A Novel GAN-Augmented EEG Classification Framework Using Graph Convolutional Networks for Autism Spectrum Disorder Detection

Melinda Melinda; Yuwaldi Away; Syahrul Gazali; W.K. Wong; Kaysa Marisa; Muhammad Mirza Rahmat

doi:10.1109/ICAIDES67265.2025.11404044

A Novel GAN-Augmented EEG Classification Framework Using Graph Convolutional Networks for Autism Spectrum Disorder Detection

Closed

Melinda Melinda, Yuwaldi Away, Syahrul Gazali, W.K. Wong, Kaysa Marisa, Muhammad Mirza Rahmat

2025 2025 International Conference on Applied Artificial Intelligence, Data Engineering and Sciences, ICAIDES 2025 Conference paper Cited by 0

Abstract

Autism spectrum disorder (ASD) is a developmental disturbance of the nervous system that has been identified through observation, behavior, and clinical interviews, which is subjective and requires a long time. This study aims to design a classification model using a Graph Convolutional Network (GCN) based on Electroencephalogram (EEG) signals, which have been converted into a heatmap image to represent activity visually. To enrich the training data, the Generative Adversarial Network (GAN) method is used to generate synthetic images. Test results show that the trained GCN model with additional data from GAN (learning rate 0.0005) yields results that are further evaluated as being taller compared to models that only use the original data. The model with additional GAN data achieved an accuracy of 94.37%, a precision of 94.42%, a specificity of 91.63%, a recall of 94.37%, and an F1-score of 94.36%. Meanwhile, the model with original data produces an accuracy of 93.90%, a precision of 93.92%, a specificity of 89.65%, a recall of 93.90%, and an F 1-score of 93.89%. This demonstrates that implementing the GAN method in the data augmentation process enhances the performance of the GCN model in more detailed classification of the ASD objective, thereby improving accuracy. © 2025 IEEE.

Affiliations

Universitas Syiah Kuala, Faculty of Engineering, Department of Electrical and Computer Engineering, Banda Aceh, Indonesia; Universitas Syiah Kuala Banda Aceh, Indonesia, Mahar Mardjono National Brain Center Hospital, Department of Neurology, Jakarta, Indonesia; Curtin University Malaysia, Faculty of Engineering and sciences, Department of Electrical and Computer Engineering, Sarawak, Malaysia

View full text (DOI) →

Research at a Glance

Research at a Glance

Topics & SDG Alignment

Collaboration

Author Profile (Selected)

References Overview

Journal & Source

Metadata & Integrity