Closed HusBunia closed 10 months ago
Hmm I haven't taken a close look at myelin maps in the DG myself before. I don't have a chance right now to try to reproduce the error myself right now, but I suspect perhaps the problem might be occurring because of the volumetric binary mask that is used to constrain the sampling.
E.g. this is the command used for sampling (from the sample_myelin_map_surf
rule) is from this line:
wb_command -volume-to-surface-mapping {input.vol} {input.mid} {output.metric} -ribbon-constrained {input.outer} {input.inner} -volume-roi {input.ribbon}
or as an example:
wb_command -volume-to-surface-mapping
work/sub-001/anat/sub-001_hemi-R_space-corobl_desc-preproc_T1wDividedByT2w.nii.gz
hippunfold/sub-001/surf/sub-001_hemi-R_space-corobl_den-0p5mm_label-dentate_midthickness.surf.gii
hippunfold/sub-001/surf/sub-001_hemi-R_space-T1w_den-0p5mm_label-dentate_myelin.shape.gii
-ribbon-constrained
hippunfold/sub-001/surf/sub-001_hemi-R_space-corobl_den-0p5mm_label-dentate_outer.surf.gii
hippunfold/sub-001/surf/sub-001_hemi-R_space-corobl_den-0p5mm_label-dentate_inner.surf.gii
-volume-roi work/sub-001/anat/sub-001_hemi-R_space-corobl_label-dentate_desc-ribbon_mask.nii.gz
You could try running the command without the -volume-roi
option? The workbench command help has some more details on the mapping approaches and options available.
Unfortunately there is no way to combine the Hipp and DG surfaces while preserving the topology!
Hmm I haven't taken a close look at myelin maps in the DG myself before. I don't have a chance right now to try to reproduce the error myself right now, but I suspect perhaps the problem might be occurring because of the volumetric binary mask that is used to constrain the sampling.
E.g. this is the command used for sampling (from the
sample_myelin_map_surf
rule) is from this line:wb_command -volume-to-surface-mapping {input.vol} {input.mid} {output.metric} -ribbon-constrained {input.outer} {input.inner} -volume-roi {input.ribbon}
or as an example:
wb_command -volume-to-surface-mapping work/sub-001/anat/sub-001_hemi-R_space-corobl_desc-preproc_T1wDividedByT2w.nii.gz hippunfold/sub-001/surf/sub-001_hemi-R_space-corobl_den-0p5mm_label-dentate_midthickness.surf.gii hippunfold/sub-001/surf/sub-001_hemi-R_space-T1w_den-0p5mm_label-dentate_myelin.shape.gii -ribbon-constrained hippunfold/sub-001/surf/sub-001_hemi-R_space-corobl_den-0p5mm_label-dentate_outer.surf.gii hippunfold/sub-001/surf/sub-001_hemi-R_space-corobl_den-0p5mm_label-dentate_inner.surf.gii -volume-roi work/sub-001/anat/sub-001_hemi-R_space-corobl_label-dentate_desc-ribbon_mask.nii.gz
You could try running the command without the
-volume-roi
option? The workbench command help has some more details on the mapping approaches and options available.Unfortunately there is no way to combine the Hipp and DG surfaces while preserving the topology!
Really, thank you for your advice! After removing the --volume-roi argument, all of the dentate vertices have a value. It seems that the --volume-roi constrains the mapping and reduces the partial volume effect. If I remove it, would it affect my results? The myelin map has the same resolution as the original T1/T2w images. If I map images with poor resolution, like dMRI or fMRI, will it impact the results? Thanks again for your help.
With the very small size and high curvature of the dentate gyrus, you are almost always going to have partial-voluming when your spatial resolution is poor, whether you include or exclude those vertices.. So yes it is likely to impact your results and is something to be mindful of when you are interpreting your data, and to state that as a limitation.
With the very small size and high curvature of the dentate gyrus, you are almost always going to have partial-voluming when your spatial resolution is poor, whether you include or exclude those vertices.. So yes it is likely to impact your results and is something to be mindful of when you are interpreting your data, and to state that as a limitation.
Thanks for your suggestion !
Hi dear Hippunfold expert: I tried to use the myelin map function in hippunfold. But myelin on DG have many zero vertices. I used HCP preprocessed T1w and T2w image, I check the image and ensure there is no 0 point for voxel-based T1 or T2 image. Can you gave me some advice about this problem? I also tried to solve the problem by combine hippocampus and DG as a whole and then mapping myelin on it. But CA and DG are two seprate structures in my result. Is there a way to combine them and also keep topology ? Really appreciate for any help!