Abstract:Stroke is the second foremost cause of death worldwide and is one of the most common causes of disability. Several approaches have been proposed to manage stroke patient rehabilitation such as robotic devices and virtual reality systems, and researchers have found that the brain-computer interfaces (BCI) approaches can provide better results. Therefore, the most challenging tasks with BCI applications involve identifying the best technique(s) that can reveal the neuron stimulus information from the patients' brains and extracting the most effective features from these signals as well. Accordingly, the main novelty of this paper is twofold: propose a new feature fusion method for motor imagery (MI)-based BCI and develop an automatic MI framework to detect the changes pre- and post-rehabilitation. This study investigated the electroencephalography (EEG) dataset from post-stroke patients with upper extremity hemiparesis. All patients performed 25 MI-based BCI sessions with follow up assessment visits to examine the functional changes before and after EEG neurorehabilitation. In the first stage, conventional filters and automatic independent component analysis with wavelet transform (AICA-WT) denoising technique were used. Next, attributes from time, entropy and frequency domains were computed, and the effective features were combined into time-entropy-frequency (TEF) attributes. Consequently, the AICA-WT and the TEF fusion set were utilised to develop an AICA-WT-TEF framework. Then, support vector machine (SVM), k-nearest neighbours (kNN) and random forest (RF) classification technique were tested for MI-based BCI rehabilitation. The proposed AICA-WT-TEF framework with RF classifier achieves the best results compared with other classifiers. Finally, the proposed framework and feature fusion set achieve a significant performance in terms of accuracy measures compared to the state-of-the-art. Therefore, the proposed methods could be crucial for improving the process of automatic MI rehabilitation and are recommended for implementation in real-time applications.

Stroke Motor Imagery Recognition based on EEG

Exploring Motor Imagery Eeg Patterns for Stroke Patients with Deep Neural Networks

A transfer learning framework based on motor imagery rehabilitation for stroke

Integrating Electroencephalography Source Localization and Residual Convolutional Neural Network for Advanced Stroke Rehabilitation

A tensor-based scheme for stroke patients' motor imagery EEG analysis in BCI-FES rehabilitation training.

Motor Imagery Classification Based on Deep Convolutional Neural Network and Its Application in Exoskeleton Controlled by EEG

EEG Neuro-markers to Enhance BCI-based Stroke Patients Rehabilitation

A multimodal electroencephalogram visualization system can promote stroke patients′ motor imagery

EEG feature fusion for motor imagery: A new robust framework towards stroke patients rehabilitation

A novel recognition and classification approach for motor imagery based on spatio-temporal features

A Deep Learning Model for Stroke Patients' Motor Function Prediction

A framework for motor imagery with LSTM neural network

Motor Imagery EEG Recognition Based on an Improved Convolutional Neural Network with Parallel Gate Recurrent Unit.

Classification Of Motor Imagery Eeg Signals With Deep Learning Models

Enhancing Motor Imagery Classification with Residual Graph Convolutional Networks and Multi-Feature Fusion

Time-Frequency-Space EEG Decoding Model Based on Dense Graph Convolutional Network for Stroke

Neurophysiological Substrates of Stroke Patients with Motor Imagery-Based Brain-Computer Interface Training

Classification of Motor Imagery Eeg Using Wavelet Envelope Analysis and LSTM Networks

Study Of Electroencephalogram Feature Extraction And Classification Of Three Tasks Of Motor Imagery

A Multifrequency Brain Network-Based Deep Learning Framework for Motor Imagery Decoding

Recognition of EEG Signal Motor Imagery Intention Based on Deep Multi-View Feature Learning.