Cross-subject EEG cognitive state detection method based on efficient multi-source capsule network

Abstract

The invention provides a cross-subject EEG cognitive state detection method based on an efficient multi-source capsule network. The method comprises the following steps: aligning the feature distribution of a target domain and the feature distribution of a multi-source domain, so as to effectively migrate inter-domain features; constructing EEG into a multi-channel one-dimensional structure, so as to improve the training efficiency, and improve the model performance at the same time; secondly, introducing a self-expression module to capture potential relations between samples, so as to well adapt cross-subject EEG data analysis with significant individual differences under different tasks; and finally, providing a space attention algorithm based on a dynamic sub-capsule to further learn fine-grained feature information on the spatial level of the EEG data, and effectively describing the spatial relationship between parts and the partial-overall hierarchical relationship of the EEG data. According to the method, the individual difference problem of electroencephalogram signals in the field of brain cognitive calculation is effectively avoided, the method can be suitable for cognitive state recognition based on EEG under any task, the generalization ability is high, and the method can be well suitable for clinical diagnosis and practical application.

Description

technical field [0001] The present invention relates to the neural electrophysiological signal analysis technology in the field of brain cognitive computing, and the multi-source domain adaptive model construction method in the field of unsupervised learning. The way to check the state. The present invention can not only effectively solve the problem of significant individual differences among different subjects, but also effectively explain the internal mechanism of the correlation between EEG features and cognitive states. In addition, it can effectively improve the training efficiency while maintaining the performance of the capsule network. Background technique [0002] Electroencephalogram signal (EEG) is the signal that can best reflect the cognitive activity of the human brain, and is a key indicator of cognitive state detection tasks. In recent years, EEG-based cognitive state detection methods have attracted more and more research attention due to their high tempor...

AI Technical Summary

Problems solved by technology

[0006] Although the existing capsule-based EEG analysis methods have made rapid progress, there are still many challenges in cross-subject EEG analysis: first, the original dynamic routing algorithm of the capsule network is sensitive to outliers caused by significant individual differences in EEG; second, , although the dynamic routing process of the capsule network can describe the hierarchical relationship from the part to the whole, it cannot explain the interaction between the underlying parts

Specifically, for multi-channel EEG analysis, it is difficult to describe the interaction between channels

Images