Abstract: Cardiovascular disease is one of the diseases with high mortality rate in China. Monitoring electrocardiograms to determine if there are abnormalities in the electrical signals of the heart can be used to prevent and screen for cardiovascular disease. Due to the large scale and complexity of electrocardiogram data, clinical medical staff have a heavy workload and are prone to misdiagnosis or missed diagnosis during electrocardiogram screening. In order to improve the screening efficiency of electrocardiogram and reduce the pressure on medical staff, a model based on convolutional neural network, long and short-term memory neural network and SE network（CNN-LSTM-SE） was proposed to divide electrocardiogram into five categories. The main research contents include: MIT-BIH arrhythmia data set is selected as the data source of ECG signals, Butterworth bandpass filter is used to de-noise ECG signals, Z-score method is used to standardize ECG signals, and unique thermal coding method is used to encode ECG labels.Finally, the proposed algorithm model is trained and tested using the processed ECG data. The experimental results show that compared with other models, the proposed model can effectively improve the accuracy of ECG classification, and the classification accuracy of the experimental data set reaches 99. 1%.