Shaun K.L. Quah

This study evaluates deep learning models for predicting task activation from resting-state fMRI, finding that architectural changes improve efficiency without sacrificing accuracy.

This paper validates data-driven factor models over RDoC-based models using fMRI and behavioral data, showing better reproducibility and clinical relevance across neural and behavioral domains.

This study compares changes in linear and nonlinear functional connectivity in MDD patients after psilocybin or escitalopram treatment, finding distinct patterns between responders and non-responders.

This study validates the RDoC framework using a latent variable approach, showing a bifactor model better fits fMRI data than the current framework and highlights underrepresentation of the arousal and regulatory systems domain.