NOrmative Morphometry Image Statistics

Disclaimer Copyright 2021 Olivier Potvin, Louis Dieumegarde, Simon Duchesne MEDICS LABORATORY - CERVO BRAIN RESEARCH CENTRE - UNIVERSITE LAVAL

NOMIS is a free tool to compute normative morphometric values for FreeSurfer 6 developed by the MEDICS laboratory at the CERVO Brain Research Centre / Universite Laval, Quebec, Canada.

NOMIS is licensed under a modified BSD. Permission to use, copy, modify, and distribute this software and its documentation for any purpose and without fee is hereby granted, provided that the above copyright notice appear in all copies. The author make no representations about the suitability of this software for any purpose. It is provided "as is" without express or implied warranty.

Instructions NOMIS computes normative Z-score effect size with a mean of 0 and a standard deviation of 1. Depending on the user need, there are four versions of Z-score adjusted on different sets of variables. All versions includes head size, image quality and scanner characteristics. The full version also includes age and sex while the other 3 versions are variants: with age, with sex, or without age and sex.

• The script automatically choose the Z-score version according to the information available in the csv input file.

If you use the normative values, please cite and refer to the following publication: Potvin O., Dieumegarde L., and Duchesne S. (2021) NOMIS: Quantifying morphometry deviation from the normality over the lifetime of the adult human brain. bioRxiv 2021.01.25.428063 https://doi.org/10.1101/2021.01.25.428063

These normative values are aimed to be applied on FreeSurfer 6.0 output already processed with fully-automated directive parameters: “recon-all -all” command without any manual or expert flag option (https://surfer.nmr.mgh.harvard.edu/fswiki/recon-all). All processed images should have finished without error and the user should verify the segmentation quality first.

This tool was built on MacOS using Python 3.7.4 with the following modules: argparse version 1.1 nibabel version 2.5.1 numpy version 1.17.2 pandas version 0.25.1 pickle version 4.0 sklearn version 0.21.3

For questions, email the author at: olivier.potvin at cervo.ulaval.ca

Script Extract the Calculator.zip file Run the python script (NOMIS.py) from a terminal

usage: python NOMIS.py [-h] -csv -s <subject_dir> [-o ] [-a ] [-v ] [-w ]

required arguments: -csv csv file containing "id" as in FreeSurfer’s subject directory, "sex" categorized as M/F, "age", "manufacturer" categorized as GE/Philips/Siemens, "mfs" (magnetic field strength) categorized as 1.5/3, and "voxel" (voxel acquisition size in mm3, for example: 1) For a csv file example, see "Calculator/Example/csv_example.csv"

-s The path to the directory containing the FreeSurfer subjects folders that you want to analyze.

optional arguments: -h, --help Show this help message -o Output directory where the normative z scores will be saved. Default is Calculator/Example/out/ -a Atlas name. Choices are DK, DKT, Destrieux, Default is DK. By default, norms for other atlases (aseg, ex vivo, brainstem subfields, hippocampal subfields) are computed. -v {on/off} Verbose. Default is on -w {on/off} Print code warnings. Default is off

Example A working example is supply with this tool and should produce normative data in the “./Calculator/Example/out” folder. In a Terminal window, change your working directory where the NOMIS.py tool is (the folder "Calculator/") and try: python NOMIS.py -csv Example/csv_example_ageandsex.csv -s Example/FreeSurfer_dir -a DK

Ouput Three folders are created. The "normative_z_scores" folder contains the nomative z scores files. The "raw_scores" folder contains the collected FreeSurfer 6 stats. The CNR folder contains the computed CNR values used to compute the norms.

Atlas files and region names are identical to those within the stats files, except: Larger regions such as brainsegvol appearing at the head of the aseg file are in files labeled as 'aseg_bigregions' cc = sum of the corpus callosum divisions: cc_posterior + cc_mid_posterior + cc_central + cc_mid_anterior + 'cc_anterior ventricles = sum of all ventricles: 3rd-ventricle + 4th-ventricle + left-inf-lat-vent + left-lateral-ventricle' + right-inf-lat-vent + right-lateral-ventricle

Other Although it is not required, if you want to compute pial surfaces stats for DKT and a2009s atlases, please see parcstats step in https://surfer.nmr.mgh.harvard.edu/fswiki/ReconAllTableStableV6.0. For the calculator we called them aparc.pial.DKT and aparc.pial.a2009s, but these names are arbitrary and can be changed easily within the NOMIS.py file (see atlaslist).

##########################################################################################