Visualizations on brain plots

[Andrea-de-Varda]

Hello, great repository! I was wondering how one could plot on the brain surface the by-voxel pearson correlations obtained with all_scripts/generate_voxel_values.py. I wanted to reproduce the figures in Appendix K of your paper (also Figure 4), but the output does not seem to be used in all_scripts/figures_paper/ nor all_scripts/plot_{}.

Thank you!

[awwkl]

Thanks for the comment.

• To visualize the brain maps, one would need the subject-level anatomical data. However, we unfortunately cannot share this because we need to protect subject anonymity. This data is sensitive as it can be potentially used to identify an individual (since it contains the actual surface of the individual's brain).
• One option can be to convert the subject-level data to the common MNI template and then visualize that. This repository includes instructions on how to do this for fMRI data: https://github.com/anikethjr/brain_syntactic_representations

[nanjunye]

Thanks for the comment.

• To visualize the brain maps, one would need the subject-level anatomical data. However, we unfortunately cannot share this because we need to protect subject anonymity. This data is sensitive as it can be potentially used to identify an individual (since it contains the actual surface of the individual's brain).
• One option can be to convert the subject-level data to the common MNI template and then visualize that. This repository includes instructions on how to do this for fMRI data: https://github.com/anikethjr/brain_syntactic_representations

Hello, thx for ur reply!. I wonder that whether the affine transforms, mni transforms and masks offered by this repository u mentioned were fit for the voxel u chose in ur paper?

[nanjunye]

Thanks for the comment.

• To visualize the brain maps, one would need the subject-level anatomical data. However, we unfortunately cannot share this because we need to protect subject anonymity. This data is sensitive as it can be potentially used to identify an individual (since it contains the actual surface of the individual's brain).
• One option can be to convert the subject-level data to the common MNI template and then visualize that. This repository includes instructions on how to do this for fMRI data: https://github.com/anikethjr/brain_syntactic_representations

I find that the number of 'true' elements in masks for every subject matched the number of voxels u used to be predicted by the encoding model. For example, 27905 for subject F. However, I'm not sure about the correspondence between these 27905 voxels and the true elements in the masks. For example, is the first voxel corresponding to the first true element in the flattened one-dimensional mask, and so on?