Volume V | Issue 7 | 2026 | Paper ID: IJATEM-V05I07P4

A Dynamic Graph Neural Network-Based Framework for EEG Signal Classification in Eye State Detection

Author(s)

, S. Prakash

Abstract

Accurate Eye State Detection (ESD) from Electroencephalogram (EEG) signal are important for understanding human visual attention, monitoring cognitive states, and developing intelligent human–computer interaction systems. However, EEG signals are highly nonlinear, non-stationary, in addition susceptible to noise, outliers, and temporal variations for ESD challenging. To overcome these issues, this paper proposes a Dynamic Graph Neural Network (DGNN)-based framework for EEG signal classification in ESD. The collected EEG eye-state dataset thereby a pre-processing based on data cleaning is executed to eliminate inconsistent samples, outlier removal to reduce the influence of abnormal signal observations. Z-score normalization is applied to standardize the EEG features and provide a consistent input distribution for subsequent learning. The normalized EEG signals are divided into temporally meaningful segments using a sliding-window segmentation technique. The segmented EEG data are then transformed into a dynamic graph representation, where EEG signal features are modelled as graph nodes and their relationships are represented through graph edges. The proposed DGNN performs graph-based feature learning by aggregating information from interconnected nodes whereas considering their changing relationships. The experimental outcome is implemented in Python software using EEG eye state dataset have the higher accuracy of 95.6%, precision of 95.4%, recall of 94.5%, F1-score of 98%, AUC of 99.2 and Cohen kappa of 0.91 using proposed DGNN for EEG signal classification in ESD. The proposed DGNN provide an effective representation of the complex spatiotemporal characteristics of EEG signals and improve the reliability of eye state classification.

References

[1] Yi Zhong; Zhenyu Wang; Xi Zhao; Tianheng Xu; Ting Zhou; Honglin Hu, Year: 2026, “Electroencephalogram-based multimodal attention level classification using deep learning techniques”, Frontiers in Human Neuroscience, Vol: 20, pp. 1791677.

CrossRef Google Scholar Publisher Link
[2] Sarunya Kanjanawattana; Isaman Sangbamrung; Dulyawat Wiriyaphong; Gun Bhakdisongkhram, Year: 2026, “EEG-Based Gait Classification in Stroke Patients Using Deep Learning”, Computers, Vol: 15, No: 6, pp. 392.

CrossRef Google Scholar Publisher Link
[3] Zhongyi Jiang; Ying Zhou; Yihan Zhang; Guanzhong Dong; Yun Chen; Qiaoyang Zhang; Ling Zou; Yin Cao, Year: 2024, “Classification of depression using machine learning methods based on eye movement variance entropy”, IEEE Access, Vol: 12, pp. 146107-146120.

CrossRef Google Scholar Publisher Link
[4] Pankaj Kunekar; Mukesh Kumar Gupta; Pramod Gaur, Year: 2024, “Detection of epileptic seizure in EEG signals using machine learning and deep learning techniques”, Journal of Engineering and Applied Science, Vol: 71, No: 1, pp. 21.

CrossRef Google Scholar Publisher Link
[5] Sumathi Balakrishnan; Raja Kumar Murugesan; Humaira Ashraf; Eng Lye Lim; Amna Faisal, Year: 2026, “EEGMetaNet: An enhanced Major Depressive Disorder detection framework using electroencephalogram and one-dimensional convolutional neural network” Applied Sciences, Vol: 16, No: 15, pp. 7564.

CrossRef Google Scholar Publisher Link
[6] Touseef Ur Rehman; Madallah Alruwaili; Muhammad Hameed Siddiqi; Yousef Alhwaiti; Sajid Anwar; Zahid Halim; Maaz Alam, Year: 2025, “Advancing EEG-based biometric identification through multi-modal data fusion and deep learning techniques”, Complex & Intelligent Systems, Vol: 11, No: 9, pp. 398.

CrossRef Google Scholar Publisher Link
[7] Şeniz Harputlu Aksu; Erman Çakıt; Metin Dağdeviren, Year: 2024, “Mental workload assessment using machine learning techniques based on EEG and eye tracking data”, Applied Sciences, Vol: 14, No: 6, pp. 2282.

CrossRef Google Scholar Publisher Link
[8] Soyiba Jawed; Ibrahima Faye; Aamir Saeed Malik, Year: 2024, “Deep learning-based assessment model for real-time identification of visual learners using raw EEG”, IEEE Transactions on Neural Systems and Rehabilitation Engineering, Vol: 32, pp. 378-390.

CrossRef Google Scholar Publisher Link
[9] Hepseeba Kode; Khaled Elleithy; Laiali Almazaydeh, Year: 2024, “Epileptic seizure detection in EEG signals using machine learning and deep learning techniques”, IEEE access, Vol: 12, pp. 80657-80668.

CrossRef Google Scholar Publisher Link
[10] Nayab Bashir; Sanam Narejo; Bushra Naz; Fatima Ismail; Muhammad Rizwan Anjum; Ayesha Butt; Sadia Anwar; Ramjee Prasad, Year: 2025, “A machine learning framework for major depressive disorder (MDD) detection using non-invasive EEG signals”, Wireless personal communications, Vol: 140, No: 1, pp. 39-61.

CrossRef Google Scholar Publisher Link
[11] Souhaila Khalfallah; William Puech; Mehdi Tlija; Kais Bouallegue, Year: 2025, “Exploring the effectiveness of machine learning and deep learning techniques for EEG signal classification in neurological disorders”, IEEE Access, Vol: 13, pp. 17002-17015.

CrossRef Google Scholar Publisher Link
[12] Abdullah Alghamdi; Mehrbakhsh Nilashi; Rabab Ali Abumalloh; Hossein Ahmadi; Mesfer Alrizq; Sultan Alyami; Waleed Abdu Zogaan; Fatima Khan Nayer, Year: 2026, “Accuracy improvements for electroencephalography (EEG) eye state classification using eXtreme gradient boosting and cluster ensembles”, Journal on Advances in Signal Processing, Vol: 2026, No: 1, pp. 20.

CrossRef Google Scholar Publisher Link

Full Length Article PDF

Article Details

A Dynamic Graph Neural Network-Based Framework for EEG Signal Classification in Eye State Detection

Author(s)

Abstract

References

Quick Info