心电信号中伪差检测的方法研究

Study on Detection Method of Error Segments in ECG Signal

目的:设计一种实用的算法,能够自动识别长时间记录心电图中的伪差波形,达到用户快速剔除干扰的目的。方法:通过分析伪差发生的原因和实际呈现形态,总结出各类伪差波形有别于正常波形的特征,设计合适的算法。结果:从大量实测心电图中,总结了伪差波形的形态,对于饱和、微小、剧烈波动、趋势漂移等伪差形态,设计了具有针对性的识别算法,并用实测信号验证了算法的有效性。结论:该伪差识别算法在24小时心电HOLTER信号分析软件中得到应用,能够比较准确地自动筛查出信号中的大量伪差,为用户快速分析心电图提供了新的有效方法。

Objective： In order to let user eliminate disturbance rapidly, we designed a practical algorithm for detecting error segments in ECG signal. Methods： After looking into a lot of ECG signal, we summarized the features of error segments and designed the algorithm suitable for detecting them automatically. Results： The algorithm was realized and tested with actual ECG signal, its validity was verified. Conclusion： The algorithm proposed in this paper is designed for detecting error segments in ECG signal. It is used in ECG Holter analysis software and let user eliminate error segnnents quickly. The satisfactory results are achieved.