SpaTeoGL: Spatiotemporal Graph Learning for Interpretable Seizure Onset Zone Analysis from Intracranial EEG
arXiv:2602.11801v2 Announce Type: replace Abstract: Accurate localization of the seizure onset zone (SOZ) from intracranial EEG (iEEG) is essential for epilepsy surgery but is challenged by complex spatiotemporal seizure dynamics. We propose SpaTeoGL, a spatiotemporal graph learning framework for interpretable seizure network analysis. SpaTeoGL jointly learns window-level spatial graphs capturing interactions among iEEG electrodes and a temporal graph linking time windows based on similarity of their spatial structure. The method is formulated within a smooth graph signal processing framework
The continuous advancements in AI and deep learning research, particularly in graph-based methods, enable more sophisticated analysis of complex biological signals like iEEG for medical applications.
This development can significantly improve the accuracy and interpretability of seizure onset zone localization, leading to more effective surgical outcomes for epilepsy patients and advancing the field of AI in healthcare.
The interpretability aspect of SpaTeoGL offers a valuable tool for clinicians to understand AI's reasoning in identifying SOZ, fostering trust and adoption in critical medical decision-making.
- · Epilepsy patients
- · Neuroscience researchers
- · Medical AI developers
- · Healthcare technology providers
- · Traditional EEG analysis methods
- · Patients whose epilepsy is not amenable to surgical intervention
Improved surgical planning and outcomes for epilepsy patients due to more precise SOZ localization.
Accelerated development of other AI-driven diagnostic tools for neurological disorders leveraging similar spatiotemporal graph learning techniques.
Enhanced understanding of brain network dynamics underlying various neurological conditions, informing new therapeutic strategies.
This signal links to a primary source. Continuum Brief monitors and indexes it as part of the live intelligence stream — we do not republish source content.
Read at arXiv cs.LG