ตรวจข้อสอบ > ฐิติพรรณ เผ่านิ่มมงคล > KOREA | Medical & Health Sciences (Secondary Level) | สาขาการแพทย์และสุขภาพ ระดับมัธยมศึกษา > Part 2 > ตรวจ

ใช้เวลาสอบ 22 นาที

Back

# คำถาม คำตอบ ถูก / ผิด สาเหตุ/ขยายความ ทฤษฎีหลักคิด/อ้างอิงในการตอบ คะแนนเต็ม ให้คะแนน
1


How does the concept of “model as a dataset” reshape traditional data-sharing practices in medical imaging?

It enables sharing of learned model weights instead of sensitive raw images.

By using the theory down below, the patients raw data will be safely kept in local storage but only weights and parameters will be the only thing Ai will learn. The theory behind this is the privacy -preserving machine Learning. From the Lancet Digital Health. 7

-.50 -.25 +.25 เต็ม 0 -35% +30% +35%

2


Which analytical conclusion can be drawn about the trade-offs between physics-informed and statistical models?

Physics-informed models are more interpretable but computationally intensive.

By using the physiscs informed models in physical test in a base learning it can give us a plysically plausible explaination. The theory behind this in the Interpretability complexity trade off. 7

-.50 -.25 +.25 เต็ม 0 -35% +30% +35%

3


Why is “mode collapse” considered a critical problem in GAN-based medical image synthesis?

It reduces image realism and variety by producing repetitive outputs.

From the theory, instead of learning the full complexity of the training data, the Generator starts producing that one successful image over and over again. Using the Nash Equilibrium and optimization instability to test the failure of Convergence. 7

-.50 -.25 +.25 เต็ม 0 -35% +30% +35%

4


Why are healthcare-specific metrics preferred over general-purpose metrics such as FID or SSIM?

They better capture clinical accuracy and diagnostic relevance.

Because it help distinguish between the beauty of the image and the medical information accuracy. Use the theory of Domain-specific Validation vs. Perceptual similarity by The Lancet Digital Health. 7

-.50 -.25 +.25 เต็ม 0 -35% +30% +35%

5


What does the article identify as the key tension between privacy preservation and image fidelity?

Higher realism may risk reproducing identifiable patient data.

By increasing the model fidelity often leads to the memorization of unique patient markers, enabling potential re-identification of training data Using the theory, privacy-utility trade-off. From the Lancet digital Health. 7

-.50 -.25 +.25 เต็ม 0 -35% +30% +35%

6


Why is the FDA’s approval of synthetic MRI technology significant for future AI-generated data?

It establishes a framework for validating synthetic data equivalence in clinical use.

It a theory confirm that All the information create AI, it's are quality are equal to the real information. Using the theory if regulatory precedent and validation framework. 7

-.50 -.25 +.25 เต็ม 0 -35% +30% +35%

7


Which strategy would best mitigate demographic bias in generative models according to the article?

Applying diversity-aware training and fairness constraints

By involving implementing specific fairness metrics and re-weighting techniques during training, helps prevent AI from perpetuating existing disparities. Use the theory of Algorithmic Fairness Theory and Debiasing thoery. 7

-.50 -.25 +.25 เต็ม 0 -35% +30% +35%

8


How do DDPMs exemplify versatility in healthcare image synthesis?

They can perform multiple tasks such as denoising, inpainting, and anomaly detection without retraining.

because their iterative denoising mechanism acts as a mathematical prior that can be steered to solve various image restoration and analysis problems without modifying the model's core architecture. Using the theory Stochastic Refinement and Zero-shot Capability. 7

-.50 -.25 +.25 เต็ม 0 -35% +30% +35%

9


What analytical insight does the article provide about integrating AI-generated medical images into education and research?

It enhances training by providing diverse, realistic datasets without ethical breaches.

This approach solves the data problem in training, offering high-fidelity, privacy-preserving simulations for diagnostic skill development. Simulation-based learning theory and data privacy protection theory. 7

-.50 -.25 +.25 เต็ม 0 -35% +30% +35%

10


Why is regional calibration essential when applying risk prediction models across countries?

