引導式測驗 · 中等 等級

膠囊內視鏡檢查中的人工智慧測驗

測試膠囊框架分類、臨床審查、閱讀時間證據以及人工智慧輔助內視鏡檢查中漏檢結果的風險。

問題 1 的 8

What does an AI frame-selection system do during capsule review?

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 an AI frame-selection system do during capsule review?
    • Ranks or selects images that may deserve closer inspection
    • Summarizes patient symptoms from a chart without reviewing the video
    • Measures bowel transit time from a single image
    • Assigns a histopathology result to a lesion
  2. Which finding was studied in the 2024 prospective capsule study?
    • It reduced reading time but had a lower lesion yield in that study
    • It proved equal performance for every capsule platform and indication
    • AI-assisted reading was non-inferior for relevant lesions and shorter in that study
    • It was a simulation using no patient recordings
  3. Which false-negative risk justifies reviewing some frames an algorithm deprioritizes?
    • A clear image may receive an unnecessary second look
    • A relevant lesion could be missed among frames the model deprioritized
    • A duplicate frame may appear more than once in a report
    • A visible segment may require a longer display load
  4. What can a poor-visibility filter do?
    • Predict a tissue diagnosis from the capsule battery level
    • Estimate the patient’s symptom severity from images alone
    • Replace bowel preparation with an image filter
    • Identify frames with poor mucosal visibility for workflow handling
  5. Which safety outcome should be checked when a frame-selection system hides images?
    • Missed clinically relevant lesions
    • Average frame count retained, without checking for missed findings
    • Reader preference for a particular capsule brand
    • How quickly the software opens the video
  6. Why can reading-time results vary between capsule studies?
    • Reading time is determined only by capsule battery life
    • The diagnosis has no effect on what the reader reviews
    • Indication, device, reader workflow and bowel preparation differ
    • Every clinic reviews the same number of frames
  7. What should a clinician do with an AI-marked red region?
    • Accept the label as a confirmed diagnosis
    • Inspect the image and decide whether it represents a finding or artifact
    • Use the marker as a substitute for reviewing surrounding frames
    • Write the report from the highlighted frame alone
  8. What does shorter reading time establish by itself?
    • That every lesion was found
    • That further testing is unnecessary
    • That the model is safe for every patient
    • Only that time was shorter under the tested workflow, not that outcomes improve