Detection of Sleep Apnea Based on Cardiopulmonary Coupling

Abstract

Sleep plays an important role in human life activities, and Obstructive sleep apnea (OSA) is a very important factor. Sleep apnea is a common sleep-related respiratory disease. Polysomnography (PSG) is the gold standard for detecting sleep apnea, but PSG is a contact device, there will be a first night effect, and some people may even be disturbed by long-term incompatibility. This study proposes to extract the cardiopulmonary coupling (CPC) strength based on the Ballistocardiogram (BCG) signal to achieve the purpose of further improving the accuracy of OSA detection. We extract heart rate and breathing through the BCG signal. Then, the time domain features and frequency domain features of the heartbeat interval sequence over a fixed length of time are extracted. The coupling strength of the two signals is further analyzed to generate cardiopulmonary coupling characteristics. A classification model of sleep apnea is used to determine whether sleep apnea occurs within a fixed length of time. And, the accuracy can be further improved by adding the cardiopulmonary coupling feature.