Research
The lab turns EEG, MEG, and SEEG recordings into interpretable markers of how the brain works in health and in neurological disease. We work across clinical and translational neuroscience, brain connectivity and network neuroscience, interpretable machine learning, and affective computing.
Research areas
Envelope-based changepoint segmentation of seizures, seizure-network dynamics for localisation, and presurgical MEG-derived network-control metrics in temporal lobe epilepsy, across non-invasive and intracranial (EEG, MEG, SEEG) recordings.
Nonlinear phase-based functional connectivity (phase-lag index, phase-locking value) mapping how distributed brain regions coordinate, together with interaction between the central and autonomic nervous systems through phase synchronization.
Data-driven whole-brain connectome markers of trait impulsivity, using connectome-based predictive modelling linking brain networks to individual differences in behaviour.
Explainable models (SHAP, feature-importance) and leakage-aware validation that reveal which physiological patterns drive a prediction, not accuracy alone.
Decoding emotional states from neural signals using advanced time–frequency analysis and machine learning.
