Deep Contrastive Learning for Seizure Onset Zone Localization
A self-supervised CNN-Transformer framework for automated ictal onset zone classification from SEEG spectrograms, with multi-center validation.
Portfolio
Seizure Propagation and Network Motif Reorganization in SEEG
Characterizing mesoscale directed network architecture during ictal evolution using three-node connectivity motifs in stereoelectroencephalography recordings.
Three-Phase Seizure Segmentation in SEEG
Semi-supervised changepoint detection for automated delineation of seizure onset, intra-ictal transition, and termination in stereoelectroencephalography recordings.
Spectral Graph Modeling of E/I Imbalance in Epilepsy
Biophysically-interpretable neural mass model parameters derived from resting-state MEG to identify excitatory-inhibitory imbalance in seizure onset zones.