An Inception-GRU-SE Based Method for Arrhythmia Diagnostic

To address the challenge that a single network is insufficient for extracting comprehensive features from ECG data, we propose an Inception-GRU-SE-based diagnostic model for ECG signal classification. This model aims to enhance the efficiency of ECG screening while alleviating the workload of healthcare professionals by classifying ECG signals into seven distinct categories. The research primarily focuses on the following aspects: merging datasets of raw electrocardiogram (ECG) signals, standardizing ECG data using the Z-score method, annotating ECG signal labels, and subsequently thoroughly evaluating the proposed model using the processed ECG data. Experimental results show a significant improvement in ECG classification accuracy with the proposed model, achieving 99.13% accuracy on the test dataset compared to existing approaches.