2) Evaluation of Medical Image Quality
1.
Rank order the three levels of image quality evaluations in terms of their objectivity from the highest to the lowest:
A) Clinical, physical, perceptual
B) Clinical, perceptual, physical
C) Perceptual, physical, clinical
D) Physical, perceptual, clinical
Answer:
D) Physical, perceptual, clinical
2.
Rank order the three levels of image quality evaluations in terms of their clinical relevance from the highest to the lowest:
A) Clinical, physical, perceptual
B) Clinical, perceptual, physical
C) Perceptual, physical, clinical
D) Physical, perceptual, clinical
Answer:
B) Clinical, perceptual, physical
3.
In the physical metrics below, which ones implies a higher image quality of the value of the metric is decreased
A) Contrast to noise ratio
B) Noise power spectrum
C) Modulation transfer function
D) Detective quantum efficiency
Answer:
B) Noise power spectrum
http://physics.rsna.org/section/default.asp?id=PHYS0610
4.
Select all correct matches between the image and the metric that is derived from it
A) Edge image-MTF
B) Edge image-NPS
C) Uniform image-MTF
D) Uniform image-NPS
Answer:
A) Edge image-MTF
D) Uniform image-NPS
5.
Measurement of the DQE involves which of the following (select all that apply):
A) MTF
B) NPS
C) Ideal SNR
D) Scattered radiation
Answer:
A) MTF
B) NPS
C) Ideal SNR
The content was from the prior module
and
6.
Effective DQE is more complete than DQE by the addition of the which of these factors (select all that apply)
A) Magnification
B) Scattered radiation
C) Anatomical noise
D) Focal spot blur
E) Anti-scatter grid
Answer:
A) Magnification
B) Scattered radiation
D) Focal spot blur
E) Anti-scatter grid
The measurement of eDQE involves imaging an actual phantom emulating the body part relevant to the imaging application, and assessing the MTF and the NPS from the images captured from the phantom. As such, the resulting eDQE encompasses all of those attributes and is a system performance indicator. However, it is difficult to standardize this measurement across multiple imaging systems, especially when manufactured by different vendors.
This is the question with 31 possible answers! One might be tempted to say that anatomical noise might be correct given that an anatomical phantom is used. Apparently the question writer felt otherwise perhaps one can only have anatomic noise with a pt with anatomy, not a phantom . (I took the quiz 11 times to find the "correct" answer!
7.
Four alternative forced choice (4AFC) evaluation involved which one of the following:
A) Selecting a superior imaging system among four alternatives
B) Selecting a target object among four alternatives
C) Selecting an ideal observer among four alternatives
D) Selecting a scoring method among four alternatives
Answer:
B) Selecting a target object among four alternatives
8.
Which one of these are main requirements of a performance based evaluation (select all that apply)
A) Knowledge of the ground truth
B) A compilation of normal cases
C) A compilation of abnormal cases
D) Five point numerical ranking
Answer:
A) Knowledge of the ground truth
B) A compilation of normal cases
C) A compilation of abnormal cases
Performance-based evaluations: This approach is perhaps the most accurate and relevant approach towards image quality evaluation. Clinical cases are presented to expert clinicians who are otherwise unaware of the case diagnosis in advance. In the most common implementation, the clinician makes a diagnosis for each case according to a numerical ranking (e.g., 1, no lesion is present; 3, uncertain; 5, a lesion is definitely present). The observer scores are then analyzed to assess the statistical likelihood for the clinician to identify an abnormal case (termed sensitivity), or to correctly identify a normal case (termed specificity). The most common analysis technique is the Receiver Operating Characteristic (ROC), which is discussed further in another module (Figure 8). Image quality evaluation through ROC analysis is considered the gold-standard for assessing image quality and diagnostic accuracy of imaging systems and is the basis of many clinical trials as well as for regulatory compliance.
Apparently the question writer did not consider the 5 point system to be a "main requirement"
9.
Which one of these evaluations are considered gold standard for clinical trials and regulatory compliance
A) Alternative forced choice evaluations
B) Receiver operating characteristic evaluations
C) Detective quantum efficiency evaluations
D) Modulation transfer function evaluations
Answer:
B) Receiver operating characteristic evaluations
10.
