The HCP data were processed using XCP-D

What is the format of the HCP dataset input tool when XCP-D processes HCP datasets?

Could you expand on what you mean? If you’re referring to the version of the HCP output structure, then XCP-D only currently works with versions >= February 2023. I’m not sure if HCP really versions its outputs (although, if it does, please let me know), so we just had the date we benchmarked against to go by.

Thank so much you reply to me.
I am using XCP_ D processes the minimum preprocessing data of HCP S1200,I have two questions :

  1. Xcp_d will perform filtering, smoothing, and other operations on cerebellar gray matter voxels?

  2. If default parameters are used, is the fmri cortical mapping space 32k or 91k?


XCP-D will automatically use CIFTI processing with HCP inputs, so if the CIFTIs include cerebellum, then XCP-D will retain it.

I believe XCP-D only supports 32k at the moment.

About the first problem,what you mean is that if the CIFTI data contain cerebellum voxel,XCP-D will also process thess voxels?
The second problem,I saw this description in the document“ If the --warp-surfaces-native2std is used, then fsnative surface files from the preprocessing derivatives will be warped to fsLR-32k space.”
If I don’t use this parameter, the cortical space is also 32k, right?


Correct. --warp-surfaces-native2std only applies to the structural surface files (i.e., pial, smoothwm), not the BOLD data. The BOLD data must be in fsLR-32k to begin with, which is the case with HCP data. XCP-D doesn’t apply any spatial transformations to the CIFTI BOLD data, so fsLR-32k CIFTIs provided by the preprocessing software (in this case HCP) will be denoised and outputted in the same space and resolution.