Steady state visual evoked potential detection method based on beamforming and cca

Abstract

The invention discloses a steady-state visual evoked potential detection method based on beamforming and CCA, and belongs to the technical field intersecting cognitive neuroscience and automatic control. In order to further improve the classification accuracy of the brain-computer interface based on the steady-state visual evoked potential, the present invention uses canonical correlation analysis as a systematic classification method to output the brain-computer interface in the early stage, and uses the steady-state visual evoked potential data at this stage As the training data for beamforming; after the beamforming builds a stable activation template, the beamforming and canonical correlation analysis are used together as a systematic classification method for brain-computer interface output. The present invention utilizes the mixed mode of beamforming and canonical correlation analysis to give full play to the advantages of canonical correlation analysis without training and high classification accuracy of beamforming, and can realize a brain-computer interface system with high detection rate without training.

Description

technical field [0001] The invention discloses a steady-state visual evoked potential detection method based on beamforming and CCA, in particular relates to a brain-computer interface system using the steady-state visual evoked potential as a paradigm, and belongs to the technical field intersecting cognitive neuroscience and automatic control. Background technique [0002] Steady-State Visual Evoked Potentials (SSVEP) is a commonly used paradigm of Brain-Computer Interface (BCI). The SSVEP signal refers to the electroencephalography (EEG) signal excited in the user's visual cortex when the user looks at a visual stimulus that flickers at a fixed frequency. The SSVEP evoking paradigm consists of multiple visual stimuli flickering at different frequencies (hereinafter referred to as SSVEP stimuli). Define each time a BCI user gazes at a SSVEP stimulus and outputs an intention as a trial. In a certain trial, when the user fixates on a fixed frequency f j When the visual st...

AI Technical Summary

Problems solved by technology

This traditional beamforming method must be trained with a certain number of data sets to obtain an effective spatio-temporal beamformer group, which cannot meet the plug-and-play requirements.

And once the spatio-temporal beamformer group is obtained through training, the dynamic update of the spatio-temporal beamformer group cannot be realized. If the effect of the spatio-temporal beamformer group is not good, it will always affect the classification effect of the subsequent system

[0005] In summary, the templates in the CCA method are not representative of the description of EEG physiological signals, which leads to the need to improve the classification effect; and the beamforming method needs to be trained before it can be used, and the lack of the ability to dynamically update the training model leads to the need to improve the classification effect

Images