Research And Implementation Of Depression Recognition Based On EEG

Although there are known effective treatments for depression,less than half of the world’s patients(less than 10%in many countries)have received these treatments.Due to the lack of biomarkers of depression,doctors often have a certain subjectivity in the diagnosis of depression.At present,there is no recognized auxiliary examination method that can be used to assist the diagnosis of depression.With the development of machine learning,artificial intelligence assisted diagnosis and treatment has become a hot research field in recent years.In this paper,we focus on the subjectivity and inconvenience of the current diagnosis of depression.We collected EEG data from a larger group of depressed and healthy people than in most similar studies and extracted a large number of features,and designed a depression recognition experiment based on EEG signals.The feature selection algorithm was used to select the EEG feature combinations with great differences between the depression patients and the healthy people,and the classification experiments were carried out to verify the results.Finally,this paper used selection algorithm of tree model,with the help of random forest model classifier,establishes a depression recognition model based on EEG signal,and achieves the effect of accurate recognition of depression.This machine learning model has been tested on the data set of this paper for 10 times 5-fold cross validation,and the average classification accuracy is 79.61%,which proves the possibility of EEG as a biomarker of depression recognition.On the premise of ensuring the recognition accuracy,this depression recognition model only uses the characteristics of 10 EEG channels as input,which reduces the number of EEG channels in the model input and effectively simplifies the depression recognition model.Based on the above machine learning model,this paper designs and implements a depression recognition system based on EEG signal.By inputting the file containing the specified channel EEG signal,the system can complete the characteristic calculation and analysis of EEG signal,so as to identify whether the people is a depression patient.The system also provides a detailed display function,users can view the specific values of the EEG features used by the recognition model and the comparison of the features with the mean values of the healthy control group and the depression group.The depression recognition system based on EEG signals provides an objective assistance and reference for the diagnosis of depression,and has important research value and potential application value in the auxiliary diagnosis of depression.