Dear Farzan,

Regarding individual-subject analyses, there are two approaches that are very much equivalent: a) running the analyses in subject-space and transforming, if necessary, any MNI-space templates or ROIs into each individual subject-space; or b) running the analyses in MNI-space and transforming, if necessary, any results of interest back to subject-space. If you already have a correct pipeline set up producing results in MNI-space, I would probably recommend simply transforming those results back to subject-space to use them for TMS neuronavigation (to do that, see for example this thread). That said, if you want to change your preprocessing and analysis pipeline to keep your data in subject-space, then see for example this thread on one way to do that. 

Regarding the unthresholded Fisher-transformed connectivity maps, these are stored in your conn project first-level analysis subfolder (e.g. connprojectfolder/results/firstlevel/SBC_01/). The files are named BETA_Subject#_Condition#_Source#.nii (with one file for each subject, condition, and seed/source). As always, both images should be in the same space to be properly overlayed. For example, if your functional data was in MNI-space and your structural image was in subject-space, then use the approach in the first link above to bring the MNI-space map back to subject-space (note: if your subject-space data was not coregistered -i.e. the structural image "subject-space" was different than the functional data "subject-space"- then simply use the structural normalization and centering transformation files instead of the files mentioned in that link when transforming back the MNI-space data to subject-space). If, on the other hand, both images are already in the same subject-space (e.g. if you used the preprocessing pipeline in the second link above, which coregisters the functional to the structural images), then you can directly display them overlayed without any additional transformation needed. 

Hope this helps

Alfonso

Originally posted by Farzan Vahedifard:

Dear CONN Team,

We are using the CONN toolbox to perform seed-based functional connectivity analyses from the subgenual ACC (sgACC) in native anatomical space, using externally preprocessed 7T fMRI data aligned to each subject’s FreeSurfer conformed space (verified in Freeview). Our goal is to export Fisher Z-transformed connectivity maps in native space for use in TMS neuronavigation.

We initially ran the analysis in MNI space, which produced good whole-brain connectivity maps.

 

However, for individualized targeting, we now need all processing and outputs to remain in native anatomical space.

Settings We Tried for Native-Space Output:

• 
  

  Disabled normalization to MNI in the preprocessing pipeline

  
  


• 
  

  Set bounding box to: [-180 -180 -180 ; 180 180 180] to avoid template constraints

  
  


• 
  

  Changed registration type from 'MNI' to 'rigid'

  
  

Despite these steps, the output maps still show misalignment between functional and anatomical images, suggesting internal transformations may still be applied.

 

We would greatly appreciate your help with:

1. 
   

   Which preprocessing steps must be modified or disabled to ensure full processing and outputs remain in native anatomical space?

   
   
   
2. 
   

   How to correctly configure CONN for use with externally preprocessed data, already aligned in native space?

   
   
   
3. 
   

   How can we visualize the raw (unthresholded) functional connectivity map from sgACC, overlaid on the subject’s anatomical image (e.g., FreeSurfer conformed T1)?

   
   
   
4. 
   

   Is there a way to view the full Fisher Z map without thresholding or restricting to predefined ROIs (e.g., to inspect whole-brain patterns)?

   
   
   

Thank you very much for your support—native-space processing is essential for our individualized TMS targeting workflow.

Best regards,
Farzan

---