摘要
为了提高P波检测准确率,利用小波变换模极大值对在多尺度上的变化规律能表征信号突变点的性质,结合人体生理特性的检测策略进行心电信号P波的跨尺度检测。同时,引入反向传播神经网络对已检出的准P波再次进行确认与识别。经MIT数据库实验表明,P波检测准确率达到97%。
In order to promote the detection rote of P wave, the signal break points, combining with body physiology synthesis strategy practice can be indicated by using character of change rule that WT＇s mould max value pairs cross multi - scales the detection of P wave in ECG signal is developed in paper. At the same time , with helpe of BP neural network , P wave is confirmed. The detection rate of P wave is above 97 % for the MIT database.
出处
《生物医学工程研究》
2005年第4期209-211,共3页
Journal Of Biomedical Engineering Research
基金
北京市教育委员会科技发展计划资助项目(项目编号:KM200410016002)
