Evaluation of diffusion imaging protocols for the Alzheimer's disease Neuroimaging Initiative

Diffusion-weighted imaging (DWI) provides information on brain integrity and connectivity, based on mapping the directional diffusion of water along the brain's neural pathways. The latest stage of the Alzheimer's Disease Neuroimaging Initiative, ADNI-2, added a 9-minute 41-gradient DWI protocol. To study the potential benefits of adjusting the number of diffusion-weighted gradients, here we scanned 3 young adults with the ADNI DWI protocol and 3 variants reducing the voxel size, increasing the repetition time (TR), and increasing the number of diffusion-weighted directions. In a small pilot evaluation, we compared the signal-to-noise ratio (SNR) of diffusion tensor-derived fractional anisotropy (FA), as well as three standard graph theory measures, to assess protocol effects on anatomical network efficiency, clustering and modularity. The protocol with smaller voxels gave rise to more detected fibers in whole-brain tractography, but no protocol advantages were detectable in the small sample; all gave comparable FA SNR, connectivity and network measures.