单导心电监护中不同电极模式的自动识别方法

Automatic Identification of Different Electrode Patterns in Single-Lead ECG Monitoring

电极脱落、电极混用等不正常电极模式会影响采集的心电信号的质量,导致误诊。人工判断电极模式是否正常过于依赖人的经验。为此对单导心电采集中4种电极模式的自动识别方法进行研究,定义了表征不同电极模式的心电信号的数字特征,结合基于LDA的特征降维和最近邻分类器来实现自动识别。实验表明,所提出的自动识别方法对单导心电设备4种电极模式的自动识别有效,对电极脱落的正确识别率可达到100%。

Some of the abnormal electrode patterns,such as electrode fall-off,mixed use of different electrodes,can seriously affect the ECG signal quality,even resulting in wrong diagnosis.Artificial identification of the electrode patterns depends on personal experiences.Hence,the automatic identification method for four kinds of electrode patterns in single-lead ECG monitoring is studied.Firstly,eight digital indices for distinguishing different electrode patterns are defined.Secondly,the LDA-based feature dimension reduction and the nearest neighbor classifier are adopted to complete the identification process.Experiments show that the proposed automatic identification method for different electrode patterns in single lead ECG monitoring is effective,with successful identification rate of 100% for electrode fall-off.