Classification of ECG Arrhythmia Using CNN, SVM and LDA

摘要

Nowadays heart disease is one of the serious diseases threatening human health, and a robust and efficient method is needed to achieve a realtime analysis and help doctors to diagnose. In this paper, we mainly propose an ECG arrhythmia classification algorithm based on convolutional neural network (CNN). Specifically we compare different CNN models, and then use them to raise the correct rate of classification combining linear discriminant analysis (LDA) and support vector machine (SVM). All cardiac arrhythmia beats are derived from MIT-BIH Arrhythmia Database, which are divided into five types according to the standard developed by the Association for the Advancement of Medical Instrumentation (AAMI). The training set and the testing set come from different people and the correction of classification is greater than 90%.