神经网络方法对心电信号主元压缩算法的改进

Data Compression of Ambulatory ECG Using Improved NN Algorithm

根据主元分析和神经网络的原理,描述了用主元分析的方法对输入信号的特征进行提取,并利用神经网络算法,实现ECG信号压缩的可行性。从压缩后信号重构的保真性和压缩计算复杂性两个角度出发,给出了压缩\重建的实验结果并分析讨论。文中的压缩算法将原始动态心电数据映射为主元分析特征参数存储,比常规的直接压缩或压缩算法有更高的压缩比。选取5个主元表示一个心动周期的特征,其压缩比可达到30∶1,均方根误差比PRD为4.75%。该ECG压缩算法克服了直接用主元分析法压缩数据时,要计算维数巨大相关矩阵的困难。实验证明,该方法在保留ECG信号临床信息的同时,具有较好的压缩效果。

The improved algorithm designed for the compression of ambulatory ECG data is described. This method is based on the principal component analysis and neural network. Moreover, ECG signal reconstruction can be achieved by principal component analysis. The algorithm records original ECG data as principal component parameters and it has higher compression ratio. When using 5 principal components to record the feature of ECG data, it also can achieve the compressing result as: PRQ 4%, CR 17:1. The experiment results of data compression/reconstruction are referred and analyzed. From the study, it can be concluded that the compression algorithm can provide superior performance.