NOMIS

NOTE: NOMIS requires Freesurfer 6.0 outputs. For more detail see Freesurfer.

To derive norm scores from Freesurfer 5.3, see the document below.

523KB

Freesurfer Norm Scores.docx

Open

Downloading NOMIS

GitHub - medicslaboratory/NOMIS: NOrmative Morphometry Image Statistics (NOMIS) for FreeSurfer 6.0GitHub

Download NOMIS from the github link above

Setting up a NOMIS conda environment

create a conda environment:

conda create -n "nomis" python==3.7.4

install dependencies

conda activate nomis
pip install argparse==1.1
pip install nibabel==2.5.1
pip install numpy==1.17.2
pip install pandas==0.25.1
pip install pickle4
pip install scikit-learn==0.21.3

Running NOMIS

For an automated script to run NOMIS on your BIDS formatted and recon-all'd data, see: https://github.com/ryn-thorn/Image_Analysis_Scripts/tree/main/NOMIS

https://github.com/ryn-thorn/Image_Analysis_Scripts/tree/main/NOMISgithub.com

Create a setup file. This is a .csv file with five columns: id, age, sex, manufacturer, and mfs.

• id: subject ID as it appears on in Freesurfer outputs

• age: can be a whole number or a decimal

• sex: M or F, uncertain of how this should be marked for people who identify as a gender they were not assigned at birth

• manufacturer: manufacturer of the MRI machine (for our purposes, usually Seimens)

• mfs: manufacturer field strength (for our purposes, usually 3)

See example below:

90B

NOMIS_setup.csv

Open

Run NOMIS using the following command structure:

conda activate nomis
python [path to nomis folder]/nomis.py -csv [path to setup file] -s [path to FS outputs] -o [path to nomis outputs]

NOMIS Outputs