Electrode selection method based on brain-derived imaging and correlation analysis

Abstract

The invention discloses an electrode selection method based on brain-derived imaging and correlation analysis. The electrode selection method includes the steps that a base electrode set is determinedaccording to initial electrode configuration of an electrode cap; then collected motor imagery electroencephalogram signals are preprocessed by band-pass filtering; then a standardized low-resolutionbrain electromagnetic tomography imaging algorithm is used for brain-derived imaging of MI-EEG, and a time series of a brain-derived domain dipole amplitude is obtained; then according to a dipole amplitude peak of each experiment, a cortical activated area is determined, and the time series of the dipole amplitude in the activated area and Pearson correlation coefficients of the MI-EEG signals of each electrode are calculated and arranged in descending order; and finally, an optimal electrode set is formed by selecting an electrode with a high correlation coefficient to be combined with thebase electrode set. According to the electrode selection method, electrodes with weak correlation with an imaginary task and disadvantageous to classification are excluded, and great significance forimproving the computational efficiency and the experimental convenience is achieved.

Description

technical field [0001] The invention belongs to the field of electrode selection in a brain-computer interface (brain-computer interface, BCI) system, in particular to an electrode selection method based on the combination of EEG Source Imaging (ESI) and correlation analysis, using standardized low-resolution The sLORETA algorithm transforms the EEG signal in the sensor domain to the brain source domain, determines the activation area of ​​the cerebral cortex based on data-driven (Data-driven), and calculates the time series of dipole amplitudes in each electrode and activation area The correlation coefficient of the optimal electrode group selection. Background technique [0002] Brain-computer interface (BCI) system is a new mode of human-computer interface based on EEG signals. BCI does not rely on the normal output channels of human peripheral nerves and muscle tissue, but directly establishes a control pathway between the brain and external communication devices. With...

AI Technical Summary

Problems solved by technology

[0007] In view of the above shortcomings, the present invention proposes an electrode selection method based on ESI and correlation analysis, that is, the sLRC method, which uses brain source imaging and correlation analysis to select electrodes, and uses uniformly distributed basic electrode groups with high correlation The optimal electrode group is composed of a combination of electrodes with positive characteristics. On the basis of ensuring the uniform distribution of electrodes, electrodes with strong correlation with the dipole in the activation area are added to effectively retain the brain-source information to the greatest extent while excluding those with weak correlation, Electrodes that are not conducive to classification improve the efficiency of calculation and the convenience of experiments, and improve the shortcomings that the selection of electrodes in the sensor domain cannot well reflect the real electrical activity of cortical neurons and the randomness and subjectivity of the selection of electrodes in the brain source domain

Images