| 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 |
|
they mentioned the directions of how to adapt and improve AI in medical field, as well as, focusing on advancements and challenges of Ai usage.
|
If we improve the quality, secured, and detailed AI in the future, it will definitely help hospitals and medical fields much more due to it's unstoppable and beneficial ability to help out with work. As it can produce higher-quality images and remove noises for clearer diagnosis or inspections.
|
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 interpret data rather than create it |
|
Based on what they mention about chatGPT, where there is a large language model. It interpret data instead of just generating it.
|
It gains substantial public attention after chatGPT is introduced.
|
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 |
|
it emphasizes the functional view of synthetic data, focusing on its complex scientific challenges than simply put out the original data itself.
|
It contain patterns as well as characteristics of the original data in their model weights
|
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 |
|
It is rule based and need incorporate knowledge and physic principles through math equations and create realistic data according to it.
|
Examples of math equation based on physics principles are fluid dynamics or radiation physics, in order to improve image accurateness.
|
7 |
-.50
-.25
+.25
เต็ม
0
-35%
+30%
+35%
|
| 5 |
According to the article, what does the “image generation trilemma” describe?
|
Balancing accuracy, ethics, and regulation |
|
They mentioned that it involves balancing high sample quality, comprehensive mode coverage, and rapid sampling rate
|
Balancing high quality sample means to create images as accurate as possible. Comprehensive coverage is to make sure it is true to ethics. As well as, rapid sampling rate is to constantly testing and improve to avoid errors in the future.
|
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 |
|
Human Turing Test require medical experts to discern between real and derived images to assure that the image is reliable.
|
To assure realism of the synthetic medical scans or images, it is need to be done by individuals with expertise.
|
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?
|
Facilitating multi-centre collaborations |
|
They want to avoid future conflicts by avoiding bias and enhance the datasets by generating additional synthetic images to allow researchers to create diversity. Patient privacy will be protected as much as possible since they do not want any of the patient's information to be out in the cloud by AI. Finally, they want to support medical educations for further improvement of the medical field.
|
However, they did not mention any collaboration between centers. Example for data augmentation is many diseases are rare, and they have limited examples available for AI training and by using AI to create diversity of dataset, it will be much more accurate and easy for the doctor and expertise.
|
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 |
|
Privacy is one of the ethic concerns that they have mentioned. AI generated images can be accurate but sometimes it is not original due to it's past data that they might have remembered.
|
AI could possibly memorize aspects of real patient data and reproduce them as output. That raises ethical concerns about privacy of the patient.
|
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 |
|
those technologies were marked as image processing software than just novel modalities and with the FDA requiring complex clinical validation to show that the diagnostic performance of the radiologist is accurate and reliable when using synthetic images versus conventional images.
|
it ensures a pathway for future synthetic data technologies with some surveillance check.
|
7 |
-.50
-.25
+.25
เต็ม
0
-35%
+30%
+35%
|
| 10 |
What is the main purpose of the article?
|
To introduce new diagnostic imaging technologies |
|
The central purpose of the paper is to improve Ai to make it more reliable and unlock fit's full potential in order to improve overall healthcare performance.
|
They believe that improved version of AI will help improving patient care, advancing scientific discovery, and transforming the landscape of medical imaging.
|
7 |
-.50
-.25
+.25
เต็ม
0
-35%
+30%
+35%
|
| 11 |
Which of the following models was originally developed for a Western population?
|
Korean Risk Prediction Model (KRPM) |
|
|
|
7 |
-.50
-.25
+.25
เต็ม
0
-35%
+30%
+35%
|
| 12 |
Why might Western-based risk prediction models overestimate ASCVD risk in East Asian populations?
|
Data collection standards are weaker in Asia |
|
Westerns have more variety of patients
|
Different lifestyle causes different predictions. And the number of East Asian immigrants living in the US is growing.
|
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 developed from European clinical trials |
|
Western have more diversity in terms of lifestyle, people, and patients. Additionally, they conducted more clinical trials
|
China's model was developed partly from the western models
|
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 |
|
genetic ancestry markers are for more accurate to be measured because it is not typically used or involve in risk prediction models
|
age, sex, blood pressure, cholesterol, and smoking status are all vital for ACVD risk predictions
|
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 was designed for a Japanese population using local epidemiological data |
|
Framingham model predict lifetime risk instead of 10 years like Suita Score.
|
They both have similar risk factors to predict but Suita is mainly for Japanese population
|
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 |
|
since there are increasing number immigrants in East Asia it is harder to measure accurately
|
Tracking the exact numbers is very complex due to the many conflicts
|
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 |
|
As they mentioned, salt intake and lifestyle is very crucial for ASCVD risk in patients
|
It may incorrectly classify patients and could lead to inappropriate treatment decisions later on.
|
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 |
|
They say that future models should improve to be validated and updated as population changes to avoid inaccurate information and treatment
|
Advances in data collection is crucial as well as machine learning
|
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. |
|
DDPM cerate data by removing noise of the images. The model starts with a sample from a simple distribution then denoises the sample using learned process. Therefore, if they repeatedly apply this process, DDPMs can produce high-quality samples
|
DDPMs are use to denoising diffusion models. GANs are use to generate networks. While VAEs are use to autoencoders diversity
|
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?
|
South Korea’s high stroke rate implies poor control of infectious diseases rather than cardiovascular conditions. |
|
stroke events in the region are primarily driven by vascular risk factors
|
|
7 |
-.50
-.25
+.25
เต็ม
0
-35%
+30%
+35%
|