A clinical study on Atrial Fibrillation, Premature Ventricular Contraction, and Premature Atrial Contraction screening based on an ECG deep learning model

Abstract

It is still a challenge to develop an electrocardiography (ECG) interpreter based on ECG basic characteristics because of the uncertainty of ECG delineation. Based on the clinical investigation in this study, ECG devices generated interpretations of Atrial Fibrillation (AF), Premature Ventricular Contraction (PVC), and Premature Atrial Contraction (PAC) have high ratios of false-positive errors. An ECG interpretation gap exists between ECG devices and cardiologists. This study aimed to develop an ECG interpreter to improve the performance of AF, PVC, and PAC based on clinical ECGs. This study first adopted a deep learning model to delineate ECG features such as P, QRS, and T waves based on 1160 8–10-s lead I or lead II ECG signals from a clinically-used 12-lead ECG device whose ECG device interpretation is AF as a training dataset. Second, a sliding window with 3-RR intervals in length is applied to the raw ECG to examine the delineated features in the window, and the ECG interpretation is then determined based on the experiences of cardiologists. The results indicate the following: (1) This delineator achieves good performance on P-, QRS-, and T- wave delineation with a sensitivity/specificity of 0.94/0.98, 1.00/0.99, and 0.97/0.98, respectively, in 48 10-s test ECGs mixed with true-positive AF and false-positive AF ECGs. (2) As compared to ECG-device generated interpretations, the precision of the detection of AF, PVC, and PAC in this study was increased from 0.77 to 0.86, 0.76 to 0.84, and 0.82 to 0.87 in 188 10-s test ECGs. Finally, (3) the F1 measure, which is a measure of the accuracy of test data but takes false-positive and false-negative into account, on the detection of AF, PVC, and PAC were 0.92, 0.91, and 0.83, respectively. In conclusion, this study overcomes the difficulties of ECG P-wave discrimination between true-positive AF and false-positive AF which are not documented well in previous research and improves the precision of ECG devices’ interpretation. We believe that this study can facilitate clinical applications of ECG, and bridge the gap between machines’ ECG interpretation and cardiologists.