Hi, I hope you are doing well. Sorry if this is posted in the wrong place, I couldn't find a forumn specific for GIFT or within the TReNDS center so thought I might be able to get help here. I'm working with a collegue to run an ICA on our fMRI data however whenever we load the fMRI data it tells us we can only run it with 2 IC components, despite the component estimate through GIFT telling us 78 components would be optimal. Interestingly we've done a similar analysis on another older dataset and were able to confirm we can still run the ICA on our old dataset, just an issue with our new dataset. So far we have tried the following solutions...

• Adjusting folder structure to not include ses-0 subfolder


• Compared headers from both fMRI datasets and everything is relatively the same expect for a small issue with our 'extents' info in the headers being 0 instead of 16384


• Adding in dataset_description.json and reading in as bids compatable, and as non-compatable

Do you have any suggestions for resolving or exploring this issue?