeoxyglucose
preoperative evaluation shows a nodule that had (18F-FDG) PET/CT (42). The malignancy rate
previously been biopsied at a different facility in hypermetabolic thyroid nodules ranges broadly
and shown to be papillary cancer. It was evalu-
ated to be a TR5 or highly suspicious nodule. from 14% to 81%, but a rate of 50% is often quoted
in the nuclear medicine literature (43). Many of
these hypermetabolic nodules subsequently un-
In this environment, it is important for radiolo- dergo US with or without FNAB (Fig 10).
gists interpreting thyroid US images to establish In a large retrospective study, the incidence of
good relationships with endocrinologists and otolar- malignancy in biopsied focal hypermetabolic thy-
yngology surgeons to gain their professional respect roid lesions was 21.4%. Nodules exhibiting a ratio
and garner trust in TI-RADS. It is also important of thyroid maximum standardized uptake value
to be familiar with ATA guidelines, as some patients (SUVmax) to background SUVmax greater than 2
may undergo repeat imaging or FNAB in a radiol- were shown to be correlated with malignancy and
ogy department after an initial endocrinologist visit. should be examined with US and possibly undergo
FNAB (43). Because of this, PET activity might be
Future Directions considered a future factor in TI-RADS by adding
As acknowledged by the TI-RADS Committee in points to a nodule with hypermetabolic activity.
the initial white paper, a number of situations are
not addressed in its current scope. The primary Additional Granularity
focus is on assessment of US features, but it may Other structured reporting systems such as BI-
be possible to make recommendations for man- RADS and LI-RADS (Liver Imaging Reporting
agement more specific through inclusion of other and Data System) contain categories that can be
clinical and imaging factors. applied to TI-RADS. One example is a thyroid
For instance, a pretest probability risk profile nodule that is known to be malignant but has
based on personal and family history (eg, ra- not been (or potentially will not be) treated. This
diation exposure, multiple endocrine neoplasia would be analogous to the BI-RADS 6 category
syndrome, and first-degree relatives with thyroid (Fig 11) and may be particularly important when
2050 November-December 2019 radiographics.rsna.org
Indeterminate and
Discrepant FNAB Results Figure 13. Discordance between imaging fea-
Analysis of thyroid nodule FNAB specimens tures of a nodule and tissue sampling results. US
shows nondiagnostic or indeterminate results in image depicts a nodule that was initially classified
15%–20% of patients (1,4). Repeat US or FNAB as TR4 and subsequently biopsied. Pathologic
analysis revealed atypia of undetermined signifi-
is often necessary to monitor the lesion, confirm cance. Repeat FNAB showed papillary cancer.
the diagnosis, or perform molecular testing. If re-
peat FNAB results are indeterminate, the patient
may undergo thyroidectomy or imaging surveil- increase in diagnostic accuracy when morpho-
lance (1). Patients with suspicious nodules and logic analysis was augmented with contrast-
negative or indeterminate FNAB results should enhanced US (44,45).
undergo repeat biopsy or be referred to a surgeon Similarly, US elastography has been used for
for possible thyroidectomy (Fig 13). assessment of mechanical properties in multiple
For example, a TR5 lesion with negative organs, including the liver, breast, and thyroid
FNAB results should undergo a repeat biopsy (46). Combining elastography with B-mode US
because a sampling error may have taken place. features has shown some increase in the sensitiv-
Providing a specific recommendation for repeat ity and specificity of diagnosis (46,47).
biopsy or surgery referral may help in such
circumstances. It is also important to coordinate Conclusion
with the referring physician to integrate this The incidence of thyroid nodules is rapidly rising,
information. When addressing the most chal- and careful risk stratification is an important ele-
lenging cases, it may be appropriate to hold an ment in preventing overdiagnosis and excessive
interdisciplinary conference. treatment. The role of the radiologist is critical to
reduce unnecessary imaging and biopsies. One
Inclusion of Contrast-enhanced US way to accomplish this is by assessing US features
and Elastography of thyroid nodules in a standardized fashion.
