Computer-Aided Detection and Diagnosis — MCQs
Digital radiography differs from conventional in
PACS in medical imaging stands for:
Best imaging modality for acute pulmonary embolism
In a screening test for DM out of 1000 population, 90 were positive. When the gold standard test was applied to the entire population, 100 were found to have the disease. Assuming all 90 screening positives were confirmed as true positives by the gold standard, calculate the sensitivity.
HIV RNA by PCR can detect as low as
A research team develops an AI algorithm using 100,000 CT scans from multiple institutions. The algorithm shows excellent performance (AUC 0.96) but requires extensive computational resources. To deploy it in resource-limited settings, they propose model compression techniques. Evaluate the potential trade-offs and propose the most balanced approach.
A radiology department is evaluating two AI algorithms for fracture detection. Algorithm A has AUC-ROC of 0.95, while Algorithm B has AUC-ROC of 0.92 but provides explainable results showing which image regions influenced its decision. Considering clinical implementation and medicolegal aspects, which statement best evaluates the choice?
A deep learning algorithm for detecting pneumonia on chest X-rays performs excellently on the validation set but poorly on external testing. Analysis reveals the algorithm learned to recognize the hospital logo and text on images from ICU patients (who more likely had pneumonia). What type of bias does this represent?
An AI model for detecting breast cancer on mammography shows sensitivity of 95% and specificity of 85% in a screening population with 1% disease prevalence. A study claims the AI outperforms radiologists who have 90% sensitivity and 90% specificity. Analyze why this comparison may be misleading.
A hospital implements an AI algorithm for detecting intracranial hemorrhage on CT scans. The algorithm was trained on data from a different population with different CT scanner protocols. The algorithm shows decreased performance. Which concept explains this phenomenon?
Computer-Aided Detection and Diagnosis Indian Medical PG Practice Questions and MCQs
Question 1: Digital radiography differs from conventional in
- A. X-rays are not required for imaging
- B. Images cannot be printed
- C. Radiation receptors are different (Correct Answer)
- D. Uses radiation other than X-rays
Explanation: ***Radiation receptors are different*** - Digital radiography uses **digital sensors** (e.g., CCD, CMOS, flat panel detectors) or **photostimulable phosphor plates** (PSP) to capture the X-ray image directly, unlike conventional radiography which uses film. - This fundamental difference in **receptor technology** allows for immediate image display, digital storage, and post-processing capabilities. *X-rays are not required for imaging* - Digital radiography is still a form of **X-ray imaging**; it uses X-rays to penetrate the body and create an image. - The difference lies in how these X-rays are **detected and processed**, not in their absence. *Images cannot be printed* - Digital images can be easily **printed** if desired, although they are primarily viewed and stored digitally. - The ability to print allows for physical copies, but the main advantage is digital storage and sharing. *Uses radiation other than X-rays* - Digital radiography exclusively uses **X-radiation** to generate images. - Techniques like MRI use radiofrequency waves and magnetic fields, and ultrasound uses sound waves; these are distinct modalities, not digital radiography.
Question 2: PACS in medical imaging stands for:
- A. Picture archiving and communication system (Correct Answer)
- B. Planned archiving computerized system
- C. Planned archiving common system
- D. Picture archiving or computerized system
Explanation: ***Picture archiving and communication system*** is the correct answer. - **PACS** is a widely used technology in medical imaging for the **storage, retrieval, management, distribution, and presentation** of medical images - It replaces traditional film-based systems with a **digital imaging and communications approach** - The system enables seamless sharing of images across departments and healthcare facilities *Planned archiving common system* - Incorrect because the "P" in PACS stands for **Picture**, referring to medical images, not "Planned" - The term emphasizes the digital images being handled, not general planning or common systems *Planned archiving computerized system* - Incorrect as PACS focuses on **Picture** and **Communication** in handling medical images - While the system is computerized, this misses the crucial picture archiving and communication functions *Picture archiving or computerized system* - Incorrect because it uses "or" instead of **"and"**, fundamentally changing the system's function - PACS is designed for both **archiving AND communication** of images, not one or the other
Question 3: Best imaging modality for acute pulmonary embolism
- A. V/Q scan
- B. CT pulmonary angiogram (Correct Answer)
- C. Chest X-ray
- D. MRI
Explanation: ***CT pulmonary angiogram*** - This is the **gold standard** imaging modality for diagnosing acute pulmonary embolism due to its high sensitivity and specificity in visualizing pulmonary arteries. - It rapidly provides detailed images of the pulmonary vasculature, allowing for direct visualization of **thrombi**. *V/Q scan* - A **V/Q scan** measures ventilation and perfusion of the lungs and is less definitive than CTPA, especially in patients with pre-existing lung disease. - It is often considered when **CTPA is contraindicated**, such as in cases of severe renal impairment or contrast allergy. *Chest X-ray* - A **chest X-ray** is generally used to rule out other causes of chest pain and shortness of breath, such as pneumonia or pneumothorax, rather than to diagnose PE directly. - It has **low sensitivity and specificity** for pulmonary embolism, as findings are often non-specific or normal even in the presence of PE. *MRI* - **Magnetic resonance angiography (MRA)** can be used, but it is typically reserved for patients who cannot undergo CTPA or V/Q scan due to contraindications like **pregnancy** or **renal failure**. - It often takes longer to perform and has lower spatial resolution compared to CTPA for pulmonary artery visualization.
