Simple Clustering-Based Algorithm for R-Peak Detection in Single-Lead ECG Chest Monitors

This study presents a simple algorithm for R-peak detection in single-lead electrocardiograms (ECG) acquired on the chest of various volunteers using a wearable device. To achieve a balance between computational efficiency and detection accuracy, the proposed algorithm consists of two optimized stages that can be implemented in portable hardware. First, the ECG signal is preprocessed using the discrete wavelet transform (DWT) to remove the baseline wandering and denoise the signal. Second, a clustering-based algorithm classifies the different peaks in the ECG signal into two clusters based on their max-min amplitude. It is demonstrated that, despite the high variability of the ECG waveform among volunteers, the proposed algorithm effectively identifies the R-peaks with high accuracy without requiring the specific cardiac profile for each individual patient, thus ensuring a reliable and robust heart rate (HR) monitoring.