Training method and device for classification model of electroencephalogram signal, and electronic equipment

Abstract

Provided are a training method, a training device and electronic equipment for a classification model used for classification of electroencephalogram signals. The training method includes: obtaining an EEG signal sample set, the EEG signal sample set includes a plurality of sample subsets respectively corresponding to a plurality of categories; for each sample subset in the plurality of sample subsets, based on time-frequency domain transformation and Sample superposition, using each sample in the sample subset to generate a superimposed sample set corresponding to the sample subset; for each sample subset in multiple sample subsets, using the sample subset and its corresponding superimposed sample set to generate an expanded sample subset of the sample subset; and using the multiple expanded sample subsets respectively corresponding to the multiple categories to train a classification model for EEG signal classification. Wherein, the trained classification model is used to classify the collected electroencephalogram signals to be classified into one of multiple categories.

Description

technical field [0001] The present application relates to the medical field, and more specifically, to a method and device for training a classification model for classification of electroencephalogram signals, and electronic equipment. Background technique [0002] Brain-computer interface (BCI) is widely used in medical scenarios. It is a technology that directly communicates between the brain and external devices such as computers without relying on the normal output pathways of the brain (peripheral nerves and muscle tissue). For example, for a stroke patient with damage to the sensorimotor cortex, a brain-computer interface (BCI) can collect signals from the damaged cortical area, and then stimulate muscles or control orthotics to improve arm movement. The brain of epilepsy patients will have abnormal discharge of neurons in a certain area. After the abnormal discharge of neurons is detected through brain-computer interface technology, corresponding electrical stimulati...

AI Technical Summary

Problems solved by technology

However, the deep learning model requires a large amount of data for training, and the small amount of training data makes it difficult to train a high generalization model for the classification task of EEG signals

Images