Cooperative Identification of Prolonged Motor Movement from EEG for BCI without Feedback

This repository contains the code used to recognise sequence of sub-gestures as well as complex fine movements (sub-gestures) from EEG-based continuous motor movement using the approach proposed in [1].

The code provided here was tested on MATLAB2022b and makes use of an adaptation of the MVGC v1.3 Toolbox [2] for the calculation of the Granger causality measures.

Citation

Please cite this repository as:

Falcon-Caro, A., Ferreira, J. F. & Sanei, S. (2024). Cooperative Identification of Prolonged Motor Movement from EEG for BCI without Feedback.[under review]

This resource is released under a MIT License.

Instructions for running DAOT code with your own data:

1. Download and add the MVGC Toolbox [2] to your MATLAB Path and initialize the Toolbox.
2. The Diffusion Adaptation with Orthogonal Target method (DAOT) can be found in file "DA_Modeling.m". You can run this file once your data has been preprocessed and the Granger causality and Bessel basis function dictionary obtained.

Instructions for running the examples code:

There are 3 examples: - Example with Dataset 1 [3] as described in the paper for the recognition of 2 sequences of sub-gestures, each with 4 sub-gestures, when the Bessel function dictionary is 2/3 the total number of sub-gestures. - Example with Dataset 1 [3] as described in the paper for the recognition of 2 sequences of sub-gestures, each with 4 sub-gestures, when the Bessel function dictionary is the same size as the total number of sub-gestures. - Example with Dataset 2 [4] as described in the paper for the recognition of 4 sub-gestures, when the Bessel function dictionary is the same size as the total number of sub-gestures.

Steps to follow: 1. Download and add the MVGC Toolbox [2] to your MATLAB Path and initialize the Toolbox. 2. Add all the provided files to your MATLAB Path. 3. Dowload and add the respective Dataset to your MATLAB Path. 4. Run the script associated with your desired example.

Abstract

This paper presents a novel approach for the recognition of a prolonged motor movement from a subject’s electroencephalogram (EEG) using orthogonal functions to model a sequence of sub-gestures. In this approach, the individual’s EEG signals corresponding to physical (or imagery) continuous movement for different gestures are divided into segments associated with their related sub-gestures. Then, a diffusion adaptation approach is introduced to model the interface between the brain neural activity and the corresponding gesture dynamics. In such a formulation, orthogonal Bessel functions are utilized to represent different gestures and used as the target for the adaptation algorithm. This method aims at detecting and evaluating the prolonged motor movements as well as identifying highly complex sub-gestures. This technique can perform satisfactory classification even in the presence of small data sizes while, unlike many regressors, maintaining a low computational cost. The method has been validated using two different publicly available EEG datasets. An average inter-subject validation accuracy of 98.10% is obtained for the smallest dataset during the classification of ten estimated sub-gestures.

Contact us

The easiest way to get in touch is via our GitHub issues.

You are also welcome to email us at aliciafalconcaro@gmail.com, to discuss this project, make suggestions, or just say "Hi"!

[1] Falcon-Caro, A., Ferreira, J. F. & Sanei, S. (2024). Cooperative Classification of Prolonged Movement from EEG for BCI without Feedback. [2] Barnett, L. and Seth, A. K. (2015). Granger causality for state-space models, Phys. Rev. E 91(4) Rapid Communication. [3] Falcon-Caro, A., Shirani, S., Ferreira, J. F., Bird, J. J., & Sanei, S. (2024). Formulation of Common Spatial Patterns for Multi-task Hyperscanning BCI. IEEE Transactions on Biomedical Engineering. doi: 10.1109/TBME.2024.3356665 [4] Cho, H., Ahn, M., Ahn, S., Kwon, M., & Jun, S. C. (2017). EEG datasets for motor imagery brain–computer interface. GigaScience, 6(7), gix034.