Research On Detection Method Of REM Sleep Behavior Disorder

In recent years,the importance of sleep-related diseases has increased significantly.Sleep diseases will reduce the body's immune system,induce various diseases,and have serious harm to human health.REM sleep disorder is one kind of sleep diseases.The disease patients will appear the phenomenon of action related to dreams during REM sleep.which severely affects the patient and other people.Multi-brain sleep monitoring is a gold standard to evaluate the quality of sleep all night,it is also the primary task of treating sleep diseases,but this method is highly specialized and complex,and it is difficult to apply to daily sleep monitoring.In order to solve the problem of sleep disease detection,this thesis proposed a REM sleep behavior disorder detection method.By combining brain wave data and EMG signals,this method can effectively select the REM sleep period.This method has three steps.The first step is to obtain brain wave data,using a series of Preprocessing operations to calculate three eigenvalues,these three eigenvalues include edge spectrum frequency EFSd,absolute power AP and relative power RP.The second step is to design a REM sleep judgment algorithm,the algorithm can analyze the possibility of REM sleep by comparing the eigenvalues with the thresholds,the last step is to effectively detect the possibility of patient's actions in REM sleep according to REM sleep judgment results.According to this detection method,this thesis designs the prototype system,which is mainly divided into brain wave module and EMG processing module.Brain wave module uses TGAM equipment,EMG module uses EMG sensor and Arduino development board.Experiments show that using brain wave module to distinguish between REM sleep and non-REM sleep is effective,and the correct rate is about 89.348%,it can effectively detect REM sleep.EMG detection module can make an effective judgment for the sleep action,and the detection accuracy is 85%.The system designed in this thesis can effectively detect REM sleep behavior disorder.The accuracy rate of detection is 83.5%.