A Standardized Pipeline for Examining Human Cerebellar Grey Matter Morphometry using Structural Magnetic Resonance Imaging

Summary

A standardized pipeline is presented for examining cerebellum grey matter morphometry. The pipeline combines high-resolution, state-of-the-art approaches for optimized and automated cerebellum parcellation and voxel-based registration of the cerebellum for volumetric quantification.

Abstract

Multiple lines of research provide compelling evidence for a role of the cerebellum in a wide array of cognitive and affective functions, going far beyond its historical association with motor control. Structural and functional neuroimaging studies have further refined understanding of the functional neuroanatomy of the cerebellum beyond its anatomical divisions, highlighting the need for the examination of individual cerebellar subunits in healthy variability and neurological diseases. This paper presents a standardized pipeline for examining cerebellum grey matter morphometry that combines high-resolution, state-of-the-art approaches for optimized and automated cerebellum parcellation (Automatic Cerebellum Anatomical Parcellation using U-Net Locally Constrained Optimization; ACAPULCO) and voxel-based registration of the cerebellum (Spatially Unbiased Infra-tentorial Template; SUIT) for volumetric quantification.

The pipeline has broad applicability to a range of neurological diseases and is fully automated, with manual intervention only required for quality control of the outputs. The pipeline is freely available, with substantial accompanying documentation, and can be run on Mac, Windows, and Linux operating systems. The pipeline is applied in a cohort of individuals with Friedreich ataxia (FRDA), and representative results, as well as recommendations on group-level inferential statistical analyses, are provided. This pipeline could facilitate reliability and reproducibility across the field, ultimately providing a powerful methodological approach for characterizing and tracking cerebellar structural changes in neurological diseases.

Introduction

The cerebellum is a part of the brain historically associated with motor control^1,2,3 and is thought to be integrally involved in only a small set of rare diseases, such as inherited ataxias⁴. However, converging lines of research from anatomical tracing studies in nonhuman primates, as well as human lesion and neuroimaging studies, provide compelling evidence for a role of the cerebellum in a wide array of cognitive^5,6,7, affective^8,9,10,11, and other nonmotor functions^7,12 (see⁶ for review). Furthermore, abnormalities of the cerebellum are increasingly implicated in a broad range of neurological and psychiatric disorders, including Parkinson's disease¹³, Alzheimer's disease^14,15, epilepsy^16,17, schizophrenia¹⁸, and autism spectrum disorder¹⁹. Therefore, it has become essential to incorporate the cerebellum into functional and structural models of human brain diseases and normative behavioral variability.

Anatomically, the cerebellum can be divided along its superior to inferior axis into three lobes: anterior, posterior, and flocculonodular. The lobes are further subdivided into 10 lobules denoted by roman numerals I-X^20,21 (Figure 1). The cerebellum can also be grouped into midline (vermis) and lateral (hemisphere) zones, which respectively receive inputs from the spinal cord and cerebral cortex. The anterior lobe, comprising lobules I-V, has traditionally been associated with motor processes and has reciprocal connections with cerebral motor cortices²². The posterior lobe, comprising lobules VI-IX, is primarily associated with nonmotor processes¹¹ and has reciprocal connections with the prefrontal cortex, posterior parietal, and superior temporal cerebral cortices^8,23. Lastly, the flocculonodular lobe, comprising lobule X, has reciprocal connections with vestibular nuclei that govern eye movements and body equilibrium during stance and gait²¹.

A growing body of recent work using functional neuroimaging has further refined understanding of the functional neuroanatomy of the cerebellum beyond its anatomical divisions. For example, resting-state functional magnetic resonance imaging (fMRI) techniques have been used to map the pattern of functional interactions between the cerebellum and cerebrum²⁴. Additionally, using a task-based parcellation approach, King and colleagues⁷ demonstrated that the cerebellum shows a rich and complex pattern of functional specialization across its breadth, evidenced by distinct functional boundaries associated with a variety of motor, affective, social, and cognitive tasks. Collectively, these studies highlight the importance of examining individual cerebellar subunits to develop complete biological characterizations of cerebellum involvement in both healthy variability and neurological diseases characterized by alterations in cerebellar structure and/or function.

