Download as docx, pdf, or txt
Download as docx, pdf, or txt
You are on page 1of 6

PRABHSHARN SINGH

CL21M012
Clinical engineering (M.tech)

Assignments
- How to evaluate image quality in conventional film screen radiography, Computed
Radiography, Flat panel detector technology, Fluoroscopy, CT and MRI. (30 marks)
- Cardiac CT scanner (10 marks)
- Troubleshooting in a Radiography un =it (10 marks)

No plagiarism allowed, write in your own words and give appropriate references when
figures are included. Include a detailed bibliography if materials are used from the internet.

To be submitted on or before 1st week March 2021

Topic 1

Film screen radiography[1]


Image quality is ability of detector to record point of object on the film screen. It shows the
index of visibility of diagnostically important detail. Aspects of Image quality are (i) contrast,
(ii) spatial resolution, (iii) noise, (iv) geometric factors, (v) detective quantum efficiency, and
(iv) sampling and aliasing
Contrast
X-ray intensities transmitted by different parts of the body is different. Generally called
subject contrast and it depends on density of the organs, atomic number, patient thickness and
photon energy. Sometimes contrast agents are used like barium and iodine to intensify image
for detailed visualization of many organs. It can be a subject contrast and film contrast above
discussed is subject contrast. Part with different thickness attenuates radiation differently. It
varies with effective atomic number Contrast media such as Barium (Z = 56), and Iodine (Z =
53) will enhance subject contrast.
The response of film to different exposure is film contrast.
It depends upon characteristic curve, film density, screen or nonscreen exposure, and film
processing. Homogeneous size of silver grains forms sharp contrast film whereas low
contrast film is made of heterogeneous grains. Chemistry of film define ability to produce
contrast e.g. single or double emulsion.
Resolution
Resolution is ability to image two separate object and visually distinguish from one another.
three types of resolution spatial resolution, contrast resolution and temporal resolution.
Spatial resolution is ability to plot in two dimensions (x, y) also known as ability which can
detect small object. Contrast resolution is ability to distinguish organs that have same
contrast. Film screen detector is known for good contrast resolution. Temporal resolution as
the name suggest it has a time component hence ability to detect image with respect to time.
Various function help in determining resolution like PSF, LSF, ESF and MTF. Point spread
function help in determining response of the imaging system. To quantify resolution of each
component in a imaging system MTF is useful.
Noise
Fluctuation in film density is termed as noise. It is unwanted as it deteriorates image quality
and hence low contrast area are not seen properly. cause of noise lies in screen or film.
Screen noise in caused by non-uniformity of screen, film noise is due to abnormality in
granular structure. The noise is mainly dominated because of discrete nature of X ray photon
this is termed as quantum noise. Noise is inversely proportional to the amount of photons but
has to be kept in certain limit so that patient is not over dosed. Noise has relation with
frequency and plot of these two is called the wiener spectrum.
Geometric Factors
1) Magnification: Image quality and resolution is better when magnification is less
2) Distortion: Due to the unequal magnification because of varying thickness distortion
can occur
3) Focal spot blur: in Xray focal spot is not sharp but it has dimension and this cause
blurriness at the edges

Computed Radiography[2]
The image Interpretation depends on two factor quality of image and ability of interpreter

Figure 1
Various method to evaluate image quality can be classified as shown in the above figure:
Physical methods include modulation transfer factor (MTF), noise, SNR, and detection
quantum efficiency (DQE). Psychophysical evaluation methods include rose model (RM),
contrast detail analysis (CDA) and subjective assessment of physical parameters. Clinical
performance measurement methods include receiver-operating characteristics (ROC) and
visual grading characteristic (VGC).
Detective Quantum efficiency (DQE)
 Measurement parameters are Modulation transfer function(MTF) and Noise power
spectrum(NPS)
MTF describe behavior of spatial resolution at all frequencies:
1. Image is imported
2. Edge spread function (ESF) is constructed
3. From ESF line spread function (LSF) is obtained
4. Fourier transformation of LSF to obtain MTF
NPS measure noise attributes and pattern at all frequencies:
1. Obtain Flood exposure image
2. Sampling is performed
3. Data is linearized
4. 2D Fourier transformation to average data

 Focus on detector “image receptor” performance

Rose model (RM)


 Using images to estimate the maximum of info can be translated into visible image by
photons numbers

Contrast detail analysis (CDA)


 Using phantom of range of different densities and sizes of sample (quantitative
measurements)

Information entropy (IE)


 Measuring(quantitatively) how much information (randomness/uncertainty) is
transmitted by signal/image

Receiver operating characteristics (ROC)


 Measuring sensitivity and specificity of imaging system by comparing images to true
disease status of patient

Visual grading characteristic (VGC)


 Using one or more reference images or quality criteria to assess visibility of specific
features on image

Flat panel detector technology[3]


Flat-panel detectors for x-ray imaging systems are a fast-emerging technology. Noise is the
major problem whereas pixel signal-to-noise ratio (SNR) increasing dose is not a option

Pre-sampling modulation transfer function (MTF)


Scintillation layer convert x ray to light photons. MTF of cascade system is multiplication of
individual multiplication MTF of each stage.

