BCG 신호와 특이 스펙트럼 분석을 이용한 반려견의 수면 심박·호흡 추정

초록

With the growth and aging of the companion animal population, continuous physiological monitoring is increasingly needed for early detection of cardiovascular diseases. This study proposes a Singular Spectrum Analysis (SSA)-based method for estimating heart rate (HR) and respiratory rate (RR) during canine sleep using ballistocardiography (BCG)-related micro-vibrations measured at the neck with a six-axis inertial measurement unit (IMU). The method includes sleep candidate detection, cardiac and respiratory signal separation, SSA-based decomposition and reconstruction, peak detection, and Signal Quality Index (SQI) evaluation. Overnight data from one small and one large dog showed mean HR and RR values of 71.9 bpm and 24.3 rpm, with mean SQI values of 86.5 and 80.7, respectively. The results suggest the feasibility of non-restrictive monitoring of cardiac and respiratory signals in sleeping dogs using the proposed wearable system.

키워드