Jin Yeongbong, Ko Bonggyun, Chang Woojin, Choi Kang-Ho, Lee Ki Hong
Department of Industrial Engineering, Seoul National University, Seoul, Korea.
Department of Mathematics and Statistics, Chonnam National University, Gwangju, Korea.
Korean J Intern Med. 2025 Mar;40(2):251-261. doi: 10.3904/kjim.2024.130. Epub 2025 Feb 21.
BACKGROUND/AIMS: Atrial fibrillation (AF) significantly contributes to global morbidity and mortality. Paroxysmal atrial fibrillation (PAF) is particularly common among patients with cryptogenic strokes or transient ischemic attacks and has a silent nature. This study aims to develop reliable artificial intelligence (AI) algorithms to detect early signs of AF in patients with normal sinus rhythm (NSR) using a 12-lead electrocardiogram (ECG).
Between 2013 and 2020, 552,372 ECG traces from 318,321 patients were collected and split into training (n = 331,422), validation (n = 110,475), and test sets (n = 110,475). Deep neural networks were then trained to predict AF onset within one month of NSR. Model performance was evaluated using the area under the receiver operating characteristic curve (AUROC). An explainable AI technique was employed to identify the inference evidence underlying the predictions of deep learning models.
The AUROC for early diagnosis of PAF was 0.905 ± 0.007. The findings reveal that the vicinity of the T wave, including the ST segment and S-peak, significantly influences the ability of the trained neural network to diagnose PAF. Additionally, comparing the summarized ECG in NSR with those in PAF revealed that nonspecific ST-T abnormalities and inverted T waves were associated with PAF.
Deep learning can predict AF onset from NSR while detecting key features that influence decisions. This suggests that identifying undetected AF may serve as a predictive tool for PAF screening, offering valuable insights into cardiac dysfunction and stroke risk.
背景/目的：心房颤动（AF）是全球发病和死亡的重要原因。阵发性心房颤动（PAF）在不明原因卒中或短暂性脑缺血发作患者中尤为常见，且具有隐匿性。本研究旨在开发可靠的人工智能（AI）算法，利用12导联心电图（ECG）检测窦性心律正常（NSR）患者的房颤早期迹象。
2013年至2020年间，收集了318321例患者的552372份心电图记录，并将其分为训练集（n = 331422）、验证集（n = 110475）和测试集（n = 110475）。然后训练深度神经网络以预测NSR后1个月内房颤的发作。使用受试者工作特征曲线下面积（AUROC）评估模型性能。采用可解释的人工智能技术来识别深度学习模型预测背后的推理证据。
PAF早期诊断的AUROC为0.905±0.007。研究结果表明，T波附近，包括ST段和S波峰，对训练后的神经网络诊断PAF的能力有显著影响。此外，将NSR中的汇总心电图与PAF中的心电图进行比较发现，非特异性ST-T异常和T波倒置与PAF相关。
深度学习可以从NSR预测房颤发作，同时检测影响决策的关键特征。这表明识别未被检测到的房颤可作为PAF筛查的预测工具，为心脏功能障碍和中风风险提供有价值的见解。