Where a and b sides of pixel in mm


Creation of guidewire images
Guidewire is used to create noise-free image. 3D model is projected onto a parallel imaging
plane. The projections is sampled on a square. This array is processed to obtain point spread
function (PSF) obtained by taking the inverse discrete Fourier’s transform (DFT). The sample
is then binned together to the desired pixel sizes so as to model partial area effects.

Noise
Noise is of two types of quantum noise and electronic noise
 Quantum noise
Fluctuation of photon counts. X-rays count follow Poisson Distribution with variance
and mean. A gray-level calibration factor converts the detected signal in photons to a
signal in gray levels. Due to sampling, noise power spectrum NPS a digital system is
aliased at frequencies above the cutoff frequency after mathematical operation like
convolution noise is given by

Quantum noise image were created using these values

 Electronic noise
Noise from sources such as thin film transistors switch on the pixel and the
preamplifier on the data line. This is independent of the exposure. Assuming pixel
gray level variance due to electronic noise is equal to the noise variance produced by
Ne x-ray photons, it is given by

 Creation of noise images


Now these two separate images of quantum and electronic images were added
together to obtain the noise images.

Fluoroscopy[4]
Image quality assessment

Contrast scale
Image uniformity and distortion
Spatial (high contrast) resolution
Noise (and Video signal)
Contrast (low contrast) resolution
Overall Image Quality (Contrast/Detail Diagram)
IQ dependence of “window” and matrix
IQ dependence of reconstruction/frame rate
IQ dependence of image processing
Subjective assessment (eyes condition)
Attenuate the X-ray output (1mm Cu)
Check all II field sizes with all test objects
Adjust TV monitor (contrast/brightness)
visible field size/distortions/homogeneity
noise, contrast resolution (contrast/detail)
lim. spatial resolution (no attenuation)

CT scan[5]
Three factor of image quality
(i) spatial resolution (ii) noise and (iii) contrast

ARTIFACTS
Artifacts are the degradation of image quality that leads to wrong or miss diagnosis
 Motion Artifacts
Patient during the scan may move that causes unpredictable data during scan and a streak can
be seen on the image, with increase in scan time motion artifacts are increased. This cause
error in reconstruction. Double image may appear o0r ghosting of image is happened.
 Streak Artifacts
Dark and light lines appear due to absence of transmitted x rays. High density material such
as metal implants.
 Beam Hardening Artifacts
Absorption of low energy beam while passing through patient cause mean energy to increase
as a result beam is hardened that cause underestimation of HU. Can be corrected by
algorithm.
 Ring Artifacts
Mis calibration of one detector in rotate-rotate system of 3rd gen CT. due to this incorrect data
will appear as a ring on every projection
 Partial Volume Artifacts
Partial volume artifact is result of averaging the linear attenuation coefficient in a given voxel
that is heterogeneous in composition.
 Rod Artifacts
Rod artifacts due to varying shape of high contrast object.
 Spiral Artifacts
A helical scan gives an image similar to partial volume averaging (PVA). In one direction,
the PVA is determined by collimation, and in the other direction, it is by collimation and
table increment per rotation. As a result, instead of a perfect circle, an ellipse like
reconstruction is done.

MRI[6]
Quality Monitoring Through Stages of Imaging Workflow:
After the scan is taken image is processed and transmitted to storage system has adverse
effect on image quality. The potential of an MRI system to generate high contrast image can
be reduced by improper system parameter settings. Also, patient motion during scan
introduce blur and artifacts which decrease sharpness and details

Limitations of Subjective Quality Evaluation:


Gold standard of quality is considered as human evaluation. However, it has certain
limitation in clinical trials as the volume is large. It depends on human emotion, lightning and
environment effect the outcomes this may lead to intra inter variability. There is little
tolerance for the cumbersomeness and the variability of the outcomes of subjective image
quality evaluation.

Inter-Site Variations in MRI System Parameters:


Cost-saving measures by the pharmaceutical companies requires that only MRI systems
available at the clinical trial sites are utilized for acquisition. This leads to variation because a
single reference image can’t act as model for different trial sites. There are no image that can
be consider as ideal for reference image. This is dependent on standardization of image
quality before data analysis

Intra-Subject Variations in Acquisition Parameters:


Images take at different time points requires registration before analysis. Also scanner change
is a possibility over a period of time. Hence intensity mismatch is common between different
images. Intensity normalization requires re-evaluation.

Conformity With Acquisition Protocols:


Image sequencing parameters vary in case of brain MRI. Post-acquisition image quality
evaluation at the CRO is one of key steps towards conformity with the acquisition protocols.

References
1. Book: The Physics of Radiology and Imaging by Thayalan K.
2. https://onlinelibrary.wiley.com/doi/pdf/10.1002/j.2051-3909.2012.tb00174.x
Figure1 https://onlinelibrary.wiley.com/doi/pdf/10.1002/j.2051-3909.2012.tb00174.x
3. https://aapm.onlinelibrary.wiley.com/doi/epdf/10.1118/1.1628278
4. Farr's physics for medical imaging
5. Book: The Physics of Radiology and Imaging by Thayalan K.
6. https://www.ncbi.nlm.nih.gov/pmc/articles/PMC6126794/

You might also like