The present work focuses on methods for quantifying local changes in cerebellar volume using structural MRI in humans. In general, there are two fundamental approaches to the quantification of regional brain volume using MRI data: feature-based segmentation and voxel-based registration. Feature-based segmentation approaches use anatomical landmarks and standardized atlases to automatically identify boundaries between subregions. Mainstream software packages for segmentation include FreeSurfer²⁵, BrainSuite²⁶, and FSL-FIRST²⁷. However, these packages provide only coarse parcellations of the cerebellum (e.g., labeling the whole grey matter and whole white matter in each hemisphere), thus overlooking the individual cerebellar lobules. These approaches are also prone to mis-segmentation, particularly overinclusion of the surrounding vasculature.

New machine-learning and multi-atlas labeling algorithms have been developed, which provide more accurate and finer-grained parcellation of the cerebellum, including Automatic Classification of Cerebellar Lobules Algorithm using Implicit Multi-boundary evolution (ACCLAIM^28,29), Cerebellar Analysis Toolkit (CATK³⁰), Multiple Automatically Generated Templates (MAGeT³¹), Rapid automatic segmentation of the human cerebellum and its lobules (RASCAL³²), graph-cut segmentation³³, and CEREbellum Segmentation (CERES³⁴). In a recent paper comparing state-of-the-art fully automated cerebellum parcellation approaches, CERES2 was found to outperform other approaches relative to gold-standard manual segmentation of the cerebellar lobules³⁵. More recently, Han and colleagues³⁶ developed a deep-learning algorithm called ACAPULCO (Automatic Cerebellum Anatomical Parcellation using U-Net with locally constrained optimization), which performs on par with CERES2, has broad applicability to both healthy and atrophied cerebellums, is available in open-source Docker and Singularity container format for 'off-the-shelf' implementation, and is more time-efficient than other approaches. ACAPULCO automatically parcellates the cerebellum into 28 anatomical regions.

In contrast to feature-based segmentation, voxel-based registration approaches operate by precisely mapping an MRI to a template image. To achieve this mapping, the voxels in the original image must be distorted in size and shape. The magnitude of this distortion effectively provides a measure of volume at each voxel relative to the gold-standard template. This form of volumetric assessment is known as 'voxel-based morphometry'³⁷. Whole-brain voxel-based registration approaches, such as FSL-FLIRT³⁸/FNIRT³⁹, SPM unified segmentation⁴⁰, and CAT12⁴¹, are commonly used for voxel-based morphometry. However, these approaches do not account well for the cerebellum, resulting in poor reliability and validity in infratentorial regions (cerebellum, brainstem⁴²). To account for these limitations, the SUIT (Spatially Unbiased Infra-tentorial Template) algorithm was developed to optimize cerebellum registration and improve the accuracy of voxel-based morphometry^42,43.

Feature-based segmentation and voxel-based registration approaches for the estimation of regional cerebellar volume have fundamental strengths and weaknesses. Segmentation approaches are substantially more accurate for quantifying the volume of anatomically defined areas (e.g., lobules³⁵). However, boundaries between distinct functional modules of the cerebellum do not map onto its anatomical folia and fissures (equivalent to gyri and sulci of the cerebrum⁷). As registration-based approaches are not constrained by anatomical landmarks, finer-grained spatial inference and high dimensional structure-function mapping of the cerebellum is possible⁴⁴. Taken together, segmentation and registration approaches are complementary to one another and can be used to answer different research questions.

Here, a new standardized pipeline is presented, which integrates these existing, validated approaches to provide optimized and automated parcellation (ACAPULCO) and voxel-based registration of the cerebellum (SUIT) for volumetric quantification (Figure 2). The pipeline builds upon the established approaches to include quality control protocols, using qualitative visualization and quantitative outlier detection, and a rapid method for obtaining an estimation of intracranial volume (ICV) using Freesurfer. The pipeline is fully automated, with manual intervention only required for checking the quality control outputs, and can be run on Mac, Windows, and Linux operating systems. The pipeline is freely available with no restrictions of its use for noncommercial purposes and can be accessed from the ENIGMA Consortium Imaging Protocols webpage (under "ENIGMA Cerebellum Volumetrics Pipeline"), following the completion of a brief registration form⁴⁵.

All required software is listed in the Table of Materials, and detailed tutorials, including a live demonstration, are available upon download of the pipeline, in addition to the protocol described below. Finally, representative results are provided, from the implementation of the pipeline in a cohort of people with Friedreich ataxia (FRDA) and age-and sex-matched healthy controls, alongside recommendations for group-level statistical inferential analyses.

