This study presents our harmonization efforts on the multi-site diffusion MRI data of the ~12,000 adolescents from the Adolescent Brain Cognitive Development (ABCD) study, collected from 22 sites using Siemens, GE and Philips scanners. On the minimally preprocessed diffusion MRI data provided by the ABCD study - release 3, we first applied brain masking using our deep learning tool which showed 99% Dice overlap performance compared to the manually corrected masks. Next, we harmonized the dMRI data from 22 sites (with 45 scanner settings) using our previously-validated software. The harmonized diffusion MRI data will be shared through the NIMH data archive.
This abstract and the presentation materials are available to members only; a login is required.