天津大学学报(英文版)
天津大學學報(英文版)
천진대학학보(영문판)
TRANSACTIONS OF TIANJIN UNIVERSITY
2004年
2期
126-129
,共4页
余辉%谢远国%周仲兴%吕扬生
餘輝%謝遠國%週仲興%呂颺生
여휘%사원국%주중흥%려양생
wavelet%data mining%signal detection%electrocardiogram%dyadic wavelet%R peak detection
A new model based on dyadic differential wavelet was developed for detecting the R peak in Holter ECG signal according to the design of data mining. The Mallat recursive filter algorithm was introduced to calculate wavelet and optimize the detection algorithm which is based on the equivalent filter technique. The detection algorithm has been verified by MIT arrhythmia database with a high efficiency of 99%. After optimization,the algorithm was put into clinical experiment and tested in the Air Force Hospital in Tianjin for about two months. After about 108 hearts beating test of more than 100 patients, the total efficient detection rate has reached 97%. Now this algorithm module has been applied in business software and shows perfect performance under the complex conditions such as the inversion of heart beating, the falling off of the electrodes, the excursion of base line and so on.
