Detection Of Apnea Syndrome Based On Non-contact Sleep Monitoring System

Sleep is a complex physiological process that plays an important role in human life.Although there are rapid economy developments and the improvement of people’s living standards,people’s sleep time and quality is decreasing.Sleep apnea is a normal sleep-related respiratory disease,which significantly affects people’s health.The detection of sleep apnea syndrome is deeply and fully investigated in order to provide people with more information about sleep health and improve their sleep quality.Traditional detection method such as polysomnography(PSG)may disturb people’s sleep and equipment is easy to fall off during the detection.In this paper,we propose a detection method based on non-contact sleep monitoring system.Firstly,the method obtains a heartbeat interval sequence by a heart rate extraction algorithm applied to Ballistocardiogram of different humans.Based on time domain features,frequency domain features,and similar features of the heartbeat interval sequence over a fixed length of time,a classification model is applied to determine whether sleep apnea occurs during that time.Under the condition of less space complexity,the detection algorithm of sleep apnea still has comparable performance using fewer features.The sensitivity,specificity and accuracy of sleep apnea detection are further improved under the help of ensemble.Sleep apnea patient can be judged from his characteristics during the whole night.The sleep apnea can be detected without changing people’s sleep environment and disturbing people’s sleep process.The proposed method has a broad application prospects in the field of family health management.