Deprecated: Please see https://github.com/FNNDSC/Fetal_Brain_MRI_Surface_Extraction_Pipeline

pl-surfaces-fetus

Table of Contents

• Abstract
• Description
• Publications
• Usage
  • Required Arguments
  • Optional Output Options
  • Output
    • Files
    • Visualization
• Example
• Build
• TODO
  • Comments
• License

Abstract

Extract surfaces (.obj) for the subplate zone from pre-segmented fetal brain MRI volume (.mnc) using CIVET's marching-cubes/sphere_mesh and ASP/surface_fit.

Description

• OHBM 2020 Poster
• Another One

Given an input of a segmented volume (*.mnc) for fetal brain MRI between 20-33 weeks of gestational age (GA), this pipeline will produce surface meshes for vertex-wise quantitative analysis of the subplate zone.

First, a marching-cubes[1] algorithm is used (sphere_mesh) to extract the gray-white boundary as a surface mesh (wm_81920.obj). Next, mesh deformation[2] by the ASP algorithm (surface_fit) using a radial distance map shrinks the surface inwards to find the intermediate zone (IZ)'s superficial boundary (iz_81920.obj), producing two surfaces over the subplate zone (SP) with one-to-one vertex correspondence. The schedule of parameters used to run surface_fit depend on age. In reality, these parameters should depend on brain size and gyrification index, which are both correlated with age in normal neurodevelopment of a healthy fetus. Finally, subplate thickness is calculated as the Euclidean distance between corresponding vertices using the "tlink"[3] method.

cortical_thickness -tlink iz_81920.obj wm_81920.obj subplate_thickness.txt

[1] Lepage C. (2017), "Human MR Evaluation of Cortical Thickness Using CIVET v2.1", OHBM, http://mcin.ca/wp-content/uploads/2017/08/HBM2017_civet.png

[2] Kim et. al, (2005). "Automated 3-D extraction and evaluation of the inner and outer cortical surfaces using a Laplacian map and partial volume effect classification". NeuroImage 27, pp. 210-221.

[3] Lerch J (2005), "In-vivo Analysis of Cortical Thickness using Magnetic Resonance Images", http://www.bic.mni.mcgill.ca/~jason/jpl-thesis-submitted.pdf

Publications

Lana Vasung, Caitlin K Rollins, Hyuk Jin Yun, Clemente Velasco-Annis, Jennings Zhang, Konrad Wagstyl, Alan Evans, Simon K Warfield, Henry A Feldman, P Ellen Grant, Ali Gholipour (2019). "Quantitative In vivo MRI Assessment of Structural Asymmetries and Sexual Dimorphism of Transient Fetal Compartments in the Human Brain." Cerebral Cortex. https://doi.org/10.1093/cercor/bhz200

Lana Vasung, Caitlin K Rollins, Clemente Velasco-Annis, Hyuk Jin Yun, Jennings Zhang, Simon K Warfield, Henry A Feldman, Ali Gholipour, P Ellen Grant (2019). "Spatiotemporal Differences in the Regional Cortical Plate and Subplate Volume Growth during Fetal Development." Cerebral Cortex. https://doi.org/10.1093/cercor/bhaa033

Lana Vasung, Chenying Zhao, Matthew Barkovich, Caitlin K Rollins, Jennings Zhang, Claude Lepage, Teddy Corcoran, Clemente Velasco-Annis, Hyuk Jin Yun, Kiho Im, Simon Keith Warfield, Alan Charles Evans, Hao Huang, Ali Gholipour, Patricia Ellen Grant, Association between Quantitative MR Markers of Cortical Evolving Organization and Gene Expression during Human Prenatal Brain Development, Cerebral Cortex, Volume 31, Issue 8, August 2021, Pages 3610–3621, https://doi.org/10.1093/cercor/bhab035

Usage

docker run -u $(id -u) --rm \
    -v <INPUTDIR>:/incoming -v <OUTPUTDIR>:/outgoing \
    surfaces_fetus --side <LEFT|RIGHT> --age <N> /incoming /outgoing

Required Arguments

<INPUTDIR> must contain a single file *.mnc of segmentations for a single brain hemisphere (either left or right) with the following labels:

Label	Region
3	gray-white matter boundary
4	superficial boundary of intermediate zone (IZ)

Also, an estimate of the subject's GA must be given. The subplate zone changes drastically between 30-32 GA, hence a different algorithm is used to fit the inner surface mesh.

