基于下采样检测QRS波的新方法

针对目前QRS波检测方法的不足,结合心电信号主要集中在0.05～35 Hz这一特点,我们提出了基于下采样的QRS波检测新方法.首先对原始心电信号进行下采样,然后利用自适应、自学习原理设置幅值和斜率阈值检测QRS波,利用不应期特性降低误检率,再利用回溯技术降低漏检率,最后对得到的QRS波进行更新来实现原始心电信号中QRS波的检测.采用MIT/BIH心电数据库的数据和重庆医科大学附属医院提供的临床数据对该方法进行验证,准确率可以达到98%.结果表明该方法简单、快速、检出率高,并且不受采样率的影响.

A New Method Based on Down-Sampling for QRS Complexes Detection

This paper presented a new method based on down-sampling for QRS complexes detection for the reason that ECG's frequency components mostly centralize 0.05-35 Hz. Firstly, the original ECG signals were down-sampled. Secondly, self adapting and self study theory were used to select the amplitude threshold and slope threshold for R-wave detection. At the same time, the refractory idiosyncrasy of heart was taken into account to reduce the misdetection. Thirdly, back-trace technique was used to reduce undetected error. Finally, the corresponding R-wave position was up-dated in the original ECG signals. The performance of the method was tested by MIT/BIH ECG database and the clinical data were collected from the affiliated hospital of Chongqing Medical University. The detection rate of 98% was achieved. Experimental results indicated that the method was simple, effective, accurate and unaffected on sample rate.