An R-peak detection method based on peaks of Shannon energy envelope

Abstract

In a real-time electrocardiogram (ECG) monitoring system, to detect the R-peak in each beat with low delay, high accuracy and high speed is an important and fundamental task. Aiming at satisfying these requirements, a novel R-peak detection algorithm based on Shannon energy envelope (PSEE) is introduced. It extracts Shannon energy envelope (SEE) from ECG firstly, then finds out the R-peaks after SEE by three new introduced sub-processes: peak detection, false-R detection and false-noise detection. Owing to these sub-processes, the whole schema can deal with long pause or asystole case, and avoid segmenting ECG. Moreover, by splitting the calculation into every sampling time slot, the delay is guaranteed. The new method has been evaluated with all records in MIT–BIH with 99.83% detection rate of R-peaks and about 1.77μs computation time for each sample and with 251 records in Chinese Cardiovascular Disease Database with total accuracy 99.78%. Therefore, it not only satisfies real-time requirements but also improves the detection accuracy rate.