How is the functional connectomes represented by 116-by-116 matrices obtained？

[cnuzh]

Hello, how to obtain the functional connectomes represented by 116-by-116 matrices from the data in http://preprocessed-connectomes-project.org/abide/？
In other words, how to get data that can be used as GNN input？ The paper does not explain how the matrix of 116 times 116 is obtained.

[basiralab]

Hi there,

You can directly download the data from:
http://fcon_1000.projects.nitrc.org/fcpClassic/FcpTable.html

Thanks for pointing this out. We will add also this link to the paper.

[cnuzh]

Hi，what are the processing operations required for the data downloaded from this link to be directly used for training GNN?

[basiralab]

No further processing of the data. You just get the connectivity matrices directly by downloading them.

[basiralab]

However, we replaced all negative values in the connectivity matrices with zeros. This is a commonly adopted approach in functional brain connectivity analysis studies.