Question 4: In a screening test for DM out of 1000 population, 90 were positive. When the gold standard test was applied to the entire population, 100 were found to have the disease. Assuming all 90 screening positives were confirmed as true positives by the gold standard, calculate the sensitivity.
- A. All positives identified by the test assumed as true positives (100%)
- B. True positives divided by total actual positives (90%) (Correct Answer)
- C. Underestimated true positives divided by total actual positives (80%)
- D. Total positives identified by the test divided by total actual positives (90%)
Explanation: ***True positives divided by total actual positives (90%)*** - **Sensitivity** is the proportion of true positives correctly identified by a screening test among all individuals who actually have the disease. It is calculated by (Number of True Positives) / (Total Number of Diseased Individuals). - In this case, 90 people screened positive and were confirmed as **true positives**. The total number of people with the disease (actual positives) is 100. So, sensitivity = 90/100 = **90%**. *Total positives identified by the test divided by total actual positives (90%)* - While this option states the correct percentage (90%), the phrasing "total positives identified by the test" is misleading terminology. In screening test evaluation, this could be confused with all test positives (which would include false positives if they existed). - The correct terminology is "true positives" divided by "total actual positives," not "total positives identified by the test." The distinction is important: true positives are confirmed cases, while test positives might include false positives. *All positives identified by the test assumed as true positives (100%)* - This option incorrectly assumes that because all 90 screening positives were confirmed as true positives, the sensitivity must be 100%. However, sensitivity measures how many of ALL diseased individuals were caught, not just those who screened positive. - There were 100 actual diseased individuals, and only 90 were identified by the screening test; therefore, the sensitivity cannot be 100%. The test missed 10 diseased individuals (false negatives). *Underestimated true positives divided by total actual positives (80%)* - This option presents an arbitrary percentage that does not reflect the given data. There is no information to suggest that the true positives were underestimated or that the calculation would result in 80%. - The actual number of true positives (90) and actual positives (100) directly leads to a sensitivity calculation of 90%, not 80%.
Question 5: HIV RNA by PCR can detect as low as
- A. 50 copies of viral RNA/ml of blood (Correct Answer)
- B. 60 copies of viral RNA/ml of blood
- C. 40 copies of viral RNA/ml of blood
- D. 30 copies of viral RNA/ml of blood
Explanation: ***50 copies of viral RNA/ml of blood*** - **HIV RNA PCR assays** used in clinical practice have a standard detection limit of **50 copies/mL**, which is the widely accepted threshold for defining "undetectable viral load." - This detection limit is used by most major **viral load testing platforms** including Abbott RealTime HIV-1 and Roche COBAS assays. - Achieving viral load **<50 copies/mL** is the goal of antiretroviral therapy (ART) and indicates effective **viral suppression**. - This has been the **standard clinical threshold** used in treatment monitoring and guidelines. *60 copies of viral RNA/ml of blood* - A detection limit of 60 copies/mL is **above the standard threshold** of 50 copies/mL used in clinical practice. - This would be considered less sensitive than conventional **HIV RNA PCR assays**. - Patients with viral loads between 50-60 copies/mL might be misclassified using this threshold. *40 copies of viral RNA/ml of blood* - While some **ultrasensitive assays** can detect down to 20-40 copies/mL, this is not the standard detection limit cited in most medical literature. - The **clinical standard** remains 50 copies/mL for defining undetectable viral load. - Detection at 40 copies/mL represents enhanced sensitivity but is not the commonly referenced threshold. *30 copies of viral RNA/ml of blood* - Some newer generation assays claim detection limits of 20-30 copies/mL, but this is not the **standard clinical threshold**. - The universally accepted detection limit for **HIV viral load testing** is **50 copies/mL**. - While greater sensitivity is theoretically better, 50 copies/mL remains the benchmark for treatment monitoring.