Protocol

NOTE: The data used in this study were part of a project approved by the Monash University Human Research Ethics Committee (project 7810). Participants provided written informed consent. While the pipeline can be run on Mac, Windows, or Linux operating systems, ACAPULCO, SUIT, and the QC pipelines have explicitly been tested on Linux (Ubuntu) and Mac (Catalina, Big Sur v11.0.1) operating systems.

1. Module 1: ACAPULCO (anatomical parcellation)

1. Data collection
   1. Collect 3D T1-weighted MRI images of the whole brain at a resolution of 1 mm³ or less. Isotropic voxel dimensions (typically 1 mm x 1 mm x 1 mm), and a 3-Telsa (or greater) scanner are recommended. Consult with an imaging specialist at their radiography center to set up and acquire data that meet these specifications.
      NOTE: T2-weighted images are sometimes useful for volumetric analyses; however, the pipeline presented here relies on T1-weighted data only, and some of the tools used are exclusive to this type of data. As such, T2-weighted images cannot be used.
   2. Undertake a visual quality assessment of images to exclude gross cerebellar malformations (e.g., large lesions) or substantial motion artifacts that prevent identification of major cerebellar landmarks (e.g., the major anatomical fissures). Do not automatically exclude atrophied cerebella, even if substantial.
   3. For group studies, also consider quantitative quality assessments using freely-available, standardized tools such as MRIQC⁴⁶ to further identify problematic data.
   4. Convert all data to NIFTI-GZ format using a tool such as dcm2niix⁴⁷.
2. Recommended data organization
   1. Obtain all necessary software as listed in the Table of Materials. Ensure Docker⁴⁸ or Singularity⁴⁹, Matlab⁵⁰, and SPM12⁵¹ are installed prior to running the pipeline.
      NOTE: Extensive written and video tutorials describing the pipeline are also available (see the Table of Materials).
   2. Once all necessary software is installed, create folders in the working directory and label them 'acapulco,' 'suit,' and 'freesurfer.' Do this using the mkdir command from the command line.
   3. In the 'acapulco' directory, create an output folder. In the output folder, create a directory for each subject in the study containing the T1-weighted image in NIFTI-GZ format.
      NOTE: It is recommended to keep a copy of the original data elsewhere.
3. Anatomical cerebellar parcellation using ACAPULCO
   1. Go to the Table of Materials and download the relevant scripts and containers required to run ACAPULCO (under acapulco pipeline files). In the 'acapulco' directory, place the (i) ACAPULCO Docker OR Singularity container ('acapulco_0.2.1.tar.gz' or '.sif', respectively), (ii) contents of the QC_scripts archive (3 files: 'QC_Master.R,' 'QC_Plots.Rmd,' and 'QC_Image_Merge.Rmd'), and (iii) 'R.sif' (singularity) OR 'calculate_icv.tar' (docker) file.
   2. Open a terminal, and from the command line, run the ACAPULCO container on a single image (replace <<subject>> in the following). Wait for ~5 min for processing to complete.
      1. Using Docker, type the command:
         docker load --input acapulco_0.2.1.tar.gz
         docker run -v $PWD:$PWD -w $PWD -t --user $(id -u):$(id -g) --rm acapulco:latest -i output/<<subject>>/<<subject>>.nii.gz -o output/<<subject>>
      2. Using Singularity, type the command:
         singularity run --cleanenv -B $PWD:$PWD acapulco-0.2.1.sif -i output/<<subject>>/<<subject>>.nii.gz -o output/<<subject>>
   3. Loop across all subjects/scans in the cohort. See the Table of Materials for a link to the ENIGMA Imaging Protocols website for downloading the pipeline (under ENIGMA Cerebellum Volumetrics Pipeline) and the tutorial manual containing examples of how to create a for-loop for processing multiple subjects serially.
   4. After processing, look for the following files generated in the subject-specific folders:
