Scholarlink

<p>AI-enabled electrocardiogram (ECG) models trained on 12-lead recordings may underperform on home single-lead devices due to device and context domain shift. We trained Lead-I models using 676,192 ECGs from 246,874 patients and externally validated them in 219,231 ECGs from 65,338 patients. We evaluated two methodological control indicators (gender and age) and seven clinical phenotypes [90-day mortality, left ventricular ejection fraction (EF), pulmonary artery systolic pressure (PASP), left atrial diameter, N-terminal pro-B-type natriuretic peptide (NT-proBNP), hemoglobin, and estimated glomerular filtration rate]. Cross-hospital validation showed minimal overall degradation (AUC decline &lt;0.03). In contrast, direct deployment to consumer ECGs from Apple Watch (<i>n</i> = 9835) and QOCA ECG102D (<i>n</i> = 31,517) produced substantial accuracy losses. After fine-tuning, AUCs across eight classification tasks for the Apple Watch and QOCA ECG102D were significantly improved, with high performance for low EF (0.85/0.86) and more modest performance for elevated PASP (0.73/0.68) and anemia (0.73/0.69). Performance was higher in sinus rhythm than atrial fibrillation. Learning curves indicated that approximately 500–1000 labeled ECGs are required for reliable fine-tuning; smaller sets can be harmful. We release an evaluation-only HOME dataset to facilitate reproducible benchmarking of cross-device generalization. Models should be interpreted as screening or triage tools, not standalone diagnostic tests.</p>

错误：搜索内容不能为空，请输入英文关键词
错误：关键词超出字数限制，请精简
高级检索

Bridging the gap from clinical to home ECG: quantifying and overcoming accuracy loss in AI-enabled single-lead ECG models

  • 期刊论文

  • 出版模式：

    开放获取

  • 发表日期：

    2026年7月6日

查看全文
  • Chin Lin,
  • Wei-Ting Liu,
  • Kai-Chieh Chen,
  • Dung-Jang Tsai,
  • Da-Wei Chang,
  • Tsung-Neng Tsai,
  • Cheng-Chung Cheng,
  • Chih-Yuan Lin,
  • Yuan-Hao Chen,
  • Chien-Sung Tsai,
  • Shih-Hua Lin,
  • Chin-Sheng Lin

摘要

AI-enabled electrocardiogram (ECG) models trained on 12-lead recordings may underperform on home single-lead devices due to device and context domain shift. We trained Lead-I models using 676,192 ECGs from 246,874 patients and externally validated them in 219,231 ECGs from 65,338 patients. We evaluated two methodological control indicators (gender and age) and seven clinical phenotypes [90-day mortality, left ventricular ejection fraction (EF), pulmonary artery systolic pressure (PASP), left atrial diameter, N-terminal pro-B-type natriuretic peptide (NT-proBNP), hemoglobin, and estimated glomerular filtration rate]. Cross-hospital validation showed minimal overall degradation (AUC decline <0.03). In contrast, direct deployment to consumer ECGs from Apple Watch (n = 9835) and QOCA ECG102D (n = 31,517) produced substantial accuracy losses. After fine-tuning, AUCs across eight classification tasks for the Apple Watch and QOCA ECG102D were significantly improved, with high performance for low EF (0.85/0.86) and more modest performance for elevated PASP (0.73/0.68) and anemia (0.73/0.69). Performance was higher in sinus rhythm than atrial fibrillation. Learning curves indicated that approximately 500–1000 labeled ECGs are required for reliable fine-tuning; smaller sets can be harmful. We release an evaluation-only HOME dataset to facilitate reproducible benchmarking of cross-device generalization. Models should be interpreted as screening or triage tools, not standalone diagnostic tests.