To Whom it may concern,
I have created a .dlabel.nii file that is a parcellation covering both cortical
and subcortical areas. Then, I have extracted mean values for each of the
parcels from a contrast map using wb_command -cifti-parcellate.
After loading the resulting ptseries.nii files in matlab (using ft_readcifti),
I would like to perform some operations on the values stored in the ptseries
field, create a new .ptseries.nii file with the new values and visualize it in
workbench.
Unfortunately, I can’t succeed in saving a CIFTI file that is read by workbench
correctly.
Given that “cif” is my matlab structure, the output from ft_readcifti, with the
values in cif.ptseries changed with the results of my calculations, I have
attempted to use ft_write_cifti in several ways:
ft_write_cifti('test', cif, 'parameter', 'ptseries', 'brainstructure',
'brainordinate.brainstructure', 'parcellation', 'brainordinate.parcellation')
Returns no errors, but then if I try:
ft_read_cifti('test.ptseries.nii')
Undefined function or variable 'pos'.
Error in ft_read_cifti (line 601)
brainordinate.brainstructure = zeros(size(pos,1),1);
If instead I try:
ft_write_cifti('test', cif, 'parameter', 'ptseries')
I get a long list of regions:
parcel L_Amygdala contains 315 vertices in CIFTI_STRUCTURE_INVALID
parcel R_Amygdala contains 332 vertices in CIFTI_STRUCTURE_INVALID
parcel L_Hippocampus contains 764 vertices in CIFTI_STRUCTURE_INVALID
parcel R_Hippocampus contains 795 vertices in CIFTI_STRUCTURE_INVALID
parcel L_Accumbens contains 135 vertices in CIFTI_STRUCTURE_INVALID
parcel R_Accumbens contains 140 vertices in CIFTI_STRUCTURE_INVALID
parcel L_Caudate contains 728 vertices in CIFTI_STRUCTURE_INVALID
parcel R_Caudate contains 755 vertices in CIFTI_STRUCTURE_INVALID
parcel L_Pallidum contains 297 vertices in CIFTI_STRUCTURE_INVALID
parcel R_Pallidum contains 260 vertices in CIFTI_STRUCTURE_INVALID
parcel L_Putamen contains 1060 vertices in CIFTI_STRUCTURE_INVALID
parcel R_Putamen contains 1010 vertices in CIFTI_STRUCTURE_INVALID
parcel L_Thalamus contains 1288 vertices in CIFTI_STRUCTURE_INVALID
parcel R_Thalamus contains 1248 vertices in CIFTI_STRUCTURE_INVALID
parcel BrainStem contains 3472 vertices in CIFTI_STRUCTURE_INVALID
parcel L_VentralDiencephalon contains 706 vertices in CIFTI_STRUCTURE_INVALID
parcel R_VentralDiencephalon contains 712 vertices in CIFTI_STRUCTURE_INVALID
parcel L_Cerebellum contains 8709 vertices in CIFTI_STRUCTURE_INVALID
parcel R_Cerebellum contains 9144 vertices in CIFTI_STRUCTURE_INVALID
parcel R_V1_ROI contains 787 vertices in
CIFTI_STRUCTURE_CORTEX_RIGHT
parcel R_MST_ROI contains 76 vertices in
CIFTI_STRUCTURE_CORTEX_RIGHT
(…)
The file gets written, but if I try to open it I get:
ft_read_cifti('test.ptseries.nii')
Dot indexing is not supported for variables of this type.
Error in ft_read_cifti (line 439)
sel = strcmp({Surface(:).BrainStructure},
Parcel(j).BrainStructure{k});
Workbench also will not open the files generated this way, with two kinds of
errors: for the first file: “xml and nifti header disagree on matrix
dimensions” and for the second one: “Cifti XML error: you must set surfaces
before adding parcels that use them”.
I suspect I am either missing something basic in the syntax of ft_write_cifti
or my original .ptseries.nii files have something strange about them.
Thank you very much.
Yours faithfully,
Leonardo Tozzi, MD, PhD
Williams PanLab | Postdoctoral Fellow
Stanford University | 401 Quarry Rd
[email protected]<mailto:[email protected]> | (650) 5615738
_______________________________________________
HCP-Users mailing list
[email protected]
http://lists.humanconnectome.org/mailman/listinfo/hcp-users