In LI-RADS, the utility of contrast-enhanced ACR TI-RADS provides a framework to gener-
US to depict hepatocellular carcinoma is fully ate structured reporting and consistently classify
recognized. As US is the dominant imaging nodules to provide appropriate management
modality used in thyroid evaluation, contrast- recommendations.
enhanced US could be used to extend TI- Implementing ACR TI-RADS affects multiple
RADS. Various studies showed a significant aspects of the imaging workflow and also has
RG • Volume 39 Number 7 Tappouni et al 2051
pitfalls and challenges. We have presented some Cancer: The American Thyroid Association Guidelines
Task Force on Thyroid Nodules and Differentiated Thyroid
solutions to issues encountered in our experi- Cancer. Thyroid 2016;26(1):1–133.
ence with this system. Ultimately, TI-RADS is 18. Shin JH, Baek JH, Chung J, et al. Ultrasonography Di-
a framework with the potential for numerous agnosis and Imaging-Based Management of Thyroid
Nodules: Revised Korean Society of Thyroid Radiology
research opportunities and has innate flexibility Consensus Statement and Recommendations. Korean J
and extensibility for future modification. Radiol 2016;17(3):370–395.
19. Russ G, Bonnema SJ, Erdogan MF, Durante C, Ngu R,
Disclosures of Conflicts of Interest.—R.R.T. Activities related Leenhardt L. European Thyroid Association Guidelines
to the present article: disclosed no relevant relationships. Ac- for Ultrasound Malignancy Risk Stratification of Thyroid
tivities not related to the present article: consultant to Behold.ai Nodules in Adults: The EU-TIRADS. Eur Thyroid J
Technologies. Other activities: disclosed no relevant relation- 2017;6(5):225–237.
ships. J.N.I. Activities related to the present article: disclosed no 20. Horvath E, Majlis S, Rossi R, et al. An ultrasonogram
relevant relationships. Activities not related to the present article: reporting system for thyroid nodules stratifying cancer
received compensation as an expert witness in multiple court risk for clinical management. J Clin Endocrinol Metab
cases; institution received funding from Coverys Community 2009;94(5):1748–1751.
Healthcare Foundation; speaker for Moffitt Cancer Center. 21. Na DG, Baek JH, Sung JY, et al. Thyroid Imaging Reporting
Other activities: disclosed no relevant relationships. and Data System Risk Stratification of Thyroid Nodules:
Categorization Based on Solidity and Echogenicity. Thyroid
2016;26(4):562–572.
References 22. Kwak JY, Han KH, Yoon JH, et al. Thyroid imaging report-
1. Fisher SB, Perrier ND. The incidental thyroid nodule. CA ing and data system for US features of nodules: a step in
Cancer J Clin 2018;68(2):97–105. establishing better stratification of cancer risk. Radiology
2. Wilhelm S. Evaluation of thyroid incidentaloma. Surg Clin 2011;260(3):892–899.
North Am 2014;94(3):485–497. 23. Frates MC, Benson CB, Charboneau JW, et al. Management
3. Guth S, Theune U, Aberle J, Galach A, Bamberger CM. of thyroid nodules detected at US: Society of Radiologists
Very high prevalence of thyroid nodules detected by high in Ultrasound consensus conference statement. Radiology
frequency (13 MHz) ultrasound examination. Eur J Clin 2005;237(3):794–800.
Invest 2009;39(8):699–706. 24. Gharib H, Papini E, Garber JR, et al. American Association
4. Mazzaferri EL. Management of a solitary thyroid nodule. of Clinical Endocrinologists, American College of Endocri-
N Engl J Med 1993;328(8):553–559. nology, and Associazione Medici Endocrinologi Medical
5. Davies L, Welch HG. Current thyroid cancer trends in Guidelines for Clinical Practice for the Diagnosis and
the United States. JAMA Otolaryngol Head Neck Surg Management of Thyroid Nodules: 2016 Update. Endocr
2014;140(4):317–322. Pract 2016;22(5):622–639.
6. Choi YM, Kim WG, Kwon H, et al. Changes in standard- 25. National Comprehensive Cancer Network. NCCN Clinical
ized mortality rates from thyroid cancer in Korea between Practice Guidelines in Oncology AND (NCCN Guidelines)
1985 and 2015: analysis of Korean national data. Cancer AND Thyroid Carcinoma version 2.2014. Fort Washington,
2017;123(24):4808–4814. Pa: National Comprehensive Cancer Network, 2014.