      1. Identify "<subject>_n4_mni_seg_post_inverse.nii.gz": parcellated cerebellum mask in original (subject space).
      2. Identify "<subject>_n4_mni_seg_post_volumes.csv": volumes (in mm³) for each of the 28 subunits generated by acapulco;
      3. Identify representative images (in 'pics' directory): sagittal, axial, and coronal.
4. Statistical outlier detection and quality control (QC)
   1. From the terminal and in the 'acapulco' directory, ensure that the contents of QC_scripts are in the 'acapulco' directory. To run the QC scripts:
      1. Using Docker, type the command:
         docker load calculate_icv.tar
         docker run -v $PWD:$PWD -w $PWD --rm -it luhancheng/calculate_icv:latest Rscript
         QC_Master.R output/
      2. Using Singularity, type the command:
         singularity exec -B $PWD:$PWD R.sif Rscript /path/to/QC_Master.R /path/to/acapulco/output
5. Examining the QC images generated by ACAPULCO
   NOTE: There is a 3-step process for quality checking the ACAPULCO parcellated images.
   1. Open the 'QC_Images.html' in a web browser and quickly (~10 s per subject) scroll through the images to identify obvious failures or systematic issues. Note the subject IDs of failed or suspect parcellated images for follow-up.
      NOTE: See Figure 3 for a guide on the neuroanatomy of the cerebellar lobules and Figure 4, Figure 5, and Figure 6 in the representative results section below for examples of 'good' parcellations, 'subtle mis-parcellations,' and 'global failure' parcellations.
   2. Open the 'Plots_for_Outliers.html' to check the boxplots for quantitative statistical outliers. Look for outliers (2.698 s.d above or below the mean) above or below the whiskers of the box plots. Hover over the data points to display the Subject ID. Identify the outliers denoted by a '1' in the relevant column in the 'Outliers.csv' file, and note the total number of segments identified as outliers for each subject in the final column in 'Outliers.csv.'
   3. Manually inspect each image having one or more outliers. CRITICAL: Using a standard NIFTI image viewer (e.g., FSLEyes or MRICron), overlay the ACAPULCO mask onto the original T1w image to check the quality of the parcellation slice-by-slice.
      1. To generate overlays for detailed QC from the command line using FSLEyes, i) change the directory to the 'acapulco' directory, ii) specify the subject to view (replace <subject>):
         subj=<subject_name>
      2. Copy/paste the following code to the terminal (without manually changing {subj} as this has been set by the previous line:
         t1_image=output/${subj}/${subj}.nii.gz
         acapulco_image=output/${subj}/${subj}_n4_mni_seg_post_inverse.nii.gz
         fsleyes ${t1_image} ${acapulco_image} --overlayType label --lut random_big --outline --outlineWidth 3 ${acapulco_image} --overlayType volume --alpha 50 --cmap random
         NOTE: A determination will need to be made whether to include the abnormal segment or not, i.e., is there a parcellation error, or is it just normal variability in the individual's anatomy? Each parcellated region is considered individually, so a few regions can be excluded for an image, while the remainder can be retained if correct.
      3. Do one or more parcellated regions need to be excluded from the final dataset?
         If Yes (outlier is confirmed), exclude this parcellation(s) from the analysis by replacing the volume estimate with NA in the corresponding cell of the 'Cerebel_vols.csv' file for that subject.
      4. Do parcellation errors result in some of the cerebellum being excluded from the mask?
         If Yes, (for example, if particular cerebellar lobules are missing from the mask or appear 'cut off'), immediately exclude the subject from further analyses (i.e., do not proceed to run the SUIT module on those subjects).

