Estimation of Sleep Stages Based on Non-contact Physiological Signal

• 1、Sino-Dutch Biomedical and Information Engineering School of Northeastern University, Liaoning Shenyang,110004
• 2、Medical Image Computing Key Laboratory of Ministry of Education, Liaoning Shenyang,110004
• 3、Sino-Dutch Biomedical and Information Engineering School, Northeastern University, Shenyang 110004, China

Abstract：Purpose: To evaluate sleep quality, this manuscript designs and implements non-contact sleep monitoring system which can identify four sleep stages. Method: In order to take the place of the method such as pasting electrode with people, a non-contact system that includes hardware and software for collecting physiological signals is designed. The hardware includes the circuit's components that collect the signal from Radio Frequency Motion Sensor (RFMS), respiration and pulse wave. The software can implement signal processing and classify sleep stages. Sleep stage is identified through muti-information fusion based on analysis on body movement, respiratory and pulse wave. Results: Based on the designed hardware, the data will be saved as TXT formation in the disk, will be fused with body movement, respiratory and pulse wave. Sleep stages are divided into Wakefulness, REM, S1, S2. Compared with Tanita, the degree to match result of this system is 70%. Conclusions: After implementation of sleep staging, a report about sleep quality evaluation will be given, and the length of time of sleep phase will be statistic, therefore the report can provide some information on people's sleep quality in sleep monitoring, and then people can take some measures to improve the quality of sleep.

Keywords： Non-contact, Sleep monitoring, Respiration, Sleep stage, Heart rate variability, Body movement