基于增量高斯混合模型的心电分类研究

ECG Classification Based on Incremental Gaussian Mixture Model

本研究提出一种基于增量高斯混合模型算法的心电分类方式。与其相对应的分为三个阶段:第一阶段,基于小波分解对心电信号进行预处理,以消除基线漂移(0.15~0.3 Hz)和背景噪声;第二阶段,首先基于短时修正希尔伯特变换(STMHT)定位R峰,然后自动确定QRS波进而提取心电特征参数;第三阶段提出一种无监督式增量高斯混合模型算法实现心电信号分类。本研究结果在MIT-BIH心律失常数据库中对48组心电数据进行分类比较,其准确率达93.31%。

In this paper, an incremental Gaussian mixture model algorithm (IGMM)-based classification method is proposed for electrocardiogram (ECG) diagnosis. The three stages corresponding to the classification system are generally divided into three stages. In the first stage, wavelet decomposition-based preprocessing method is employed to eliminate the baseline drift (0.15 - 0.3 hz) and background noise. In the second stage, the automatic R peak is firstly located via the short time modified Hilbert transform (STMHT), and then the QRS complex is automatically determined to extract the ECG parameters. And an unsupervised incremental Gaussian mixture model algorithm is proposed to classify the ECG signal in the third stage. As comparative accuracies of classifying 48 groups of ECG data in MIT-BIH database, the higher classification accuracy of 93.31% greater than other methods is obtained.