LAUSR

Heart rate variability (HRV) is an important physiological parameter to assess the status of the autonomic nervous system (ANS) and diagnose cardiovascular diseases (CVDs). HRV is generally extracted from the cardiac activity monitored at the chest using electrocardiography (ECG). This article presents a novel algorithm to accurately extract HRV, from signals obtained using an innovative wearable approach based on acoustic sensing of cardiac activity at the wrist. The proposed methodology is based on the relative energy method, which utilizes the relative information between short and long-term energies of the acoustic signal to enhance the S1’s characteristics in comparison to other signal transitions. The algorithm was tested on a dataset consisting of acoustic recordings of 5-min duration obtained from 12 subjects. For the instantaneous heart rate (IHR) determination, 98.50% and 98.96% of the total samples achieved a relative error of less than ±5% when compared to ECG and photoplethysmography (PPG) signals, respectively. The comparison of the HRV parameters in the time-domain and the frequency-domain, with the reference signals, also demonstrated a strong statistical agreement and correlation of close to one, hence proving the feasibility of extracting the HRV from acoustic signals recorded at the wrist.