Dear Chaleece,

The issue is likely with the 'analysis mask' definition (found in the Setup.Options tab), which may be perhaps using the default mask in CONN (which is defined in MNI-space). Simply switch that option to 'implicit mask' (to have CONN use a subject-specific mask instead derived from the functional data) or use 'explicit mask' and select another functional analysis mask (appropriately coregistered to your functional data), and see if that fixes this issue.

Typically the analysis mask only affects voxel-level analyses, so ROI-to-ROI analyses are likely perfectly fine (to double-check, go to the Setup.ROIs tab, select your ROI and click on the option 'ROI tools -> display slice viewer with functional overlay' in order to check visually whether your ROI is properly aligned with the functional data)

Hope this helps

Alfonso

Originally posted by Chaleece Sandberg:

Hello! 

I have been browsing the forum regarding native space seed-voxel analysis and have not seen the issue I am running into.

I only have one subject loaded. I have rs data that are already preprocessed. I loaded the native space structural and functionals. I loaded the subject-specific ROIs (cluster-based ROIs from a localizer task) and the native space gray, white, and csf segmented images. I loaded the output from ART as covariates (motion, timeseries, outliers). Under options, I chose ROI-ROI and seed-voxel (and all the output files). Denoising ran fine. First level ran fine.

BUT, when I view the first-level results, the activation map is offset from the structural map. What is happening? It looks as though the seed-voxel analysis was forced into MNI space (which is not what I want). Is it just the seed-voxel analysis that is affected? If I extract the correlation matrix for the ROI-ROI analysis for this subject, will it be accurate (i.e., all in native space)? How do I keep the seed-voxel analysis in native space? See attached image.

Any help/insight is greatly apprecicated. 

Thanks!