help
help > RE: Implicit vs explicit mask in setup
Apr 6, 2022 01:04 PM | Alfonso Nieto-Castanon - Boston University
RE: Implicit vs explicit mask in setup
Hi Andrea
Yes, sorry second-level analyses are not able to deal properly with missing-data that varies across voxels so your group-level analyses are only going to cover those voxels where you have data for all subjects (i.e. it is going to be restricted to the intersection of all participants masks). Typically implicit masking is used in the context of subject-specific ROI-level analyses, where one does not care about inter-subject coregistration at the level of individual voxels, while explicit masking is used in the context of voxel-level analyses (which require inter-subject coregistration at the level of individual voxels). Is the "chopped" data from those participants arising from a restricted field of view of their functional acquisition or is it arising from other sources (e.g. lesion masks, problems during normalization, etc.)?
Best
Alfonso
Originally posted by Andrea Burgess:
Yes, sorry second-level analyses are not able to deal properly with missing-data that varies across voxels so your group-level analyses are only going to cover those voxels where you have data for all subjects (i.e. it is going to be restricted to the intersection of all participants masks). Typically implicit masking is used in the context of subject-specific ROI-level analyses, where one does not care about inter-subject coregistration at the level of individual voxels, while explicit masking is used in the context of voxel-level analyses (which require inter-subject coregistration at the level of individual voxels). Is the "chopped" data from those participants arising from a restricted field of view of their functional acquisition or is it arising from other sources (e.g. lesion masks, problems during normalization, etc.)?
Best
Alfonso
Originally posted by Andrea Burgess:
Hi CONN experts,
I have several participants with functional data that have "chopped" images (please see attached). If I select "implicit mask" in the setup options, will the subsequent group-level analyses be restricted to the participant with the most severe chop? I tried rerunning the setup with the "explicit mask" option, but denoising QC-FC associations seem more off compared to the "implicit mask" option. Any advice is greatly appreciated!
Thanks in advance,
Andrea Burgess
I have several participants with functional data that have "chopped" images (please see attached). If I select "implicit mask" in the setup options, will the subsequent group-level analyses be restricted to the participant with the most severe chop? I tried rerunning the setup with the "explicit mask" option, but denoising QC-FC associations seem more off compared to the "implicit mask" option. Any advice is greatly appreciated!
Thanks in advance,
Andrea Burgess
Threaded View
| Title | Author | Date |
|---|---|---|
| Andrea Burgess | Apr 3, 2022 | |
| Alfonso Nieto-Castanon | Apr 6, 2022 | |
| Andrea Burgess | Apr 11, 2022 | |