Abstract:In the Motor Imagery (MI)-based Brain Computer Interface (BCI), users' intention is converted into a control signal through processing a specific pattern in brain signals reflecting motor characteristics. There are such restrictions as the limited size of the existing datasets and low signal to noise ratio in the classification of MI Electroencephalogram (EEG) signals. Machine learning (ML) methods, particularly Deep Learning (DL), have overcome these limitations relatively. In this study, three hybrid models were proposed to classify the EEG signal in the MI-based BCI. The proposed hybrid models consist of the convolutional neural networks (CNN) and the Long-Short Term Memory (LSTM). In the first model, the CNN with different number of convolutional-pooling blocks (from shallow to deep CNN) was examined; a two-block CNN model not affected by the vanishing gradient descent and yet able to extract desirable features employed; the second and third models contained pre-trained CNNs conducing to the exploration of more complex features. The transfer learning strategy and data augmentation methods were applied to overcome the limited size of the datasets by transferring learning from one model to another. This was achieved by employing two powerful pre-trained convolutional neural networks namely ResNet-50 and Inception-v3. The continuous wavelet transform (CWT) was used to generate images for the CNN. The performance of the proposed models was evaluated on the BCI Competition IV dataset 2a. The mean accuracy vlaues of 86%, 90%, and 92%, and mean Kappa values of 81%, 86%, and 88% were obtained for the hybrid neural network with the customized CNN, the hybrid neural network with ResNet-50 and the hybrid neural network with Inception-v3, respectively. Despite the promising performance of the three proposed models, the hybrid neural network with Inception-v3 outperformed the two other models. The best obtained result in the present study improved the previous best result in the literature by 7% in terms of classification accuracy. From the findings, it can be concluded that transfer learning based on a pre-trained CNN in combination with LSTM is a novel method in MI-based BCI. The study also has implications for the discrimination of motor imagery tasks in each EEG recording channel and in different brain regions which can reduce computational time in future works by only selecting the most effective channels.

A Transfer Learning Method based on VGG-16 Convolutional Neural Network for MI Classification

Classification of motor imagery electroencephalogram signals by using adaptive cross-subject transfer learning

Parameter transfer learning based on shallow visual geometry group network and its application in motor imagery classification

Transferring a deep learning model from healthy subjects to stroke patients in a motor imagery brain–computer interface

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

A Deep Transfer Convolutional Neural Network Framework for EEG Signal Classification

A transfer learning framework based on motor imagery rehabilitation for stroke

An Algorithm for Motor Imagery Classification Based on Transfer Learning and Feature Fusion

A multi-scale fusion CNN model based on adaptive transfer learning for multi-class MI-classification in BCI system

A transfer learning-based CNN and LSTM hybrid deep learning model to classify motor imagery EEG signals

Classification of Motor Imagery EEG Signals Based on Data Augmentation and Convolutional Neural Networks

Motor imagery electroencephalogram classification algorithm based on joint features in the spatial and frequency domains and instance transfer

Improving Motor Imagery EEG Classification by CNN with Data Augmentation

Three-stage transfer learning for motor imagery EEG recognition

Hybrid deep neural network using transfer learning for EEG motor imagery decoding

Adaptive transfer learning for EEG motor imagery classification with deep Convolutional Neural Network

Instance Transfer Subject-Dependent Strategy for Motor Imagery Signal Classification Using Deep Convolutional Neural Networks.

Improved Domain Adaptation Network Based on Wasserstein Distance for Motor Imagery EEG Classification

Transfer learning with data alignment and optimal transport for EEG based motor imagery classification

A Model Combining Multi Branch Spectral-Temporal CNN, Efficient Channel Attention, and LightGBM for MI-BCI Classification

A 1D CNN for high accuracy classification and transfer learning in motor imagery EEG-based brain-computer interface