Sleep disorders such as obstructive sleep apnea (OSA) are now regarded as an important risk factor for cardiovascular diseases. The increasing demands for home-based sleep monitoring have prompted studies to develop devices that monitor sleep using fewer sensors. This paper proposes an electrocardiogram (ECG)-based sleep monitoring system that uses a personal digital assistant to display signals and a wireless transmission system. An OSA detection algorithm based only on the overnight-ECG-derived heart-rate-variability low-frequency component (0.02-0.04 Hz) exhibited detection sensitivity of 68.97% and specificity of 100%. This system could meet the future demands of home-based sleep monitoring and thereby reduce the current considerable burdens on hospital sleep centers.
