Application of ECG signal processing and classification based on CNN in wearable devices

Abstract

To address the issue of ECG signals being susceptible to internal and external noise interference during the acquisition process, in this paper, a novel ECG signal processing and classification method based on convolutional neural network (CNN) is proposed, which combines advanced Wavelet transform technology to achieve high precision classification of real-time ECG signals. Specifically, the method first uses Wavelet transform to preprocess and de-noise ECG signals to improve signal quality. Then, a customized CNN model is used to extract and classify the preprocessed ECG signals. The CNN model can be trained to accurately predict the type of ECG signal. The innovation of this paper lies in combining the highly efficient denoising capability of Wavelet transform with the powerful feature extraction capability of CNN algorithm to realize the high-precision classification of ECG signals. This approach not only improves the accuracy and reliability of ECG signal monitoring, but also opens the possibility of more effective health monitoring and disease prevention. The experimental results demonstrate significant improvements in the clarity and classification accuracy of the preprocessed ECG signals, with the trained CNN model achieving a 99% accuracy rate in predicting ECG signal types. This advancement not only enhances the accuracy and reliability of ECG signal monitoring by wearable devices but also provides possibilities for more effective health monitoring and disease prevention.