If quantitative diffusion measures are acquired at different sites but identical hardware and protocol settings, it remains unclear if a site bias would require a data homogenization in order to pool the data for analysis. A traveling volunteer (including four subjects) approach was applied and different complex diffusion measure as well as diffusion tensor metrics were computed based on the measurements at two different sites. Our results suggest that the inter-site differences are much smaller than the inter-subject differences in the ROI based analysis. The voxel-wise analysis was found to be more susceptible to incomplete artifact compensation and registration errors.
This abstract and the presentation materials are available to members only; a login is required.