Gia H. Ngo

This paper introduces BrainSurfCNN, a surface-based CNN that predicts task-based contrast maps from resting-state fMRI, achieving state-of-the-art accuracy and generalization.

This study shows deep neural networks can predict cortical responses to naturalistic stimuli across large brain areas beyond visual and auditory regions.

This paper applies an author-topic model to ALE-type meta-analyses to explore heterogeneity, revealing functional sub-domains and task-dependent co-activation patterns.

This paper evaluates methods for mapping between MNI152/Colin27 and fsaverage coordinate systems, finding that registration fusion with ANTs performs best for new subjects.