| 1 |
What is the primary goal of the article according to its introduction?
|
To explore advancements, applications, and challenges of generative AI in medical imaging |
|
The article mainly discusses the development of artificial intelligence in medical imaging.
|
The article states that AI has the potential to aid medical imaging.
|
7 |
-.50
-.25
+.25
เต็ม
0
-35%
+30%
+35%
|
| 2 |
How do generative AI models differ from traditional discriminative models in healthcare applications?
|
Generative models produce new data rather than only classify or interpret |
|
Generative AI can create new content, while traditional models can only sort and understand existing data.
|
Generative AI is capable of creating new content, different from traditional discriminative models that can't.
|
7 |
-.50
-.25
+.25
เต็ม
0
-35%
+30%
+35%
|
| 3 |
What is meant by the term “model as a dataset”?
|
Sharing trained model weights instead of raw data |
|
Source data are stored in the weights of the Generative AI. During data sharing, synthetic data/model weights (source data that has been altered and learned by the generative AI) will be shared instead of the source data (raw data).
|
Model as a dataset refers to a new concept in data sharing, where generative AI learn patterns of original data and stores it in its weights. Sharing model weights allows compressed synthetic data to be shared, which closely resembles its original data.
|
7 |
-.50
-.25
+.25
เต็ม
0
-35%
+30%
+35%
|
| 4 |
Which statement correctly distinguishes physics-informed and statistical models?
|
Physics-informed models incorporate biological or physical principles |
|
Physics-informed models are rule-based, meaning that it learns from existing laws found by experts e.g., tissue biomechanics.
|
Physics-informed models learn from existing laws, meaning factual ideas. Statistical models learn from data patterns, not from rules.
|
7 |
-.50
-.25
+.25
เต็ม
0
-35%
+30%
+35%
|
| 5 |
According to the article, what does the “image generation trilemma” describe?
|
Trade-offs among image diversity, quality, and speed |
|
Different models have different mainly 2 advantages to be traded off for 1 disadvantage. For example, DDPMs are capable of generating high-quality and diverse data, but they have a low generation speed.
|
Each model has 3 measurements. Speed, quality, and image diversity. Each model can have 2 of these, which would be its strong points, and 1 that would be its weak point. For example, VAEs have good diversity and speed of data generation, but are held back its image quality.
|
7 |
-.50
-.25
+.25
เต็ม
0
-35%
+30%
+35%
|
| 6 |
What is the Human Turing Test used for in medical image synthesis?
|
To assess realism of synthetic medical images by experts |
|
The Human Turing test involves medical experts recognizing the differences between real data and synthetic data.
|
The test uses experts to identify real and synthetic data, which is important to assess the quality of synthetic data to be used
|
7 |
-.50
-.25
+.25
เต็ม
0
-35%
+30%
+35%
|
| 7 |
Which of the following is NOT mentioned as a potential benefit of synthetic data in healthcare?
|
Eliminating all medical biases permanently |
|
Synthetic data in healthcare cannot eliminate all medical biases permanently
|
Models are still developing, and as they learn from multiple data sources, it wouldn't be 100% accurate during data generation. This can leave multiple biases and limitations, as generative AI cannot recognize its mistakes.
|
7 |
-.50
-.25
+.25
เต็ม
0
-35%
+30%
+35%
|
| 8 |
What is one major ethical concern associated with generative AI in medical imaging?
|
Data copying and patient reidentification |
|
The model might accidentally reveal sensitive patient data, raising concerns for the patient's safety.
|
Synthetic datasets may allow reidentification of patients, for example, through facial features in brain MRIs.
|
7 |
-.50
-.25
+.25
เต็ม
0
-35%
+30%
+35%
|
| 9 |
What regulatory precedent did the article cite for synthetic data technologies?
|
FDA clearance of synthetic MRI as image-processing software |
|
FDA creates frameworks for assessing synthetic medical imaging
|
FDA removing synthetic MRI allows reducing ethical issues.
|
7 |
-.50
-.25
+.25
เต็ม
0
-35%
+30%
+35%
|
| 10 |
What is the main purpose of the article?
|
To compare and evaluate ASCVD risk prediction models in East Asia |
|
Eastern countries needs a different ASCVD model compared to western countries.
|
Western risk calculators overestimates risk of ASCVD for east asians.
