Fast ECG anomaly detection on Android platform

This paper is to present a fast and real-time algorithm for detecting abnormal ECG signals by using Welch Method to preprocess the data of normal sinus rhythm database of MIT-BIH, and then analyze with K-means algorithm to get the normal form of ECG categories and store them in Android mobile terminal. In real-time detection, extract the ECG waveform and match with the normal ECG categories to get whether it is abnormal. The time of measuring 10s electrocardiogram was 3.03s, the precision rate was 94.05%, and the recall rate was 93.12%. This method can measure the ECG abnormity rapidly in real time, and the accuracy is similar to the traditional method.