Inquiry About Visualizing Functional Gradient Values with wb_view

[RuiWang0805]

I am currently working on functional connectivity gradient analysis and would like to use Connectome Workbench (wb_view) to visualize the results. Specifically, I have computed the functional gradients based on the DK308 atlas, and obtained a gradient vector of size 308×1, where each value corresponds to a brain region in the DK308 parcellation.

I would like to ask how I can visualize this gradient vector using wb_view, ideally by mapping the values onto a standard surface (e.g., fs_LR 32k). Could you kindly provide a step-by-step guide or workflow on how to achieve this? I am particularly interested in:

What file format should the gradient vector be saved in (e.g., .csv, .txt, .dscalar.nii)?

How to associate the gradient values with the DK308 parcels on the surface?

What commands or GUI steps should be followed in wb_view to visualize the gradient?

Any guidance or example commands/scripts would be greatly appreciated.

Thank you very much for your help

[coalsont]

Use -cifti-parcellate with some random 91k grayordinates dscalar file and a dlabel file of DK308 to obtain a .pscalar.nii file. You can then either load that into matlab/python to be able to use the parcels dimension from that file when saving your data to a new file, or you can save your data as a .csv or other simple delimited text format and use -cifti-convert with -from-text to make a new .pscalar.nii file that contains your data, which can be loaded in wb_view. Be sure to double-check the order of the parcels (-cifti-parcellate sorts the parcels by their label key value).

Gradients in functional connectivity generally occur on the edges between functional areas, so it is unusual to parcellate them (if the parcellation is a good representation of functional areas, the gradients will be right on the edge most of the time, and the peaks will fall into one area or the other based largely on how good or individualized the parcellation is).