Abstract
Acquiring Electroencephalography (EEG) data is often time-consuming, laborious, and costly, posing practical challenges to train powerful but data-demanding deep learning models. This study proposes a surrogate EEG data-generation system based on cycle-consistent adversarial networks (CycleGAN) that can expand the number of training data. This study used EEG2Image based on a modified S-transform (MST) to convert EEG data into EEG-topography. This method retains the frequency-domain characteristics and spatial information of the EEG signals. Then, the CycleGAN is used to learn and generate motor-imagery EEG data of stroke patients. From the visual inspection, there is no difference between the EEG topographies of the generated and original EEG data collected from the stroke patients. Finally, we used convolutional neural networks (CNN) to evaluate and analyze the generated EEG data. The experimental results show that the generated data effectively improved the classification accuracy.
Highlights
GENERALLY speaking, any form of communication between people and the external environment needs the involvement of peripheral nerves and muscles
This study examines the feasibility of using CycleGAN to generate surrogate motor-imagery (MI) EEG data of stroke patients for improving the performance of Deep learning (DL)-based classifiers
CycleGAN, which is equivalent to two Generative Adversarial Networks (GAN)’s, has two generators and two discriminators. This structure can restore the generated EEG topographies to their original data (i.e the EEG topographies of the healthy subjects) through the other generator to ensure the universality of the generator in the whole distribution domains
Summary
GENERALLY speaking, any form of communication between people and the external environment needs the involvement of peripheral nerves and muscles. A Brain-Computer Interface (BCI) is an intermediary of communication between people with nerve or muscle damage and nature. Instead of depending on peripheral nerves and muscles, the BCI converts the signals generated by the brain into the output and directly transmits the user's intention to external devices[1,2]. The whole BCI system includes signal acquisition, signal processing, and application (Fig. 1). Electrical signals generated by the brain can be detected either invasively or non-invasively. Electroencephalography (EEG) is non-invasive brain signals obtained by EEG electrodes and equipment[3]
Talk to us
Join us for a 30 min session where you can share your feedback and ask us any queries you have
More From: IEEE Transactions on Neural Systems and Rehabilitation Engineering
Disclaimer: All third-party content on this website/platform is and will remain the property of their respective owners and is provided on "as is" basis without any warranties, express or implied. Use of third-party content does not indicate any affiliation, sponsorship with or endorsement by them. Any references to third-party content is to identify the corresponding services and shall be considered fair use under The CopyrightLaw.