|
7 |
-.50
-.25
+.25
เต็ม
0
-35%
+30%
+35%
|
| 11 |
Which of the following models was originally developed for a Western population?
|
Framingham Risk Score |
|
It is the only risk model developed by the U.S.
|
The other 4 models are developed by China, Japan, and South Korea.
|
7 |
-.50
-.25
+.25
เต็ม
0
-35%
+30%
+35%
|
| 12 |
Why might Western-based risk prediction models overestimate ASCVD risk in East Asian populations?
|
East Asians have lower baseline incidence of ASCVD |
|
East asians has less incidences of ASCVD compared to western-people
|
A lower incidence of ASCVD risk in East Asian population means lowered risk of getting ASCVD. This means that Western-based risk models have a higher chance to overestimate risk of ASCVD in East Asians if East asians were to have the same lifestyle habits as Westerners.
|
7 |
-.50
-.25
+.25
เต็ม
0
-35%
+30%
+35%
|
| 13 |
What is the key advantage of the China-PAR model compared to Western-based models?
|
It was calibrated using national data representing diverse regions in China |
|
It allows for the China-PAR model to be used across the country to assess different ASCVD risks in multiple areas
|
Allows for an average of data to be used across the entire country
|
7 |
-.50
-.25
+.25
เต็ม
0
-35%
+30%
+35%
|
| 14 |
Which of the following variables is not typically included in ASCVD risk prediction models discussed in the article?
|
Genetic ancestry markers |
|
Most prediction models used lifestyle habits rather than genetic markers
|
Genetic ancestry markers were not used in assesing risk of ASCVD, as it is too difficult to assess, especially in East asians who are migrants and live in Western countries.
|
7 |
-.50
-.25
+.25
เต็ม
0
-35%
+30%
+35%
|
| 15 |
What is a major difference between the Suita Score and the Framingham Risk Score?
|
Suita Score is based solely on hospital inpatients |
|
Individuals are screened by experts to be assessed.
|
Internal validation is needed from experts to make an accurate model.
|
7 |
-.50
-.25
+.25
เต็ม
0
-35%
+30%
+35%
|
| 16 |
According to the article, what is a potential benefit of developing East Asia–specific risk models?
|
They improve accuracy and reduce overestimation of risk |
|
An East Asia-specific risk model will allow for ASCVD risk to be assessed more properly in east-asians with different lifestyle habits compared to Westerners.
|
Western risk models overestimate risk of ASCVD in East asians
|
7 |
-.50
-.25
+.25
เต็ม
0
-35%
+30%
+35%
|
| 17 |
Which factor was highlighted as influencing ASCVD risk differences among East Asian countries?
|
Cultural and dietary variations, such as salt intake and lifestyle |
|
ASCVD risk differs throughout East asian countries due to different lifestyle habits
|
Different east asian countries have different food and lifestyles, affecting their cardiovascular health.
|
7 |
-.50
-.25
+.25
เต็ม
0
-35%
+30%
+35%
|
| 18 |
What future direction does the article suggest for improving ASCVD risk prediction?
|
Using multimodal AI-based prediction integrated with regional data |
|
It allows different regions to have their own ASCVD risk model, which would be more accurate as they have different lifestyle habits.
|
AI-Based prediction allows higher efficiency while interpreting regional data to be used to create specific risk models for each regions.
|
7 |
-.50
-.25
+.25
เต็ม
0
-35%
+30%
+35%
|
| 19 |
Which statement best explains the key difference in how VAEs, GANs, and DDPMs generate medical images according to the figure?
|
DDPMs iteratively remove noise through reverse diffusion rather than using encoder–decoder or discriminator structures. |
|
DDPMs reverse noising process, slowly refining the data
|
VAEs and GANs are one way and are much faster than DDPMs.
|
7 |
-.50
-.25
+.25
เต็ม
0
-35%
+30%
+35%
|
| 20 |
Which of the following best explains the trend shown in Figure comparing age-standardized and crude CVD mortality rates among East Asian countries?
|
Japan and South Korea show low age-standardized CVD mortality rates because of smaller populations. |
|
Smaller population means that data gathered would be less reliable
|
Lower population leads to unreliable data.
|
7 |
-.50
-.25
+.25
เต็ม
0
-35%
+30%
+35%
|