基于小波神经网络的心电数据压缩方法研究

ECG Data Compressing Based on the Wavelet Neural Network

本文在研究心电信号特征的基础上提出了动态心电数据的小波神经网络压缩算法。本算法对原始心电数据实行实时分桢;采用最速梯度下降法优化网络参数。对MIT-BIH数据库中的几组数据进行压缩,给出压缩/重建的实验结果并分析讨论。研究表明:该算法具有参数收敛速度快、数据压缩性能良好的特点。

This paper presents an ECG data compression algorithm based on the wavelet neural network. The original ECG data is divided into frames, and the fastest descending gradient method is used to optimize the network parameters. Compress several groups Experimental results based on the ECG data of MIT/BIH database, are given in the paper.