Compressed estimation of heart and respiratory rates from a photoplethysmogram

Abstract

Many mobile healthcare (m-healthcare) devices, such as smart watch and smart band, use a photoplethysmogram (PPG) sensor to measure the user's heart rate (HR) and respiratory rate (RR). Since, when such devices measure PPG, it should illuminates skin using a light emitted diode, their battery life depend on sampling rate. Hence, reducing sampling rate is important problem for these m-healthcare devices which utilize a PPG sensor. Several compression schemes were introduced, but most of them were insufficient for PPG compression. In this study, to enhance the efficiency of m-healthcare devices, we introduced a new compression scheme for PPG using compressive covariance sensing. It can estimate HR and RR from compressed PPG sampled at a lower rate than the Nyquist sampling rate. Its compression and estimation performances were satisfactory, so we expect this technique will contribute to m-healthcare.