Abstract：Objective To analyze user's ECG data recorded and uploaded by remote ECG monitoring smartphone in natural population, and to compare the diagnostic effects between intelligent analysis and artificial diagnosis. Methods Smart handheld ECG measuring equipment was used in our study. A total of 9259 volunteers in 11 cities of Shanxi province were enrolled, among whom were 3872 males and 5387 females at an average age of (42.7±18.9) years old. A total of 12 109 ECGs were obtained. Results(i) In the artificial diagnosis method, 602 abnormal ECGs are identified with a detection rate of 65%; by using intelligent analysis software, 1296 abnormal ECGs are identified with a detection rate of 14.0%, and the difference is of statistical significance(P<0.005). (ii) Tachyarrhythmia is detected in 277 cases(46.0%) while atrial fibrillation is detected in 30 cases(5.0%). (iii) By using intelligent analysis software, 10 789 ECGs are correctly diagnosed, and its coincidence rate with artificial diagnosis is 89.1%. Among the 1320 misdiagnosed ECGs, there are 1213 cases(91.9%) caused by interference fluctuations. Conclusion By making intelligent analysis of user's ECGs recorded and uploaded by handheld singlelead ECG equipment, arrhythmias can be identified, and atrial fibrillation can be detected. The coincidence rate of autodiagnosis by intelligent software is tolerable, but it still requires for further enhancement of abilities of interference identification.
