Phenotype discovery of traumatic brain injury segmentations from heterogeneous multi-site data
We harmonized 5,600 subjects from the NIH’s FITBIR repository to identify phenotypes of traumatic brain injury.
We harmonized 5,600 subjects from the NIH’s FITBIR repository to identify phenotypes of traumatic brain injury.
We developed DeepFixel, a deep learning approach for separating crossing white matter fibers in diffusion MRI data.
We developed a method for adaptive spatial smoothing based on the vasculature in the brain.
We developed an atlas for adaptive smoothing in the brain based on white matter tracts.
We found and corrected a bias in quantitative T1 methods using a patch-based deep learning model.
We published our work on creating eye atlases from low-resolution MRI.
One password to rule them all, one password to find them, one password to bring them all and in the darkness bind them.