2. Module 2: SUIT cerebellum-optimized voxel-based morphometry

1. Voxel-based morphometry analyses using SUIT
   CRITICAL: This pipeline requires the ACAPULCO module to have already been run, as it relies on the generation of a subject-specific cerebellar mask for optimization of the registration and normalization of the cerebellum to the SUIT template. If the subject-specific mask generated by ACAPULCO does not include the whole cerebellum, this warrants exclusion from the SUIT module. For instructions on running SUIT standalone, see⁵².
   1. Obtain all necessary software listed in the Table of Materials. Ensure the SPM12 folder and all subfolders are in the MATLAB path. Ensure enigma_suit scripts are saved in 'spm12/toolbox' directory and added to the MATLAB path. To check the MATLAB path, type pathtool in the MATLAB command window, then click Add with subfolders to add the relevant folders.
   2. Run the SUIT pipeline for one or more subjects. Wait for ~15-20 min (if using the graphical user interface [GUI]) and ~5-7 min if running from the terminal (bash/shell) for processing to complete.
      1. To use the GUI (subjects will be run in serial), from the MATLAB command window, type the command:
         suit_enigma_all
      2. In the first pop-up window, select the subject folders from the 'acapulco/output' directory to include in the analysis. Click on the individual folders on the right side of the window, or right-click and Select All. Press Done. In the second pop-up window, select the SUIT directory, where the analyses will be written.
      3. OR Call the function from the MATLAB command line for a single subject, type the command:
         suit_enigma_all('/path/to/acapulco/output/subjdir','/path/to/suitoutputdir')
      4. OR Call the function from the terminal window, outside of MATLAB, for a single subject by typing the command:
         matlab -nodisplay -nosplash -r "suit_enigma_all('/path/to/acapulco/output/subjdir','/path/to/suitoutputdir'), exit"
   3. See the Table of Materials for a link to the ENIGMA Imaging Protocols website for downloading the pipeline (under ENIGMA Cerebellum Volumetrics Pipeline) and the tutorial manual containing examples of how to create a for-loop for processing multiple subjects serially.
   4. Look for the following points regarding the script.
      1. Ensure that the script copies the N4 bias-corrected, MNI-aligned (rigid-body) T1 image and the ACAPULCO cerebellum mask into the output directory.
      2. Ensure that the script segments the grey and white matter of the cerebellum.
      3. Ensure that the script corrects for overinclusion errors in the parcellation using the ACAPULCO mask.
      4. Ensure that the script DARTEL normalizes and reslices the data into SUIT space with Jacobian modulation so that the value of each voxel is proportional to its original volume.
      5. Check each subject's folder for the following final outputs: 'wd<subject>_seg1.nii' (grey matter) and 'wd<subject>_seg2.nii' (white matter).
2. Statistical outlier detection and quality control
   1. Visually inspect the normalized, modulated images (wd*) for major failures. In MATLAB, type the command:
      spm_display_4D
   2. Manually select the 'wd*seg1' images from the suit subfolders, or navigate to the 'suit' directory; insert '^wd.*seg1' in the Filter box (no quotations) and press Rec button. Press Done.
   3. Scroll through the images to ensure they are all well-aligned. See Figure 7 for correctly normalized images from healthy controls (A,B) and an individual with a heavily atrophic cerebellum (D).
      NOTE: At this stage, the between-subject anatomy is very similar (as they have been registered to the same template), and volume differences are instead encoded by differing voxel intensities. Major failures will be obvious, e.g., blank images, large areas of missing tissue, unusual intensity gradients (i.e., bright voxels all at the top, dark voxels all at the bottom). These images should be excluded from subsequent steps.
   4. Check spatial covariance for outliers. In MATLAB, type the command:
      check_spatial_cov
      1. Select the 'wd*seg1' images as per the previous step. When prompted, select the following options: Prop scaling: Yes; Variable to covary out: No; Slice (mm): - 48 , Gap: 1.
      2. Look at the boxplot displaying the mean spatial covariance of each image relative to all others in the sample. Identify data points that are >2s.d. below the mean in the MATLAB command window. For these, inspect the "<subj>_n4_mni.nii.gz" image in the SUIT folder for artifacts (motion, anatomical abnormalities), image quality issues, or preprocessing errors.
      3. If the image quality and preprocessing are acceptable and visual inspection of the modulated images in the previous step does not indicate an issue with segmentation and normalization, retain these data in the sample. Otherwise, exclude these data.

3. MODULE 3 (optional): Intracranial Volume (ICV) estimation using FreeSurfer

NOTE: This module will use the FreeSurfer pipeline to calculate ICV. It does not need to be re-run if there are existing Freesurfer outputs for the cohort (any version).

1. Setting up FreeSurfer
   1. Ensure FreeSurfer is downloaded and installed⁵³. Go to the Table of Materials and download the relevant scripts to run this Module (under ICV pipeline files). When working with FreeSurfer, set the following variables:
      export FREESURFER_HOME=<freesurfer_installation_ directory>
      source $FREESURFER_HOME/SetUpFreeSurfer.sh
   2. Replace <path> in the following:
      export SUBJECTS_DIR=<path>/enigma/Freesurfer
2. Running Freesurfer autorecon1
   1. For a single subject, from inside the 'freesurfer' directory (processing time ~20 min), type the command:
      cd <path>/enigma/freesurfer
      recon-all -i ../input/<subject>.nii.gz -s <subject> -autorecon1
   2. See the tutorial manual for examples of how to create a for-loop for processing multiple subjects serially.
3. Calculation of ICV
   1. Data organization
      1. In the 'freesurfer' directory, place the (i) Docker OR Singularity container used in Module 1 ('calculate_icv.tar' or 'R.sif,' respectively) and (ii) xfm2det script (see the Table of Materials). Then, do a git clone to clone the required ICV script:
         git clone https://github.com/Characterisation-Virtual-Laboratory/calculate_icv
   2. Running ICV extraction (processing time ~5 min)
      1. From 'freesurfer' directory, with singularity ('R.sif') container, type:
         singularity exec --cleanenv -B $PWD:$PWD R.sif calculate_icv/calculate_icv.py --freesurfer_dir=/path/to/freesurfer --acapulco_dir=/path/to/acapulco/QC/Cerebelvolsfile --output_csv_name=Cerebel_vols.csv calculate_icv
      2. From 'freesurfer' directory, with docker container, type:
         docker run -v $PWD:$PWD -w $PWD -rm -it luhancheng/calculate_icv:latest
         calculate_icv/calculate_icv.py --freesurfer_dir=/path/to/Freesurfer --
         acapulco_dir=/path/to/acapulco/QC/Cerebelvolsfile --output_csv_name=Cerebel_vols.csv
         calculate_icv
      3. Running script without container-see the Table of Materials for additional required software and dependencies. From the 'freesurfer' directory, type:
         ./calculate_icv/ calculate_icv.py ---freesurfer_dir=/path/to/freesurfer --
         acapulco_dir=/path/to/acapulco/QC/Cerebelvolsfile --
         output_csv_name=Cerebel_vols.csv calculate_icv
         NOTE: This will calculate the ICV for each subject and append a column with ICV to the end of the 'Cerebel_vols.csv' file.

