Brain-control hybrid intelligent rehabilitation method based on novel deep learning model

Abstract

A brain-controlled hybrid intelligent rehabilitation method based on a novel deep learning model comprises the steps of acquiring a motor imagery electroencephalogram signal of a testee through a brain-computer interface composed of an electrode cap and portable electroencephalogram signal collection equipment, and the motor intention of the testee is decoded through a decoding model; controllingrehabilitation equipment to assist the testee in performing limb movement according to the movement intention of the testee; wherein the decoding model is obtained by training a deep learning model through a training data set, and the decoding model is continuously updated along with advancing of rehabilitation training of the testee so as to adapt to electroencephalogram characteristic changes ofthe testee; wherein the training data set comes from a motor imagery database, and the motor imagery database is used for storing collected electroencephalogram signal samples of a testee and markingtime marks and action labels. According to the method, the decoding model can be updated in real time to adapt to electroencephalogram characteristic changes in the training process of a patient, sothat the classification effect of electroencephalogram signals is effectively improved, and a more efficient rehabilitation training effect is achieved.

Description

technical field [0001] The invention relates to a brain-controlled hybrid intelligent rehabilitation method. In particular, it involves a brain-controlled hybrid intelligent rehabilitation method based on a new deep learning model. Background technique [0002] Deep learning is a branch of machine learning, which has developed into a multi-field interdisciplinary subject, involving probability theory, statistics, approximation theory, convex analysis, computational complexity theory and other disciplines. It attempts to use multiple processing layers that contain complex structures or multiple nonlinear transformations to perform high-level abstraction on data. It is a method based on representation learning of data and is widely used in many fields, especially motor imagination brain-machine It has a wide range of application prospects in the interface system. Brain-computer interface technology provides us with an effective way to realize human-machine hybrid intelligenc...

AI Technical Summary

Problems solved by technology

For the classification algorithm of motor imagery, it is difficult to obtain the EEG data of patients, and the amount of effective data is small, so it is difficult to train a high-quality model

At the same time, the individual differences and inter-individual differences of EEG signals are relatively large, that is, the differences in the characteristics of EEG signals of different individuals and the differences of EEG signals of the same individual with the advancement of training are very large, resulting in poor classification performance. , thus affecting the accuracy of model training

Images