top of page

Modeling Latent Signals

 

With collaborators from the Universiti Teknologi Malaysia (CM Ting and S Hussain), we develop a structured factor analytic model to study dependence in a default mode network from massive whole brain fMRI data. When compared to the naïve approach of summarizing ROI-based hemodynamic activity via averaging across voxels, our proposed factor analytic method gives high power and reduced false positive rates. 

bottom of page