What is the Nyquist frequency for an imaging system with a pixel size of 0.2 mm?
A) 2.5 cycles/mm
B) 5 cycles/mm
C) 25 cycles/mm
D) 0.5 cycles/mm
Answer:
A) 2.5 cycles/mm
In terms of spatial sampling, no imaging detail can be represented beyond a cutoff frequency equal to ½ of the sampling frequency of the image. In other words, at least two measurements are necessary to accurately depict any one cycle of oscillation. For example, if an image has a pixel size of 0.2 mm (or a sampling frequency of 1 pixel per 0.2 mm = 5 mm-1), the image cannot represent spatial frequencies larger than 2.5 mm-1. This cutoff frequency is also known as the Nyquist frequency (NF):
In this equation, p is the pixel size. Image information that is being depicted by the imaging system can obviously have fine details at spatial frequencies beyond the Nuquist limit of the imaging system. As the imaging system cannot properly reflect those details, they get distorted in the final image in a process known as aliasing.
11.
Aliasing refers to an artifact
A) when low special frequency details get displayed at higher spatial frequencies
B) when high special frequency details get displayed at lower spatial frequencies
C) when pixels get distorted by detector element misalignments
D) when x-rays get scattered within the detector
Answer:
B) when high special frequency details get displayed at lower spatial frequencies
12.
Bit depth affects (select all that apply)
A) Contrast of fine structures in the image
B) Resolution of fine structures in the image
C) Image size
D) Image uniformity
Answer:
A) Contrast of fine structures in the image
For a given matrix, increased bit depth would increase the storage requirement for the image, so I would have considered checking off image size as well.
One could argue that image uniformity (or at least ability to see image uniformity) could be affected by bit depth as well. In my opinion this question would have been best as single best answer.
13.
Scattered radiation causes except (select all that apply)
A) Loss of contrast
B) Loss of CNR
C) Loss of sharpness
Answer:
C) Loss of sharpness
Scattered radiation reduces subject contrast and impacts the level of noise within the image. The impact varies for different systems. Furthermore, from an imaging system perspective, the impact of scattered radiation is dependent on the specifics of the methods used to reduce scattered radiation, such as slot-scanning, air gaps, or anti-scatter grids. Thus, an overall performance assessment of a digital radiographic detector in terms of scatter sensitivity should be made considering the whole acquisition system, including the detector and the anti-scatter element.
The wording of the question is confusing, should be:
Scattered radiation causes all except (select all that apply)
One could also argue that with enough scatter, your ability to detect sharpness is affected. Not a great question.
14.
Acceptance testing and quality control processes are essential to ensure (select all that apply)
A) Highest image quality at lowest dose
B) Consistent imaging performance
C) Safe and efficient operation of the imaging system
D) Proper installation of the imaging system
Answer:
A) Highest image quality at lowest dose
B) Consistent imaging performance
C) Safe and efficient operation of the imaging system
D) Proper installation of the imaging system
It is imperative that an imaging system be tested upon installation through the process of acceptance testing by a qualified medical physicist to ensure the system delivers the desired image quality performance and dose response. Such evaluations can ensure that that system is “safe” for clinical use.
Even though an imaging system may behave perfectly at the time of acceptance testing, its performance may vary over time. Variations might be due to normal deteriorations (e.g., mechanical changes or changes in the scanning or erasure components in CR), or susceptibility to environmental conditions such as temperature, if the temperature of the electronic components is not well regulated. Thus, there is a need to assure that a system performance is stable over time and has not deteriorated beyond a point that might negatively impact its clinical utility or patient dose. To do so, a quality control program may be implemented within which the system characteristics most susceptible to change are tracked over time. Many manufacturers offer QC tools with their products that may be used for this purpose. However, as the QC testing methods might be simpler and more limited than those used in more comprehensive evaluations, it is important to perform the QC tests at the same time as the comprehensive evaluations in order to establish benchmark performance levels for tracking over time.