Adaptive Slope-Based R-Peak Detection in Multi-Activity Chest-Strap ECG Signals
Year: 2026
Type: preprint
Abstract: Reliable R-peak detection in wearable electrocardiogram (ECG) signals remains challenging during physical activities because motion-related disturbances and variations in signal morphology can affect detection accuracy. This study presents an adaptive slope-based R-peak detection framework for activity-dependent chest-strap ECG signals, implemented and evaluated in MATLAB/Simulink. The framework integrates digital filtering, derivative-based slope extraction, absolute-slope processing, moving-average integration, adaptive thresholding, and local peak refinement to identify R-peak locations. The proposed framework was evaluated using 123 annotated ECG recordings from 25 participants in the ECG-GUDB dataset, covering five activities: sitting, walking, mathematics, jogging, and hand biking. Four adaptive threshold configurations were compared using a primary temporal matching tolerance of 48 ms (12 samples at 250 Hz). The original maximum-based configuration achieved a precision of 97.299%, recall of 96.968%, and F1-score of 97.134%, with 22,518 true positives, 625 false positives, and 704 false negatives. It achieved the highest overall F1-score among the evaluated configurations. Activity-wise and subject-wise analyses examined performance variations, while timing-error and tolerance-sensitivity analyses assessed temporal localization and evaluation consistency. The findings demonstrate the feasibility of implementing and evaluating an adaptive slope-based R-peak detector in a discrete-time Simulink environment using multi-activity chest-strap ECG recordings. Further validation using independent participants and external ECG datasets is required to establish its generalizability.
Author: Sobhan, Abdus
Institution: Varendra University
Language: English
Topic: ECG Monitoring and Analysis
Subfield: Cardiology and Cardiovascular Medicine
Field: Medicine
Domain: Health Sciences
Open Access status: green
OpenAlex ID: W7236597803
