Development of Heart Rate Monitoring System Based on Wrist Acoustic Sensor

This study presents an acoustic sensor system for heart rate detection at the wrist. Initially, the system was used to detect resting heart rate, followed by the development of an adaptive denoising heart rate detection algorithm called RMEE (Kernel Recursive Minimum Error Entropy Adaptive Filter) to remove motion artifacts. The algorithm utilizes a dual-microphone setup to synchronously capture wrist acoustic signals: one channel detects the pulse wave signal, while the other captures motion noise. The RMEE algorithm employs an adaptive filtering technique based on the minimum error entropy criterion to effectively eliminate motion-induced noise. Additionally, spectral peak tracking is applied to reduce random interference, thereby enabling accurate heart rate estimation. Tests conducted on data collected from 12 volunteers, involving various hand movements, demonstrate that the proposed system exhibits strong resistance to motion artifacts, such as those caused by wrist movement, maintaining high stability in heart rate monitoring. The proposed method shows significant practical value, providing a reliable solution for heart rate detection in dynamic environments, with potential applications in next-generation wearable devices for continuous health monitoring.