Hi HCP, In the fMRI Volume processing, we noticed that although the spin-echo fieldmaps are calculated using topup, there are a set of custom applywarp commands to generate distortion-corrected final-space fieldmaps instead of applywarp: (e.g. the DistortionCorrection script DistortionCorrectionAndEPIToT1wReg_FLIRTBBRAndFreeSurferBBRbased.sh<https://github.com/Washington-University/HCPpipelines/blob/0a27221d0c44e45db6bdc5a8f38f05c3f32adbdb/fMRIVolume/scripts/DistortionCorrectionAndEPIToT1wReg_FLIRTBBRAndFreeSurferBBRbased.sh#L328> called from the GenericFMRIVolume.sh<https://github.com/Washington-University/HCPpipelines/blob/0a27221d0c44e45db6bdc5a8f38f05c3f32adbdb/fMRIVolume/GenericfMRIVolumeProcessingPipeline.sh#L377>).
It seems like the warps are used instead of apply_topup to 1) do 1-step resampling, 2) handle modulation (Least Squares Restoration _or_ jacobian) and 3) bring along the sbref. I just wanted to check that I understood the reasoning and that I’m not I’m not missing something obvious. The reason this came up is because we acquired HCP-style se-based fieldmaps for another study and want to be sure that we’re applying them in a consistent way. Finally, we also noticed that the topup config file used in HCP is nearly-identical with the standard FSL-distributed topup config, except that the final set of iterations was increased from 20 to 50. Do you recommend this change for all data, or was it something specific that you saw in the HCP fieldmaps? Thanks for any advice, Erik _______________________________________________ HCP-Users mailing list HCP-Users@humanconnectome.org http://lists.humanconnectome.org/mailman/listinfo/hcp-users
