基于ART2神经网络的动态ECG数据压缩算法

ART2 NEURAL NETWORKBASED DATA COMPRESSION OF THE DYNAMIC ECG

采用ＡＲＴ（ＡｄａｐｔｉｖｅＲｅｓｏｎａｎｃｅＴｈｅｏｒｙ）神经网络中的适应于模拟信号输入的ＡＲＴ２网络结构，提出了一种动态ＥＣＧ数据压缩算法。借助于ＡＲＴ２优良的模式识别特性，该算法能实现高压缩比和高压缩精度的动态ＥＣＧ数据压缩。ＡＲＴ２网络能迅速识别已学习样本及事先未知的新样本，因此该算法不仅能快速适应各种未知ＥＣＧ信号的动态变化，而且提高了算法的运行速度和数据压缩比。

his paper proposed a data compression algorithm of the dynamic ECG based on ART2 neural network structure. ART2 is one of the ART(Adaptive Resonance Theory) based neural networks and suited to analog input signal patterns. Because the ART2 had the feature of good category recognition, high compress rate and high compression precision of the ECG could be reached by the algorithm. ART2 network can rapidly recognize the learned samples and the new unknown samples, thus improving the running speed and the data compression rate of the algorithm.