Innovative R peak detecting algorithm using enhanced artificial bee colony

R peaks is the most significant point in electrocardiogram (ECG) because many studies adopt interval between adjacent R peaks to diagnose arrhythmia disease. The previous effective R detection algorithm consisted of band-pass filter, interesting block and threshold to obtain the high performance of detecting R peaks. However, this previous study using a knowledge-based theory spent a lot of time to produce the parameters (F1, F2, W1, W2 and β). Our previous study adopted the optimization algorithm to replace the knowledge-based theory and found out the best parameters quickly. In addition, our previous study added the one parameter (order of filter, N) in the algorithm and obtained the high performance. This study revises the definition of threshold for determining the meaningful interesting block and adds one parameter (multiple of standard deviation in threshold, μ) in the algorithm and adopts enhance artificial bee colony (EABC) to find the suitable parameter in the algorithm. This study adopts the ECG signals in MIT-BIH arrhythmia database as training and testing data. The results of this study show that revised ABC can faster than knowledge-based theory to find the parameters in R detection algorithm and have a high accuracy (99.7737%). In the future, the proposed study can combine with the medical treatment systems to achieve real-time diagnosis system.