DWI multiband sequence from CMRR (Minnesota) on Prisma does not support storage of complex data
Are you sure? I’m doing this on our Prisma here. It’s a very general setting in the protocol editor that exists at a lower level than the particulars of any specific sequence. You need to change “Reconstruction” to “Magnitude and phase”, and then convert the resulting two DICOM series to native complex data using mrcalc, specifically scaling the phase data to Radians and then using the -polar option that was added to the dev branch in #1659.