Continuous Dynamic Physiological Signal Anomaly Detection Algorithm

The continuous dynamic physiological signals of the human body record the physiological waveforms and vital signs data of the human body,including electrophysiological signals such as brain electricity,ECG,eye electricity,and myoelectricity,as well as vital signs parameters such as respiratory rate,blood oxygen saturation,and blood pressure.Among these continuous dynamic physiological signals,the human brain wave is a kind of bioelectric signal generated by the advanced central nervous system,which contains rich human physiological and pathological information,and has the characteristics of low signal-to-noise ratio,low frequency,and strong randomness.Using electroencephalogram(EEG)to judge the physiological state of the human body has become a mainstream research direction.In addition,as more and more people suffer from sleep disorders such as insomnia and dreaminess in recent years,many countries in the world are investing more and more in human sleep research.Realizing the assessment of sleep quality is the primary purpose of the study,and the assessment of sleep state,that is,sleep staging,is the prerequisite for realizing sleep quality assessment.At the same time,the first step in studying sleep-related diseases is also a very important step in sleep staging.Therefore,the study of sleep staging is of great significance,and the assessment of sleep quality and the diagnosis of sleep diseases have also become research hotspots.The main research content of this paper is the use of anomaly detection algorithm based on divergence to find anomalies in EEG signals,and combined with the anomaly detection algorithm,the use of neural networks to study sleep staging methods.First,this article discusses the research background and significance of abnormality detection and sleep staging,as well as their respective research status at home and abroad.Secondly,the background knowledge involved in this research is introduced,including related theories such as abnormality detection,EEG sleep and other related theories,and the key technology of artificial neural network.Then,the process of using anomaly detection algorithm based on divergence measurement to detect EEG signals,the principles and methods of each part are introduced,and optimization is made,and the result graph is given.Subsequently,the shortcomings of traditional sleep staging were pointed out,and the problem was solved by combining anomaly detection algorithm.Next,set up a one-dimensional convolutional neural network and formulate a training strategy.Use the network to classify EEG signals and realize sleep staging.Finally,89.3% and 85.8% of the data set on the SHHS1 data set and Sleep EDFx data set were obtained.Accuracy,and gives the statistical results of the performance of the anomaly detection algorithm and the comparison results of the sleep staging performance with other algorithms.Finally,this article combines the changes in social concepts and related laws and regulations in recent years to look forward to the development trend of sleep staging.