引导式测验 · 中等 等级

人工智能在心脏病学测验中的应用

使用 ECG 和回波数据、CT 衍生血流分析、群体验证以及可穿戴 ECG 工具的规定局限性来回顾心脏 AI 示例。

相关引导路径人工智能在心脏病学中的应用
问题 1 的 8

根据 Apple 的文档，Apple 手表心电图应用程序会生成什么样的记录？

预览每个问题8 问题

当您参加测验时，就会出现答案和解释。

  1. 根据 Apple 的文档，Apple 手表心电图应用程序会生成什么样的记录？
    • A waveform similar to a single-lead ECG
    • A full coronary CT angiogram
    • A cholesterol measurement from a blood sample
    • A twelve-lead ECG with twelve independent waveforms
  2. 2019 年梅奥诊所的 AI 辅助心电图研究评估了哪项任务？
    • Measuring blood cholesterol without a blood test
    • Performing cardiac surgery autonomously
    • Screening for low left-ventricular ejection fraction from ECG data
    • Detecting every heart condition from every normal-looking tracing
  3. 什么物理信息为基于心电图的心脏模型提供输入？
    • Only the patient's appointment filename
    • Only the colors in an ultrasound image
    • Only the dimensions of a CT scanner's casing
    • A time series representing the heart's electrical activity
  4. HeartFlow 如何描述其 FFRCT 分析从冠状动脉 CT 衍生的解剖结构中估算的内容？
    • The patient's genetic sequence directly from image filenames
    • Coronary physiology using deep-learning and computational-fluid-dynamics methods
    • Every cardiac diagnosis without clinical interpretation
    • The exact time of every future cardiac event
  5. 此处讨论的 2023 年盲法随机超声心动图试验比较了什么？
    • An AI prescription against automatic surgery in every patient
    • A wearable ECG against a cholesterol blood test
    • Initial AI and sonographer ejection-fraction assessments followed by cardiologist evaluation
    • Whether images could be eliminated from echocardiography
  6. Why should a cardiac AI evaluation describe its patient population and clinical setting?
    • Performance in the studied data may not transfer unchanged to other patients or workflows
    • One reported score guarantees equal performance in every setting
    • Clinical setting cannot affect input quality
    • A model name alone specifies every intended use
  7. Which capability does Apple's ECG app documentation explicitly exclude?
    • Recording an ECG waveform for the specified supported device
    • Detecting a heart attack
    • Offering its specified rhythm classification
    • Exporting a recording for review
  8. How should an ECG-based low-ejection-fraction screening score be distinguished from an imaging measurement?
    • It directly measures every ventricular volume without imaging
    • It establishes that every CT-based result is interchangeable
    • It removes the need to understand the model's prediction target
    • It predicts a screening target from electrical data rather than directly measuring ventricular volumes from images