7. Ahn HS, Kim HJ, Welch HG. Korea’s thyroid-cancer 26. Middleton WD, Teefey SA, Reading CC, et al. Multiin-
“epidemic”: screening and overdiagnosis. N Engl J Med stitutional Analysis of Thyroid Nodule Risk Stratification
2014;371(19):1765–1767. Using the American College of Radiology Thyroid Imag-
8. Aspinall SR, Ong SG, Wilson MS, Lennard TW. How shall ing Reporting and Data System. AJR Am J Roentgenol
we manage the incidentally found thyroid nodule? Surgeon 2017;208(6):1331–1341.
2013;11(2):96–104. 27. Hoang JK, Middleton WD, Farjat AE, et al. Reduction
9. Ezzat S, Sarti DA, Cain DR, Braunstein GD. Thyroid inci- in Thyroid Nodule Biopsies and Improved Accuracy with
dentalomas: prevalence by palpation and ultrasonography. American College of Radiology Thyroid Imaging Reporting
Arch Intern Med 1994;154(16):1838–1840. and Data System. Radiology 2018;287(1):185–193.
10. Griffin A, Brito JP, Bahl M, Hoang JK. Applying Criteria of 28. Zheng Y, Xu S, Kang H, Zhan W. A Single-Center Ret-
Active Surveillance to Low-Risk Papillary Thyroid Cancer rospective Validation Study of the American College of
Over a Decade: How Many Surgeries and Complications Radiology Thyroid Imaging Reporting and Data System.
Can Be Avoided? Thyroid 2017;27(4):518–523. Ultrasound Q 2018;34(2):77–83.
11. Ito Y, Miyauchi A, Oda H. Low-risk papillary microcarci- 29. Middleton WD, Teefey SA, Reading CC, et al. Comparison
noma of the thyroid: a review of active surveillance trials. of Performance Characteristics of American College of
Eur J Surg Oncol 2018;44(3):307–315. Radiology TI-RADS, Korean Society of Thyroid Radiology
12. Oda H, Miyauchi A, Ito Y, et al. Comparison of the costs TIRADS, and American Thyroid Association Guidelines.
of active surveillance and immediate surgery in the manage- AJR Am J Roentgenol 2018;210(5):1148–1154.
ment of low-risk papillary microcarcinoma of the thyroid. 30. Koseoglu Atilla FD, Ozgen Saydam B, Erarslan NA, et al.
Endocr J 2017;64(1):59–64. Does the ACR TI-RADS scoring allow us to safely avoid
13. Aschebrook-Kilfoy B, Schechter RB, Shih YC, et al. The unnecessary thyroid biopsy? single center analysis in a large
clinical and economic burden of a sustained increase in cohort. Endocrine 2018;61(3):398–402.
thyroid cancer incidence. Cancer Epidemiol Biomarkers 31. Ha EJ, Na DG, Baek JH, Sung JY, Kim JH, Kang SYUS.
Prev 2013;22(7):1252–1259. US Fine-Needle Aspiration Biopsy for Thyroid Malignancy:
14. Tessler FN, Middleton WD, Grant EG, et al. ACR Thyroid Diagnostic Performance of Seven Society Guidelines Applied
Imaging, Reporting and Data System (TI-RADS): White to 2000 Thyroid Nodules. Radiology 2018;287(3):893–900.
Paper of the ACR TI-RADS Committee. J Am Coll Radiol 32. Gholamrezanezhad A, Kessler M, Hayeri SM. The Need
2017;14(5):587–595. for Standardization of Musculoskeletal Practice Reporting:
15. Grant EG, Tessler FN, Hoang JK, et al. Thyroid Ultrasound Learning From ACR BI-RADS, Liver Imaging-Reporting
Reporting Lexicon: White Paper of the ACR Thyroid Imag- and Data System, and Prostate Imaging-Reporting and Data
ing, Reporting and Data System (TIRADS) Committee. J System. J Am Coll Radiol 2017;14(12):1585–1587.
