Motor imagery electroencephalogram signal processing method and device and computer readable storage medium

Abstract

The invention discloses a motor imagery electroencephalogram signal processing method. The method comprises the steps that motor imagery electroencephalogram signals of a subject are collected to serve as a training data set; according to the training data set, obtaining a tensor model of the motor imagery electroencephalogram signals; according to the motor imagery electroencephalogram signal tensor model, space-time frequency tensor signal models of the motor imagery electroencephalogram signals in different action modes are constructed through tensor decomposition; optimizing filter parameters of the space-time frequency tensor signal model to obtain a space-time frequency filter with the highest feature separation degree; training the plurality of time-space frequency filters with the highest feature separation degree by adopting a machine learning method, and optimizing to obtain a time-space frequency filter based on multiple same tasks of the same subject; according to different individuals, parameters of the space-time frequency filter with the highest feature separation degree are preset, and the space-time frequency filter corresponding to the parameters is obtained through pre-learning; and classifying the motor imagery electroencephalogram signals of the subject by adopting a space-time frequency filter corresponding to the parameters obtained by pre-learning.

Description

technical field [0001] The present invention relates to the field of network communication, in particular to a method, device, equipment and computer-readable storage medium for processing motor imagery EEG signals. Background technique [0002] Motor imagery refers to imagining limb movements in the brain according to memory without any motor output, activating a specific area of ​​an activity, and is widely used in sports rehabilitation and other fields. The use of EEG signals for motor imagery recognition has gradually attracted people's attention in recent years, especially in the rehabilitation of paralyzed patients and the control of human-computer interaction, which has important practical application value. [0003] The brain is the high-level nerve center of the human body. The brain generates certain characteristic signals to realize communication and control with electronic equipment, which is called the Brain Computer Interface (BCI). The key to the brain-compute...

AI Technical Summary

Problems solved by technology

After CSP was proposed, a variety of improved methods based on CSP emerged, but the improvement methods were limited to constructing a filter set that is more in line with the characteristics of EEG signals or a more complete filter set to improve signal processing and pattern recognition in the time-space domain. The accuracy and robustness of the EEG signal pattern recognition accuracy is limited, still needs to be improved

Images