1.
Adaptation is the process by which
A) The focusing properties of the eyes are adapted to the scene.
B) The contrast sensitivity changes with brightness.
C) The accommodation and vergence processes are correlated.
Answer:
B) The contrast sensitivity changes with brightness.
2.
In the state of fixed adaptation when a radiologist is examining a medical image, the contrast sensitivity is greatest in
A) bright scenes
B) dark scenes
C) intermediate luminance scenes
Answer:
C) intermediate luminance scenes
3.
Which one of the following four steps of the interpretation process is usually the fastest:
A) Global impression
B) Local attention
C) Recognition
D) Decision making
Answer:
A) Global impression
4.
Visual search is necessitated primarily by the
A) Subtlety of pathological abnormalities
B) Limited angular coverage of the foveal vision
C) Anatomical noise
D) Quantum noise
Answer:
B) Limited angular coverage of the foveal vision
5.
What is the time frame associated with foveal fixation believe to be necessary to recognize an image feature:
A) 0.1 sec
B) 0.3 sec
C) 1 sec
D) 3 sec
Answer:
B) 0.3 sec
6.
Interpretation accuracy increases with the number of cases read per year in a
A) Linear fashion
B) Quadratic fashion
C) Logarithmic fashion
D) Exponential fashion
Answer:
C) Logarithmic fashion
7.
Generally speaking, about half of the interpretation errors is due to lack of visual recognition, while the other half is due to cognitive recognition, when an abnormality is seen but dismissed
A) True
B) False
Answer:
A) True
8.
Satisfaction of search is the process by which
A) Visual search leads to a satisfactory recognition and an abnormality
B) Visual search is prematurely discontinued since no abnormality is found.
C) An abnormality is overlooked because of the presence of another abnormality in the image.
D) An abnormality is overlooked because the search process is confused by the anatomical clutter.
Answer:
C) An abnormality is overlooked because of the presence of another abnormality in the image.
9.
Image quality is affected by all the following except
A) Image resolution
B) Image noise
C) Image presentation
D) Image memory allocation
E) Anatomical noise
Answer:
D) Image memory allocation
10.
What is the FDA approved role of a mammography CADe device?
A) first reader
B) second reader
C) concurrent reader
D) the radiologist may choose how to use the CAD device
Answer:
B) second reader
11.
Which performance metrics depend on disease prevalence?
A) sensitivity and specificity
B) PPV and NPV
C) accuracy only
D) PPV, NPV, accuracy
E) all (sensitivity, specificity, PPV, NPV, accuracy)
Answer:
D) PPV, NPV, accuracy
12.
The standalone performance of a CAD device depends on:
A) the difficulty level of the test cases
B) the experience level of the CAD users
C) both a and b
Answer:
A) the difficulty level of the test cases
13.
A study reported that the standalone performance of a mammography CADe device was shown to be 92% case-based sensitivity. What does that mean?
A) The system detected on 92% of all cancerous lesions in at least one mammographic view.
B) The system detected on 92% of all cancerous lesions in both mammographic views.
C) The system detected 92% of every type of breast malignancy (calcifications, masses, architectural distortions, etc.) in at least one mammographic view.
D) The system detected 92% of every type of breast malignancy (calcifications, masses, architectural distortions, etc.) in both mammographic views.
E) Radiologists using the CAD system detected 92% of of every type of breast malignancy (calcifications, masses, architectural distortions, etc.) in at least one mammographic view.
F) Radiologists using the CAD system detected 92% of of every type of breast malignancy (calcifications, masses, architectural distortions, etc.) in both views.
Answer:
A) The system detected on 92% of all cancerous lesions in at least one mammographic view.
14.
If a CADe system has 80% sensitivity at 0.4 false positives per case and it is developed to detect a disease that has 0.5% prevalence, what is the ratio of TP to FP cues generated by the system?
A) 1:5
B) 1:10
C) 1:100
D) 1:1000
Answer:
C) 1:100
Consider 1000 pts (cases)
TP would equal (0.005)(1000)(0.8) or 4
FP would be (1000)(0.4) or 400
15.
The ROC area index of 3 different CAD systems developed to detect the same disease sign are: System A=0.90, System B=0.80, and System C=0.70. Order the systems from best to worst in terms of their sensitivity.
A) C,B,A
B) A,B,C
C) A,C,B
D) not enough information to answer the question
Answer:
D) not enough information to answer the question
16.
What is the recommended action if a radiologist perceives an abnormality that is not marked by a CADe system?
A) there is no recommended action
B) the radiologist should trust the CADe opinion
C) the radiologist should disregard the CADe opinion
Answer:
C) the radiologist should disregard the CADe opinion
17.
A woman discovers a breast lump and gets a mammogram. The prevalence of breast cancer is 25%. The sensitivity of mammography is 60% and the specificity of mammography is 80%. If the mammogram is positive, what is the probability that the lump is cancerous?
A) 60%
B) 85.7%
C) 50%
D) None of the above
Answer:
C) 50%
Specificity is 80%
so if prevalence is 25%, for 1000 pts.
0.8= 600 true negatives/750 pts without disease
or 150 cases are false positives
0.60= 150 true positives/250 pts with disease
or 100 cases are false negatives
So if test is positive, 150 times it is true positive, 150 times it is false positive, or 50% chance that positive study is cancer.
18.
If the decision threshold of a CAD system is changed so that the system becomes more conservative (thus overcalling cases), how is its performance affected?
A) Specificity increases but sensitivity decreases
B) Specificity decreases but sensitivity increases
C) Both specificity and sensitivity increase
D) Both specificity and sensitivity decrease
Answer:
B) Specificity decreases but sensitivity increases