SemioKAT-FL: explainable, background-invariant, and privacy-preserving video seizure detection with a Kolmogorov–Arnold Transformer and federated learning

Automatic detection of epileptic seizures from video promises non-invasive, continuous patient monitoring, yet web-sourced seizure corpora are dominated by background confounds: recording-specific cues (studio, home, clinic) that are spuriously correlated with the seizure label. Models that appear highly accurate under conventional random splits may therefore be recognizing studios rather than seizures, and they remain opaque and dependent on centralizing sensitive patient video. We present SemioKAT-FL, a framework that addresses all three issues jointly.

Spatial features from an ImageNet-pretrained MobileNetV2 encoder are temporally modeled by a Kolmogorov–Arnold Transformer (KAT) that couples multi-head self-attention with learnable B-spline (KAN) feed-forward layers; a temporal-attention pooling layer yields per-second saliency; and an adversarial background-invariance head, driven by a gradient-reversal layer, penalizes any ability to recover the source recording. Training uses Federated Averaging so that raw video never leaves a site. On a corpus of 1,148 clips from 275 source recordings evaluated under leakage-controlled subject-wise (GroupKFold) cross-validation, we first show that a faithful re-implementation of the reference method drops from 99.5% under random splits to 94.9% subject-wise, and that a background-only model still reaches 90.4%—direct evidence of confounding.

The adversarial head substantially reduces a source-recovery probe from 70% to 50% (chance \(\approx 1.3\%\)); the representation retains reduced but non-negligible source information, which we characterize as partial rather than complete invariance. The full calibrated ensemble attains \(96.6\%\pm 1.8\%\) accuracy (F1 \(=96.8\%\), AUC \(=0.93\), expected calibration error 0.02–0.03) under source-level GroupKFold, and the federated model—trained on simulated site partitions without formal differential privacy—matches its centralized counterpart within 0.4%. To our knowledge, SemioKAT-FL is the first video seizure detector that jointly addresses background confounding through adversarial invariance, provides intrinsic explainability, and avoids centralization of raw video through federated training.

Faculty of Computer and Information Technology, Sana’a University, Sana’a, Yemen Department of Computer Science, Dr. Babasaheb Ambedkar Marathwada University, Chhatrapati Sambhajinagar, India Alhasan A. Alharbi & Mukta Dhopeshwarkar Department of Cyber Security, Faculty of Information Technology, Ajloun National University, P.O. 43, Ajloun, 26810, Jordan The authors declare no competing interests.

This study uses a publicly available dataset of video clips hosted on Kaggle. The videos were originally collected from publicly accessible online platforms (e.g., YouTube) where they were uploaded by patients, caregivers, or medical educators. No protected health information (PHI) or personally identifiable metadata accompanies the dataset.

Eyes are blurred in any released frames to further protect privacy. Because the study analyzes only pre-existing, de-identified, publicly available data and does not involve direct interaction with human subjects, prospective recruitment, or collection of new data, it is exempt from institutional review board (IRB) approval under the common exemption for secondary analysis of publicly available, de-identified datasets. No ethics approval was required or sought.