Pure idea control intelligent rehabilitation method and application based on distillation learning and deep learning

Abstract

A pure mind control intelligent rehabilitation method based on distillation learning and deep learning, including: constructing a teacher network and a student network according to the collected EEG signals of existing subjects, including constructing a spatiotemporal convolution block, constructing a teacher network, and constructing The student network; pre-training the teacher network by using the collected subjects' EEG signals; pre-training the student network on the basis of the pre-trained teacher network; pre-training the student network on the basis of the pre-trained student network Carry out retraining; collect the EEG signals of the subjects through the intelligent rehabilitation system controlled by pure ideas, and identify the input EEG signals on the basis of the trained student network. When the subjects corresponding to the EEG categories are identified After the exercise intention of the subject, the exoskeleton rehabilitation equipment of the intelligent rehabilitation system is controlled by pure ideas to assist the subject to perform the corresponding body movements and complete the rehabilitation training. The invention can realize active rehabilitation training for users.

Description

technical field [0001] The invention relates to an intelligent rehabilitation system controlled by pure ideas. In particular, it relates to an intelligent rehabilitation method and application based on distillation learning and deep learning to control intelligent rehabilitation. Background technique [0002] EEG signals are the overall reflection of the electrophysiological activities of brain nerve cells on the surface of the cerebral cortex or scalp. EEG signals contain a large amount of physiological and disease information. By analyzing EEG signals, information related to certain diseases can be obtained, so that effective methods can be provided for brain disease diagnosis and rehabilitation based on these information. Therefore, the analysis and processing of EEG signals are widely used in clinical neurology research, rehabilitation medicine and other fields. In the field of brain-computer interface, deep learning models have been widely used in the analysis and pro...

AI Technical Summary

Problems solved by technology

The main problem is the balance between the complexity and accuracy of the deep learning model

In the classification of EEG signals, deep learning models often have better classification performance than traditional classification models, but their model structures are often more complex

A deep learning network with high model complexity will have two main negative effects in practical applications: high computational power consumption and weak real-time responsiveness

These two effects are also the two bottlenecks that have restricted the deployment of deep learning models to mobile terminals or embedded devices.

Images