Results

Cerebellum parcellation (ACAPULCO)

Quality control of cerebellum parcellated masks:
The following examples demonstrate the ACAPULCO parcellated outputs and guide decision-making about a) the quality of the parcellated mask at the individual level and b) subsequent inclusion or exclusion of a particular lobule(s) from the statistical analyses. Ultimately, the decision to include or exclude a subject is subjective; examples of 'good parcellations,'...

Discussion

The cerebellum is critical to a wide range of human motor³, cognitive⁵⁸, affective¹⁰, and language^7,59 functions and is implicated in many neurological and psychiatric diseases. The availability of a standardized and easily implementable approach for the quantification of regional cerebellar volumes will contribute to increasingly detailed 'whole-brain' structure-function mapping...

Materials

Name	Company	Catalog Number	Comments
ACAPULCO pipeline files	0.2.1	http://enigma.ini.usc.edu/protocols/imaging-protocols/	Please make sure to use acapulco version 0.2.1
Docker for Mac		https://docs.docker.com/desktop/mac/install/	macOS must be version 10.14 or newer Docker requires sudo priviledges Docker imposes a memory (RAM) constraint on Mac OS. To increase the RAM, open Docker Desktop, go to Preferences and click on resources. Increase the Memory to the maximum
Docker for Windows		https://docs.docker.com/docker-for-windows/install/
ENIGMA SUIT scripts		http://enigma.ini.usc.edu/protocols/imaging-protocols/
FreeSurfer	7	https://surfer.nmr.mgh.harvard.edu/fswiki/DownloadAndInstall	Following variables need to be set everytime you work with Freesurfer: export FREESURFER_HOME=freesurfer _installation_directory source $FREESURFER_HOME/SetUpFreeSurfer.sh
			export SUBJECTS_DIR=path/enigma/Freesurfer
FSL (for FSLeyes). Optional	6	https://fsl.fmrib.ox.ac.uk/fsl/fslwiki/FslInstallation
ICV pipeline files		http://enigma.ini.usc.edu/protocols/imaging-protocols/	ICV pipeline can be run in two ways: 1) with docker/singularity. You will not require additionl software; 2) without docker/singularity- this involves running the ICV script (calculate_icv.py) manually. You will require the following additional software: Python version =3.5 Python module pandas Python module fire Python module tabulate Python module Colorama
		https://github.com/Characterisation-Virtual-Laboratory/calculate_icv
MATLAB*	2019 or newer	https://au.mathworks.com/	An academic license is required
Singularity	3.7 or newer	https://www.sylabs.io/docs/	Prefered for high performance computing (HPC) clusters
SPM	12	http://www.fil.ion.ucl.ac.uk/spm/software/spm12/	Make sure spm12 and all subfolders are in your MATLAB path
SUIT Toolbox	3.4	http://www.diedrichsenlab.org/imaging/suit_download.htm	Make sure you place SUIT toolbox in spm12/toolbox directory
Troubleshooting manual and segmentation output examples		http://enigma.ini.usc.edu/protocols/imaging-protocols/
Tutorial manual and video		http://enigma.ini.usc.edu/protocols/imaging-protocols/	Manual and accompanying live demonstration provide detailed step-by-step instructions on how to run the pipeline from start to finish.
*Not freely available; an academic license is required