Question 6: A research team develops an AI algorithm using 100,000 CT scans from multiple institutions. The algorithm shows excellent performance (AUC 0.96) but requires extensive computational resources. To deploy it in resource-limited settings, they propose model compression techniques. Evaluate the potential trade-offs and propose the most balanced approach.
- A. Model compression always maintains performance while reducing size
- B. Use knowledge distillation to train a smaller model that mimics the larger model while accepting minimal performance decrease (Correct Answer)
- C. Avoid compression as any performance loss is unacceptable in medical AI
- D. Random pruning of neural network connections is sufficient
Explanation: ***Use knowledge distillation to train a smaller model that mimics the larger model while accepting minimal performance decrease*** - **Knowledge distillation** allows a "student" model to learn the complex features of a "teacher" model, significantly reducing **computational footprint** while preserving high **diagnostic accuracy**. - This approach is the most balanced for **resource-limited settings**, as it optimizes the trade-off between **model size** and the high **AUC** required for clinical safety. *Model compression always maintains performance while reducing size* - This is incorrect because compression techniques like **quantization** or **pruning** often result in some degree of **information loss** or degradation in metric sensitivity. - The goal of compression is to minimize this loss, but it is not a guaranteed consequence of the process. *Avoid compression as any performance loss is unacceptable in medical AI* - While accuracy is critical, failing to compress the model makes it unusable in **edge devices** or areas with low **processing power**, hindering medical access. - Medical AI deployment requires a pragmatic balance between **idealistic performance** and **practical utility** in real-world clinical environments. *Random pruning of neural network connections is sufficient* - **Random pruning** is suboptimal and lacks the strategic precision needed to maintain the **AUC 0.96** performance level required for radiology. - Effective model optimization requires **structured pruning** or **weight-based selection** to ensure critical diagnostic features are not inadvertently deleted.
Question 7: A radiology department is evaluating two AI algorithms for fracture detection. Algorithm A has AUC-ROC of 0.95, while Algorithm B has AUC-ROC of 0.92 but provides explainable results showing which image regions influenced its decision. Considering clinical implementation and medicolegal aspects, which statement best evaluates the choice?
- A. Algorithm A should always be chosen due to superior performance metrics
- B. Algorithm B may be preferred despite lower AUC due to interpretability and accountability (Correct Answer)
- C. AUC-ROC is the only relevant metric for clinical decision making
- D. The difference in AUC is clinically insignificant so both are equivalent
Explanation: ***Algorithm B may be preferred despite lower AUC due to interpretability and accountability*** - **Explainable AI (XAI)** is critical in medicine because it allows clinicians to verify the **reasoning process**, ensuring the algorithm isn't relying on irrelevant artifacts. - High **interpretability** facilitates **medicolegal accountability** and builds trust, which are often prioritized over marginal gains in statistical performance metrics like **AUC-ROC**. *Algorithm A should always be chosen due to superior performance metrics* - Relying solely on **performance metrics** ignores the "black box" problem, where a model may have high accuracy but fail unexpectedly in **real-world clinical scenarios**. - Without **spatial localization** or explanation, clinicians cannot easily distinguish between a true positive and a **spurious correlation** detected by the AI. *AUC-ROC is the only relevant metric for clinical decision making* - **AUC-ROC** measures general discriminatory power but does not account for **clinical utility**, workflow integration, or the safety implications of **false negatives**. - Other metrics such as **Positive Predictive Value (PPV)** and **Explainability** are equally vital for determining if a tool is safe and effective for bedside use. *The difference in AUC is clinically insignificant so both are equivalent* - A difference between **0.95 and 0.92** can be statistically and clinically significant depending on the **prevalence** of the condition and the volume of images processed. - Labeling them as **equivalent** overlooks the qualitative advantage of **explainability**, which fundamentally changes how the radiologist interacts with the software.
Question 8: A deep learning algorithm for detecting pneumonia on chest X-rays performs excellently on the validation set but poorly on external testing. Analysis reveals the algorithm learned to recognize the hospital logo and text on images from ICU patients (who more likely had pneumonia). What type of bias does this represent?
