麻豆精品视频在线_韩国三级av_亚洲一区中文_涩涩的网址_91极品尤物_日本猛少妇色xxxxx_在线性视频_亚洲第一综合_大牛影视剧免费播放在线_被触手肉干高h潮文

Application of Lightweight Convolutional Neural Network in ECG Classification

• 1、School of Electronics and Information, Liaoning Technical University

Abstract：Objective To address the inefficiency of traditional ECG analysis relying on manual diagnosis, this paper proposes an automatic arrhythmia classification method based on one-dimensional convolutional neural network. Methods ECG signals were obtained from the MIT-BIH arrhythmia database of PhysioNet. After preprocessing, denoising and heartbeat segmentation, a heartbeat dataset containing normal beats (N) and premature ventricular contractions (V) was constructed. A lightweight one-dimensional convolutional neural network model was designed to extract deep features of ECG signals through multi-layer convolution and pooling operations, and Dropout regularization was adopted to prevent overfitting. Results On the dataset containing 100 heartbeat samples, the model achieved 100% classification accuracy on the test set, with precision, recall and F1 score all reaching 1.00. The model converged after only 6 training epochs, demonstrating excellent classification performance and training efficiency. Conclusion This method can effectively realize automatic classification of ECG signals and provides a reliable auxiliary tool for rapid diagnosis of arrhythmia.

Keywords： ECG signal arrhythmia deep learning convolutional neural network automatic diagnosis