To adjust for population-specific incidence and lifestyle differences

Since cardiovascular risk models developed in one region often fail in another due to variations in baseline disease rates, genetics, and cultural habits , requiring recalibration to ensure clinical accuracy for the local population. Baseline risk and model recalibration 7

-.50 -.25 +.25 เต็ม 0 -35% +30% +35%

11


What analytical conclusion can be drawn when comparing the China-PAR and Framingham models?

China-PAR uses local epidemiological data, leading to improved predictive validity.

it utilizes local epidemiological data, correcting for overestimation of risk. The principle of population-specific validity. 7

-.50 -.25 +.25 เต็ม 0 -35% +30% +35%

12


Based on CVD mortality data, what analytical inference can be made about Japan’s position compared to neighboring countries?

Japan’s mortality reflects poor access to screening.

because statistical data from major epidemiological studies shows that Japan maintains some of the lowest age-standardized CVD mortality rates globally. Analytical Inference theory. 7

-.50 -.25 +.25 เต็ม 0 -35% +30% +35%

13


What analytical limitation arises when using Western-derived coefficients in East Asian models?

It introduces systematic overestimation of ASCVD probability.

because Western-derived coefficients are calculated based on high-incidence populations, causing them to project an inaccurately high level of risk when applied to East Asian who have lower baseline disease rates. Calibration drift and baseline hazard misalignment. 7

-.50 -.25 +.25 เต็ม 0 -35% +30% +35%

14


What policy implication can be derived from country-specific risk models?

They allow for targeted national prevention programs.

because country-specific models provide accurate data that enables governments to design precise public health interventions and allocate resources efficiently based on the unique risk profile of their local population. Use Resource Allocation and precision public health theory. 7

-.50 -.25 +.25 เต็ม 0 -35% +30% +35%

15


If a model excludes socioeconomic variables, what analytical consequence might occur?

Improved accuracy

If a model exclude social economic variables. The accuray can be improved. Using the theory of the social determinants of health 7

-.50 -.25 +.25 เต็ม 0 -35% +30% +35%

16


How might AI improve next-generation ASCVD risk prediction in East Asia?

By removing human oversight in risk assessment

Removing humans oversight error. It can improve the next generation AsCVD risk. Data fusion theory 7

-.50 -.25 +.25 เต็ม 0 -35% +30% +35%

17


What conclusion can be drawn from comparing Mongolia’s and South Korea’s CVD mortality rates?

Mortality differences reflect varying effectiveness of national prevention programs.

By reflecting on the diversith. It can help discover a new a Way to prevent a risk. Health system and risk factor comtrol 7

-.50 -.25 +.25 เต็ม 0 -35% +30% +35%

18


What is the most logical future direction for improving ASCVD models across East Asia?

Establishing multinational data-sharing platforms to harmonize regional models

It can help share insightful data with each other Collaborative Data Intelligence 7

-.50 -.25 +.25 เต็ม 0 -35% +30% +35%

19


According to the “image generation trilemma” shown in the figure, what analytical conclusion can be drawn about the relative strengths of VAEs, GANs, and DDPMs in medical image synthesis?

GANs provide a balance between image quality and diversity but may suffer from mode collapse.

It highlights the inherent trade-offs, where GANs prioritize high-fidelity outputs but struggle with maintaining full distribution coverage. The Generative learning 7

-.50 -.25 +.25 เต็ม 0 -35% +30% +35%

20


Based on Figure, what analytical conclusion can be drawn regarding the distribution of cardiovascular disease (CVD) subtypes across East Asian countries?

The proportion of IHD and stroke deaths is uniform across all regions of East Asia.

7

-.50 -.25 +.25 เต็ม 0 -35% +30% +35%

ผลคะแนน 112.75 เต็ม 140

แท๊ก หลักคิด
แท๊ก อธิบาย
แท๊ก ภาษา