[--age <N>]
Estimate of subject's gestational age (GA), in weeks.

[--side <left|right>]
Specify brain hemisphere

Optional Output Options

Generate additional output which can be helpful for debugging or visualization.

[--keep-intermediate]
Keep intermediate files (e.g. *mask.mnc, *chanfer.mnc)

[--qc]
Save surface_fit logs and produce vertex-wise quality check files,
such as fitting distance error, curvature, and triangle area.

Output

Files

Output File Name	Purpose
wm_81920.obj	subplate outer surface
iz_81920.obj	subplate inner surface
intermediates/wm_mask.mnc	subplate outer mask
intermediates/iz_mask.mnc	subplate inner mask
intermediates/iz_chamfer.mnc	distance map to inner surface
qc/wm_cubes.log	surface extraction log, preprocessing and surface_fit
qc/iz_fit.log	fitting surface_fit log
qc/wm_dist.txt	marching-cubes distance error
qc/wm_smth.txt	marching-cubes smoothness quality
qc/wm_area.txt	marching-cubes triangle areas quality
qc/iz_dist.txt	fitting distance error
qc/iz_smth.txt	fitting smoothness quality
qc/iz_area.txt	fitting triangle areas quality
sp_thickness_tlink.txt	-tlink thickness between surfaces
qc/sp_thickness_tnear.txt	-tnear thickness between surfaces
qc/tlink_minus_tnear.txt	vertex-wise difference between -tlink and -tnear thicknesses
diemask.txt	vertex mask for diencephalon
not_subplate_mask.txt	vertex mask over region where the subplate is discontinuous
intermediates/mid_81920.obj	midpoints between inner and outer surface
qc/distortion_angles.txt	distortion angles between 0-pi

Visualization

MNI Display should be used to view *.mnc files. It supports overlay of *.obj files.

Display -spectral labels.mnc wm_81920.obj iz_81920.obj

brain-view is used to view *.obj surface meshes in 3D, along with overlaying data or a mask *.txt.

brain-view wm_81920.obj subplate_thickness.txt

Warning: brain-view is legacy software, I have spent 30 hours trying to compile it on a modern GNU/Linux and have made no progress.

Example

TODO sample data doesn't exist yet, because our lab is stingy about data sharing :(

I found some public stuff here:

• 10.18112/openneuro.ds003105.v1.0.1 - CC0 license
  • https://arxiv.org/pdf/2009.06275.pdf
• http://neuroimaging.ch/feta

docker run -u $(id -u) --rm \
    -v $PWD/examples/brain:/incoming:ro $PWD/out:/outgoing:rw \
    fnndsc/pl-surfaces-fetus surfaces_fetus \
    --age 31.29 --side left  \
    --qc --keep-intermediate \
    /incoming /outgoing

Build

Most simply,

docker build -t pl-surfaces-fetus .

Or cross-compile a multi-platform manifest list, and push to Dockerhub

docker run --rm --privileged aptman/qus -s -- -p ppc64le
docker buildx create --name moc_builder --use
docker buildx build -t fnndsc/pl-surfaces-fetus --platform linux/amd64,linux/ppc64le .
# optional clean up
docker buildx rm
docker run --rm --privileged aptman/qus -- -r

TODO

• --verbose option to display helpful messages when running locally
• expansion to cortical plate
• --clobber option
• forward arguments to cortical_thickness
• reuse chamfer for both surface_fit and volume_object_evaluate

Comments

• parm.py mathematical model, we're workin' on it
• edgy.py and triangle_aspect.py not currently used, since they are confounding with depth_potential -area_simple

License

Civet core

• open source
• no commercial use