Presacan Oriana, Dorobanţiu Alexandru, Isaksen Jonas L, Willi Tobias, Graff Claus, Riegler Michael A, Sridhar Arun R, Kanters Jørgen K, Thambawita Vajira
Oslo Metropolitan University, 0167, Oslo, Norway.
Lucian Blaga University of Sibiu, 550024, Sibiu, Romania.
Commun Med (Lond). 2025 Apr 25;5(1):139. doi: 10.1038/s43856-025-00814-w.
Wearables with integrated electrocardiogram (ECG) acquisition have made single-lead ECGs widely accessible to patients and consumers. However, the 12-lead ECG remains the gold standard for most clinical cardiac assessments. In this study, we developed a neural network to reconstruct 12-lead ECGs from single-lead and dual-lead ECGs, and evaluated the mathematical accuracy.
We used lead I or leads I and II from 9514 individuals from the Physikalisch-Technische Bundesanstalt (PTB-XL) cohort and a generative adversarial network, with the aim of recreating the missing leads from the 12-lead ECG. ECGs were divided into training, validation, and testing (10%). Original and recreated leads were measured with a commercially available algorithm. Differences in means and variances were assessed with Student's t-tests and F-tests, respectively. Calibration and bias were assessed with Bland-Altman plots. Inter-lead correlations were compared in original and recreated ECGs.
The variability of precordial ECG amplitudes is significantly reduced in recreated ECGs compared to real ECGs (all p < 0.05), indicating regression-to-the-mean. Amplitude averages are recreated with bias (p < 0.05 for most leads). Reconstruction errors depend on the real amplitudes, suggesting regression-to-the-mean (R between target and error in R-peak amplitude in lead V3: 0.92). The relations between lead markers have a similar slope but are much stronger due to reduced variance (R-peak amplitude R between leads I and V3, real ECGs: 0.04, recreated ECGs: 0.49). Using two leads does not significantly improve 12-lead recreation.
AI-based 12-lead ECG reconstruction results in a regression-to-the-mean effect rather than personalized output, rendering it unsuitable for clinical use.
集成心电图（ECG）采集功能的可穿戴设备已使单导联心电图广泛应用于患者和消费者。然而，12导联心电图仍是大多数临床心脏评估的金标准。在本研究中，我们开发了一种神经网络，用于从单导联和双导联心电图重建12导联心电图，并评估其数学准确性。
我们使用了德国物理技术联邦研究所（PTB-XL）队列中9514名个体的I导联或I导联和II导联，以及一个生成对抗网络，旨在从12导联心电图中重建缺失的导联。心电图被分为训练集、验证集和测试集（10%）。使用市售算法测量原始导联和重建导联。分别用学生t检验和F检验评估均值和方差的差异。用布兰德-奥特曼图评估校准和偏差。比较原始心电图和重建心电图中的导联间相关性。
与真实心电图相比，重建心电图中胸前导联心电图振幅的变异性显著降低（所有p<0.05），表明存在均值回归。振幅平均值重建存在偏差（大多数导联p<0.05）。重建误差取决于真实振幅，表明存在均值回归（V3导联R波峰值振幅的目标值与误差值之间的R为0.92）。导联标记之间的关系具有相似的斜率，但由于方差减小而更强（I导联和V3导联之间的R波峰值振幅R，真实心电图：0.04，重建心电图：0.49）。使用两个导联并不能显著改善12导联重建。
基于人工智能的12导联心电图重建会产生均值回归效应而非个性化输出，因此不适合临床使用。
