引導式測驗 · 中等 等級

ONC HTI-1演算法透明度規則測驗

測試 HTI-1 範圍、經過認證的健康 IT、預測決策支援、來源屬性以及與 FDA 授權的差異。

問題 1 的 8

What does the HTI-1 algorithm transparency rule cover?

預覽每個問題8 問題

當您參加測驗時,就會出現答案和解釋。

  1. What does the HTI-1 algorithm transparency rule cover?
    • Only FDA-cleared imaging software
    • All consumer health apps automatically
    • Predictive decision support interventions within ONC-certified health IT scope
    • Every AI product used in a hospital
  2. What can required source attributes help users assess?
    • Whether FDA has approved every algorithm
    • Fairness, appropriateness, validity, effectiveness, and safety
    • Whether a vendor is profitable
    • The patient’s diagnosis with certainty
  3. Does HTI-1 transparency prove a model is safe?
    • Yes, because attributes are disclosed
    • Yes, if the tool is certified
    • No, transparency has no use
    • No; disclosure supports evaluation but does not establish performance
  4. For an AI-enabled health IT function, who should identify whether the function falls within certified-health-IT scope?
    • Developers and users should identify whether the function is in certified health IT scope
    • The certification body alone, based only on the product’s marketing category
    • The vendor, based on whether its website calls the product “AI”
    • The individual patient, by deciding whether the feature influenced care
  5. How does HTI-1 relate to FDA authorization?
    • It applies only after FDA denies a device
    • It replaces privacy law
    • It does not replace FDA device authorization or other obligations
    • It automatically grants FDA clearance
  6. What should a health system do with source-attribute disclosures?
    • Use them as one input to local risk assessment and validation
    • Assume they prove local performance
    • Ignore data quality and workflow
    • Use them to bypass clinician review
  7. When did the rule take effect according to ONC materials?
    • It has no effective date
    • March 2024
    • March 2020
    • August 2026
  8. Why map each predictive intervention to criteria?
    • All algorithms use the same attributes
    • Every software feature is a medical device
    • A model’s architecture determines certification
    • Different functions and products may have different obligations