基于模式识别的胸阻抗信号自动检测算法

Automatic Detection Algorithm for Transthoracic Impedance Signal Based on Pattern Recognition

为了自动识别胸阻抗(TTI)信号中的按压和通气波形,完成相关重要参数的计算,并结合先验知识和机器智能从而完成对心肺复苏质量的监测评估,提出了一种基于模式识别的胸阻抗信号自动检测算法。基于实验采集的猪的电诱导心脏骤停模型TTI信号,设计结合小波和形态学的除噪算法对信号进行预处理,再由多分辨率窗口搜索法完成潜在按压和通气波形的定位,最后采用线性判别分析法对定位的按压和通气波形进行分类识别。实验结果表明,该算法对TTI信号中按压波形和波形分析识别的正确率和敏感度可达到98.237%、94.947%和99.651%、97.282%,稳定性好,且运行时间(0.485±0.07 s)满足实时性要求。

In order to recognize the compression and ventilation waveforms, obtain the important parameters, and evaluate the CPR quality by combining with prior knowledge, this paper proposes an automatic detection algorithm for transthoracic impedance （TTI） signal based on pattern recognition. The TTI signals that come from pig model based on electrically induced cardiac arrest are reprocessed by denoising algorithm based on wavelet and morphology firstly. Then the potential compression and ventilation waveforms are located by using the searching algorithm of multiresolution window. Finally, the linear discriminant analysis algorithm is used to classify and recognize the located compression and ventilation waveforms. The results show that both the recognition accuracy and sensitivity of the compression and ventilation waveforms are 98.237%, 94.947% and 99.651%, 97.282%, and the running time （0.485±0.07s） satisfies the requirement of clinical applications.