Adaptive thresholding based low complex QRS detection algorithm

Abstract

In this study, a QRS detection algorithm with a low processing load based on time-domain thresholding is proposed. The ECG signal is filtered only with a low pass filter to reduce the computational load. After the filtering, derivation and squaring are also performed. In the Thresholding stage, a linear decreasing threshold voltage method using addition operation instead of multiplication is proposed. Simulations on MIT-BIT Arrhythmia Database have yielded 99.2925% sensitivity (% Se) and 99.6759% positive predictivity (+ P). The proposed algorithm is compared with two similar algorithms in terms of both performance and processing load. It is shown that the proposed algorithm is better than its counterparts especially in terms of processing load. However, it is observed that it gave worse results in terms of Sensitivity (% Se).