- A. Selection bias
- B. Confounding bias (Correct Answer)
- C. Information bias
- D. Spectrum bias
Explanation: ***Confounding bias*** - In machine learning, this occurs when an algorithm learns a **spurious correlation** between a feature (like a hospital logo) and the outcome (pneumonia) because that feature is non-causally associated with the disease. - The **hospital logo** acts as a **confounding variable** that provides a shortcut for the model, leading to high internal accuracy but poor **generalizability** to external datasets without that logo. *Selection bias* - This involves errors in the **recruitment or retention** of study participants, leading to a sample that does not accurately represent the target population. - While the ICU population represents a specific subset, the core issue here is the algorithm identifying **irrelevant visual markers**, not just the patient selection process. *Information bias* - This refers to errors in how data is **measured, collected, or recorded**, such as recall bias or measurement error. - In this scenario, the images themselves were recorded correctly, but the model's **interpretation logic** was flawed due to external markers rather than an error in the data collection tool. *Spectrum bias* - This occurs when the study population does not reflect the **full range** of disease severity seen in clinical practice, often using only very sick patients and healthy controls. - While using ICU patients could contribute to this, the specific problem of identifying **hospital-specific text or logos** is a hallmark of confounding, not just a narrow disease spectrum.
Question 9: An AI model for detecting breast cancer on mammography shows sensitivity of 95% and specificity of 85% in a screening population with 1% disease prevalence. A study claims the AI outperforms radiologists who have 90% sensitivity and 90% specificity. Analyze why this comparison may be misleading.
- A. The AI has lower positive predictive value despite higher sensitivity (Correct Answer)
- B. The AI has higher negative predictive value in all cases
- C. Specificity is more important than sensitivity in screening
- D. The prevalence is too high for meaningful comparison
Explanation: ***The AI has lower positive predictive value despite higher sensitivity*** - In a low **prevalence** environment (1%), even a small drop in **specificity** leads to a significant increase in **false positives**, which markedly reduces the **Positive Predictive Value (PPV)**. - Despite a sensitivity of 95%, the AI's lower specificity (85% vs 90%) results in more unnecessary follow-up procedures and **recall rates** compared to the radiologist. *The AI has higher negative predictive value in all cases* - While higher sensitivity generally improves **Negative Predictive Value (NPV)**, the NPV is already exceedingly high for both (approx. 99.9%) due to the low **prevalence** of the disease. - A marginal gain in NPV does not necessarily justify a substantial increase in **false alarms** caused by lower specificity. *Specificity is more important than sensitivity in screening* - Neither metric is universally "more important"; the ideal screening tool requires a **balance** to ensure high **sensitivity** (catching cases) without overwhelming the system with **false positives**. - However, in this specific clinical context, the radiologist's higher **specificity** maintains a better diagnostic yield (PPV) than the AI model. *The prevalence is too high for meaningful comparison* - A **prevalence** of 1% is actually typical for **screening mammography** populations; it is not considered too high for statistical analysis. - The comparison is misleading due to the **trade-off** between sensitivity and specificity, not because the prevalence rate is an outlier.
Question 10: A hospital implements an AI algorithm for detecting intracranial hemorrhage on CT scans. The algorithm was trained on data from a different population with different CT scanner protocols. The algorithm shows decreased performance. Which concept explains this phenomenon?
- A. Overfitting of the training data
- B. Dataset shift and lack of generalizability (Correct Answer)
- C. Insufficient neural network layers
- D. Poor image preprocessing
Explanation: ***Dataset shift and lack of generalizability*** - **Dataset shift** occurs when the distribution of data used during training differs significantly from the data encountered in clinical practice, such as different **scanner protocols**. - This leads to a lack of **generalizability**, where the AI performs poorly in new environments because it cannot adapt to variations in **population demographics** or imaging hardware. *Overfitting of the training data* - **Overfitting** happens when a model learns the noise and specific details of the training set too well, failing to predict outcomes on any new data. - While it affects generalizability, the specific issue of switching **scanner protocols** and **populations** is more accurately described as a shift in data domains. *Insufficient neural network layers* - Insufficient layers or **lack of depth** typically results in **underfitting**, where the model is too simple to capture the underlying patterns in the training data. - This is a structural limitation of the model architecture rather than an issue related to the **external validation** or the source of the data. *Poor image preprocessing* - **Preprocessing** involves cleaning or standardizing images before feeding them into the model; errors here would affect consistency across all datasets. - While standardized preprocessing helps mitigate differences, the root cause of decreased performance across different **institutional protocols** is the mismatch in the data distribution itself.
Want unlimited practice?
Get full access to all questions, explanations, and performance tracking.
Start For Free