A gender classification method based on resting-state EEG data

Abstract

The present invention relates to a gender classification method based on resting-state EEG data, comprising the following steps: S1, collecting original resting-state EEG data corresponding to each gender; S2, preprocessing the original resting-state EEG data, Get the resting-state EEG data with artifacts removed; S3, recombine the resting-state EEG data without artifacts; S4, construct a convolutional neural network, and input the reconstructed resting-state EEG data with artifacts removed to The convolutional neural network is used to train and test the convolutional neural network to obtain a trained convolutional neural network; S5, the trained convolutional neural network is used to classify the actual resting state EEG data. Compared with the prior art, the present invention de-artifacts and reorganizes resting-state EEG data, and extracts and analyzes features through convolutional neural networks based on the features of brain functional connections, which not only reduces the need for EEG data preprocessing It also solves the problem of not being able to choose a suitable model for EEG data.

Description

technical field [0001] The invention relates to the technical fields of deep learning neural network and brain science, in particular to a gender classification method based on resting-state EEG data. Background technique [0002] In neuroscience, there have been numerous studies showing differences in the brains of different sexes. In the study of sex differences in the brain, fMRI technology is usually used to study the differences in the structure, chemistry and functional connectivity of the brains of different sexes. The existing research results show that the brain volume of men is generally larger than that of women However, the proportion of gray matter in the female brain is higher, while the proportion of white matter in the male brain is higher; at the same time, the overall blood flow in the female brain is also significantly higher than that of males; as for the network connections of the brain, the connections in the male brain are mostly left and right The co...

AI Technical Summary

Problems solved by technology

[0007] To sum up, these two types of studies, brain composition and brain functional differences, can be analyzed by fMRI or task-state fMRI, but differences in brain functional connectivity cannot be analyzed using existing fMRI techniques. In fact, brain The functional connection of the EEG data can be reflected through the activation process of the resting state EEG data, but because the EEG data has no standardized collection steps, it is difficult to preprocess, and the performance is abstract, so the machine learning method is used to analyze the EEG data. Sometimes, it is impossible to find a suitable model: firstly, the EEG data acquisition process will be accompanied by a lot of noise, and it is difficult to filter it in the preprocessing process; secondly, when choosing a machine learning classifier, such as SVM or Decision tree, different data preprocessing will have different effects on the results of the model, making the accuracy of the model results not high, which also leads to the use of resting state EEG data for gender classification research has not yet achieved results

Images