Hi Guy

We don’t automatically deface, nor do we transform into a normalized coordinate system by default. We do “conform” – make the input data 1mm isotropic and 256^3, but it remains in the same RAS coordinate system

Cheers

Bruce

From: freesurfer-bounces@nmr.mgh.harvard.edu <freesurfer-bounces@nmr.mgh.harvard.edu> On Behalf Of Guy Fons F Rens
Sent: Monday, June 28, 2021 7:49 AM
To: freesurfer@nmr.mgh.harvard.edu
Subject: [Freesurfer] Question regarding tissue segmentation

        External Email - Use Caution        

        External Email - Use Caution        

Dear FreeSurfer community,

I was wondering whether someone could please help me.

Background

In my supervisor’s lab, we are currently doing neuroimaging with functional Near-Infrared Spectroscopy (fNIRS). Parts of our data analysis are performed in Homer3 & AtlasViewer, software designated for fNIRS analysis.

In order to perform analyses on individual (and not template) brains, AtlasViewer can read brain segmentation files generated by FreeSurfer. Accordingly, only anatomical and no functional files are needed.

Current status

• I am running Freesurfer  7.1.0-1 on my windows 10 computer through a Linux terminal (centos7).
• I am able to process single participants’ anatomicals through the recon-all command
• I can export the datasets from the Linux environment back to the Windows environment
• I can load the generated freesurfer files into AtlasViewer.

The issue

The recon-all command is highly convenient for generating the required files that can be read into AtlasViewer.

However, some of the included processing steps interfere with proper analysis in AtlasViewer. These steps are:

1. Automatic defacing of the anatomical: As we need anatomical landmarks (such as nasion) for fNIRS analysis, defacing makes defining this landmark tedious (if not impossible)
2. Transformation into normalized space: The dataset is automatically transformed into Talairach space. As we need to process our data in native space (because the fNRIS montage is in native space),  this step is interfering with our ability to align the anatomical scan and the fNIRS montage.

The specific questions

For our data analysis in AtlasViewer, we primarily need the anatomical with 5 tissue segmentation (skin scalp, CSF, WM and GM). Ideally, without (Issue 1) defacing and (Issue 2) transformation into normalized space.

As such my specific question is:

• How can I run FreeSurfer through the centos7 terminal to acquire the segmentation files without defacing and space-normalization.
• Or, if these files are actually generated through recon-all, how/where can I access them?

Last words

I would greatly appreciate any help with this issue. Although I have some coding experience, my knowledge of Linux is very limited. I tried to find this information on the FreeSurfer webpage but was unable to do so. Many thanks in advance and feel free to provide any support by reaching out to grens@uwo.ca [grens AT uwo dot ca].

Best wishes and, again, many thanks

Guy Rens