ECG Signal Classification Algorithm Based on Improved VGG Network

Efficient, accurate, and fast classification of ECG signals has always been a key focus in the field of ECG diagnosis. This paper presents an algorithm for ECG signal classification that enhances the VGG network. The traditional VGG network is improved for different categories by optimizing the connection layer structure. Additionally, depth separable convolution is introduced to reduce model parameters and improve training efficiency. Since ECG signals contain a significant amount of temporal feature information, a long short-term memory network model is incorporated to capture temporal correlations within the data. Four classifications are then implemented using softmax. Experimental results demonstrate that the enhanced model achieves a classification accuracy of $\mathbf{9 3. 6 \%}$.