I have only used the HCP database for dMRI analysis before. Now I need to collect my own data, so I would like to ask for advice on the current standard Diffusion MRI scan parameters. The expected scan time is approximately 10-15 minutes.
Do I need to scan two runs? AP and PA。
The regions of interest are focused on the frontal lobe and the default network.
The scanning machine is a 3T prisma mri.
You’re unlikely to get a consensus on acquisition parameters. Different people may choose to prioritize different things, or tune for a specific model, and have different time allocations.
The HCP-Aging protocol is designed for Prismas, so might be a reasonable starting point for you. From memory it’s > 20 minutes, but you could get that down to < 15 minutes by modifying the gradient table. Personally I’d suggest splitting the DWI gradient directions between two phase encoding directions rather than duplicating all DWI gradient directions across both phase encoding directions. The recon won’t be as good in regions of strong susceptibility artifacts, and you won’t have the same SNR, but it’d still be decent. Also you can extract the b=0’s from the start of those two series for inhomogeneity field estimation.
I’ll myself have a suggested DWI acquisition released early next year, but the project has not yet set the protocol in stone.