Am Coll Radiol 2015;12(Part A):1272–1279. 33. Cramer JA, Eisenmenger LB, Pierson NS, Dhatt HS, Heilbrun
16. Ha EJ, Baek JH, Na DG. Risk Stratification of Thyroid ME. Structured and templated reporting: an overview. Appl
Nodules on Ultrasonography: Current Status and Perspec- Radiol 2014;43:18–21. https://www.appliedradiology.com/
tives. Thyroid 2017;27(12):1463–1468. articles/structured-and-templated-reporting-an-overview.
17. Haugen BR, Alexander EK, Bible KC, et al. 2015 American 34. Weiss DL, Langlotz CP. Structured reporting: patient
Thyroid Association Management Guidelines for Adult care enhancement or productivity nightmare? Radiology
Patients with Thyroid Nodules and Differentiated Thyroid 2008;249(3):739–747.
2052 November-December 2019 radiographics.rsna.org
35. Hoang JK, Middleton WD, Farjat AE, et al. Interobserver 42. Ozkol V, Alper E, Aydin N, Ozkol HF, Topal NB, Akpinar
Variability of Sonographic Features Used in the American AT. The clinical value of incidental 18F-fluorodeoxyglucose-
College of Radiology Thyroid Imaging Reporting and Data avid foci detected on positron emission tomography/computed
System. AJR Am J Roentgenol 2018;211(1):162–167. tomography. Nucl Med Commun 2010;31 (2):128–136.
36. Itani M, Assaker R, Moshiri M, Dubinsky TJ, Dighe MK. 43. Barrio M, Czernin J, Yeh MW, et al. The incidence of
Inter-observer Variability in the American College of Radiol- thyroid cancer in focal hypermetabolic thyroid lesions: an
ogy Thyroid Imaging Reporting and Data System: In-Depth 18F-FDG PET/CT study in more than 6000 patients. Nucl
Analysis and Areas for Improvement. Ultrasound Med Biol Med Commun 2016;37(12):1290–1296.
2019;45(2):461–470. 44. Zhang Y, Zhou P, Tian SM, Zhao YF, Li JL, Li L. Useful-
37. Grani G, Lamartina L, Cantisani V, Maranghi M, Lucia ness of combined use of contrast-enhanced ultrasound and
P, Durante C. Interobserver agreement of various thyroid TI-RADS classification for the differentiation of benign
imaging reporting and data systems. Endocr Connect from malignant lesions of thyroid nodules. Eur Radiol
2018;7(1):1–7. 2017;27(4):1527–1536.
38. ACR National Radiology Data Registry. American College 45. Zhao H, Liu X, Lei B, et al. Diagnostic performance of
of Radiology. https://www.acr.org/Practice-Management- thyroid imaging reporting and data system (TI-RADS) alone
Quality-Informatics/Registries. Accessed February 15, 2019. and in combination with contrast-enhanced ultrasonography
39. Itri JN, Raghavan K, Patel SB, et al. Developing Quality for the characterization of thyroid nodules. Clin Hemorheol
Measures for Diagnostic Radiologists: Part 2. J Am Coll Microcirc 2019;72(1):95–106.
Radiol 2018;15(10):1366–1384. 46. Sigrist RMS, Liau J, Kaffas AE, Chammas MC, Willmann
40. Itri JN. Patient-centered Radiology. RadioGraphics JK. Ultrasound Elastography: Review of Techniques and
2015;35(6):1835–1846. Clinical Applications. Theranostics 2017;7(5):1303–1329.
41. Are C, Hsu JF, Schoder H, Shah JP, Larson SM, Shaha 47. Du YR, Ji CL, Wu Y, Gu XG. Combination of ultrasound
AR. FDG-PET detected thyroid incidentalomas: need elastography with TI-RADS in the diagnosis of small thyroid
for further investigation? Ann Surg Oncol 2007;14(1): nodules (≤10 mm): a new method to increase the diagnostic
239–247. performance. Eur J Radiol 2018;109:33–40.
TM
This journal-based SA-CME activity has been approved for AMA PRA Category 1 Credit . See rsna.org/learning-center-rg.