| 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 |
|
Stated from the head of the article, and from the contents, the article fully explains all four of the mentioned subjects in the choice.
|
The first few articles and the introduction explain potential and uprising of Chat GPT, and other AI language models. Adding that they will rapidly improve incrementally, and have been showing promising work. Including the challenges, as mentioned in the back of the introduction.
|
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 |
|
Probability density estimation of ai models differ due to its descriptive, prescriptive machine learning.
|
In contrast to modern model AI or statistical models learn from data patterns and distributions and among them variational autoencoders derive texts into a less dimensional data which is also known as latent space, and then reconstructing the data, therefore effective capturing the data. (from Synthetic datasets, Generative models)
|
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 |
|
generative models like GAN or Diffusion model acts like a knowledge dense cube of thought for the original sensitive one
|
The concept redefines (quote on quote) data as a functional model because instead of transferring the sensitive dicom files, an institution shares the parameters of a diffusion model. Therefore acts as a proxy allowing the recipient to generate an infinite cohort of synthetic patients locally. (from Synthetic datasets, Generative models)
|
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 |
|
Unlike AI statistical models that rely on pattern recognition the Phys-informed models, they are by hard constraints; the laws of electromagnetism in MRI or fluid dynamics in ultrasound to ensure the generated images are physiologically plausible.
|
(from Synthetic datasets, Generative models), Physics-informed models offer high fidelity and interpretability but might require extensive domain expertise and computational resources. Again, In contrast to physics-informed models, statistical models learn from data patterns and distributions. A generator creates data samples and a discriminator evaluates these data samples and provides feedback to the generator.
|
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 |
|
The trilemma suggests that a single architecture rarely excels in all three dimensions, simultaneously.
|
The Diffusion models often provide high quality and diversity but suffer from slow sampling speeds, whereas GANs offer speed but may struggle with diversity, leading to low mode coverage, known as mode collapse. The image generation trilemma describes the real hard engineering challenge of balancing high sample quality, high mode coverage, and fast sampling speed within a single generative framework. (from Synthetic datasets, Generative models, figure 1)
|
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 |
|
Its a subjective evaluation protocol requires experienced clinicians differentiate between real and synthetic scans in a blinded trial.
|
(in Human Evaluation), The Human Turing Test is and will remain a standard for evaluating the fidelity of synthetic medical images, and it utilizes diagnostic expertise of clinicians to 'validate' whether the seemingly synthesized data is indistinguishable from real clinical scans. Aa wide range of participants with different experience levels is required to involve in the image evaluation process.
|
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 can make less of a bias by starting underrepresented groups, (can't "permanently eliminate" it.)
|
If the training data contains systemic biases, the generative model may inadvertently learn and propagate those biases into the synthetic output. So it's not a panacea for bias; researchers must remain vigilant for the preposterous, eccentric risk of the model replacing or changing the existing clinical disparities. (in Use Cases in Medical Imaging)
|
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 |
|
Although synthetic datasets can help to preserve patient privacy by generating non personal data the concerns regarding potential data copying still exist.
|
A primary ethical risk is memorization in which the model overfits to the training set and produces synthetic images that are nearly identical to real patients. This, could potentially lead to a breach of privacy if a generated image reveals a unique anatomical signature of a real individual, quite plausibly as well. Concerns include data leakage of patients and the mentioned memorization. (Patient privacy and data copying)
|
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 |
|
based on the article it highlights a shift in regulatory acceptance, and noting that the FDA has already been clearing AI's mistakes.
|
Regulatory bodies have established a precedent for these technologies, exemplified by the FDAs clearance of Artificial Intelligence based image processing tools that use generative models to bloom up the raw MRI data. (eg, denoising) (Image super resolution, denoising, and inpainting) (from Subtle Medical’s SubtleHDTM wins FDA clearance, setting a new benchmark for MRI image quality and speed) (and from Future directions)
|
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 |
|
The study evaluates the performance and calibration of various risk scores within East Asian cohorts to determine which models provide the most accurate clinical guidance for local populations.
|
The primary objective of this research is to evaluate and compare the predictive accuracy of various ASCVD risk models in both Western and region specific within the unique demographic landscape of the East Asia.
|
7 |
-.50
-.25
+.25
เต็ม
0
-35%
+30%
+35%
|
| 11 |
Which of the following models was originally developed for a Western population?
|
|
|
|
|
7 |
-.50
-.25
+.25
เต็ม
0
-35%
+30%
+35%
|
| 12 |
Why might Western-based risk prediction models overestimate ASCVD risk in East Asian populations?
|
|
|
|
|
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?
|
|
|
|
|
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?
|
|
|
|
|
7 |
-.50
-.25
+.25
เต็ม
0
-35%
+30%
+35%
|
| 15 |
What is a major difference between the Suita Score and the Framingham Risk Score?
|
|
|
|
|
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?
|
|
|
|
|
7 |
-.50
-.25
+.25
เต็ม
0
-35%
+30%
+35%
|
| 17 |
Which factor was highlighted as influencing ASCVD risk differences among East Asian countries?
|
|
|
|
|
7 |
-.50
-.25
+.25
เต็ม
0
-35%
+30%
+35%
|
| 18 |
What future direction does the article suggest for improving ASCVD risk prediction?
|
|
|
|
|
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?
|
|
|
|
|
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?
|
|
|
|
|
7 |
-.50
-.25
+.25
เต็ม
0
-35%
+30%
+35%
|