[2603.26821] Epileptic Seizure Prediction Using Patient-Adaptive Transformer Networks

Computer Science > Machine Learning arXiv:2603.26821 (cs) [Submitted on 26 Mar 2026] Title:Epileptic Seizure Prediction Using Patient-Adaptive Transformer Networks Authors:Mohamed Mahdi, Asma Baghdadi View a PDF of the paper titled Epileptic Seizure Prediction Using Patient-Adaptive Transformer Networks, by Mohamed Mahdi and 1 other authors View PDF HTML (experimental) Abstract:Epileptic seizure prediction from electroencephalographic (EEG) recordings remains challenging due to strong inter-patient variability and the complex temporal structure of neural signals. This paper presents a patient-adaptive transformer framework for short-horizon seizure forecasting. The proposed approach employs a two-stage training strategy: self-supervised pretraining is first used to learn general EEG temporal representations through autoregressive sequence modeling, followed by patient-specific fine-tuning for binary prediction of seizure onset within a 30-second horizon. To enable transformer-based sequence learning, multichannel EEG signals are processed using noise-aware preprocessing and discretized into tokenized temporal sequences. Experiments conducted on subjects from the TUH EEG dataset demonstrate that the proposed method achieves validation accuracies above 90% and F1 scores exceeding 0.80 across evaluated patients, supporting the effectiveness of combining self-supervised representation learning with patient-specific adaptation for individualized seizure prediction. Subjects: Ma...