Computational Neuroimaging for Individualized Brain Function
Predicting individualized brain function from resting-state fMRI.
At Stanford, I model how individual brains respond to tasks and naturalistic stimuli, aiming for interpretable markers relevant to cognition and mental health.
- Deep learning models predict individualized, task-evoked brain activity from resting-state functional connectivity (Human Brain Mapping, 2026; co-first author).
- Dynamic responses to naturalistic movie-watching modeled with graph- and topology-based methods to derive interpretable neural markers.
- Validated with spatial similarity, Dice overlap, cross-validation, and test-retest baselines rather than single-number metrics.