Research On The Algorithm Of Epileptic EEG Automatic Detection And Sleep EEG Automatic Staging

EEG is a main tool to record the electrophysiological activities of nerve cell groups in human brain,which is often used to detect neurological diseases.Taking patients with epilepsy and patients with sleep disorders as the research object,this paper uses intelligent algorithm to analyze EEG signals to realize the automatic detection of epilepsy and the automatic staging of sleep stage.Epilepsy is due to the abnormal discharge between neurons in the human brain,which leads to the functional disorder of the human brain.Sleep is a common physiological phenomenon.The brain sensitivity is low during sleep.Some latent diseases are easy to appear during sleep,and the timing of occurrence in the whole sleep process is also different.Therefore,analyzing the sleep cycle is the basis of studying sleep and related diseases,which has important clinical significance.The method of directly monitoring EEG in patients with epilepsy or sleep disorders is time-consuming and labor-consuming,and the detection efficiency is low.In recent years,the use of artificial intelligence algorithm to analyze EEG signals and assist doctors in diagnosis has gradually become the mainstream.The main work of this paper is as follows:1)Based on EEG data,an automatic detection algorithm of epileptic EEG based on common space mode is proposed in this paper.Firstly,wavelet packet decomposition and common space mode are combined to extract EEG features,and frequency band energy and standard deviation are selected as complementary features and combined.Then,particle swarm optimization algorithm is combined with support vector machine to adjust the super parameters in support vector machine and improve the performance of classifier.Finally,double classification is used to further improve the classification accuracy.This paper is verified on the epilepsy data set of Bonn University,and compared with the mainstream methods.The experimental results show that the proposed method can obtain better performance.2)In order to further verify the generality of this method,the above method is applied to the sleep stage staging task.On this basis,the preprocessing and classification methods are improved and verified on the sleep EDF data set.The results show that it is good,but there is still some room for improvement.Through the reflection and summary of the experiment,this paper plans to introduce the deep learning algorithm to complete the automatic staging task of sleep stage.3)An automatic sleep stage staging algorithm based on deep learning is proposed.Firstly,one-dimensional sleep EEG signal is transformed into two-dimensional form through wavelet packet transform,and convolutional neural network is selected as the basic network.Convolution attention module is added to the network to improve the network performance.The conventional convolution layer in the network is replaced by deep separable convolution to improve the efficiency of the network.Through experimental comparison,the results show that the method proposed in this paper has higher accuracy and efficiency,and provides a new research idea for automatic sleep staging task...