Questions about the HCPTask and DynHCP-Task datasets

[Xeena2812]

Hello, I'm working on a project where we try to do graph diffusion to create synthetic ROI samples, to then condition a LDM model with them for synthetic fMRI generation. We've previously worked with the HCP 1200 dataset, using the fMRIs for training an LDM, and trying graph diffusion on ROIs we manually preprocessed. However, the preprocessing part was very difficult as I'm a Computer Science students and so is my supervisor, so we don't have a background in neuroimaging. That's when we came across your paper, and found it very promising because you've done the hard part for us and we can just use it to train our models. However, after looking at the NeuroGraphDataset a bit more carefully I've noticed that we've only been working with the "Motor" class part of the HCPTask dataset, and I have a few questions about the dataset:

1. Do you know if it's possible to map the samples in the HCPTask dataset's Motor class to know which subject they come from?
2. I noticed in the code that the NeuroGraphDynamic class's description that it has the DynHCPActivity dataset. Am I correct that this is the DynHCP-Task dataset from the paper, just with a different name?