03-06-2019: Meeting w/ Camille (processing)

[cmaumet]

Table: https://www.frontiersin.org/files/Articles/204386/fninf-10-00024-HTML/image_m/fninf-10-00024-t001.jpg

[gkiar]

HCP Dataset(s):

• Motor
• Gambling
• (optional) UKBB Task

From Alex's pipelines, the task component is here. Once encoded for a task, will remain fixed.

For Alex's pipelines, the preprocessing is done here. This takes task info into consideration, and we want to sweep over implementation details here, such as preproc. steps. The encoding of these preprocessing steps is variable for each software, i.e. for FSL in an .fsf file.

Greg TODO:

• [ ] look at HCP paper motor analysis to see how tasks are encoded
• [ ] try to translate the above into Alex/Camille's representation
• [ ] Look at the table above for preprocessing steps and try to understand it :)

[gkiar]

For processing pipelines, use canonical choices per software as starting point

[gkiar]

unrelated to this issue but the paper in general: https://miccai2018.org/en/MICCAI-2018-PROCEEDINGS.html

[gkiar]

HCP task fMRI info: http://protocols.humanconnectome.org/HCP/3T/task-fMRI-protocol-details.html

[cmaumet]

HCP document including .fsf files (Appendix 3D, linked from the one above)

[gkiar]

https://www.sciencedirect.com/science/article/pii/S1053811913005272