Automatic Identification of Systolic Time Intervals in Seismocardiogram.

摘要: Continuous and non-invasive monitoring of hemodynamic parameters through unobtrusive wearable sensors can potentially aid in early detection cardiac abnormalities, provides a viable solution for long-term follow-up patients with chronic cardiovascular diseases without disrupting the daily life activities. Electrocardiogram (ECG) siesmocardiogram (SCG) signals be readily acquired from light-weight electrodes accelerometers respectively, which employed to derive systolic time intervals (STI). For this purpose, automated accurate annotation relevant peaks these is required, challenging due inter-subject morphological variability noise prone nature SCG signal. In paper, an approach proposed automatically annotate desired signal that are related STI by utilizing information peak detected sliding template narrow-down search actual Experimental validation performed conventional/controlled supine realistic/challenging seated conditions, containing over 5600 heart beat cycles shows good performance robustness noisy conditions. Automated measurement configuration provide quantified health index patients, elderly people at risk health-enthusiasts.