fetalsvrtk/svrtk repository overview

⁠SVRTK - slice to volume reconstruction toolkit

SVR reconstruction package from King's College London based on MIRTK library for fetal MRI motion correction including:

• 3D brain
• 3D body
• 4D whole fetal heart, including magnitude and blood flow reconstructions
• 3D placenta
• 3D and 4D multi-channel T2*
• SH brain diffusion (HARDI)

The description of the SVRTK package is available at the gitbub repository: https://github.com/SVRTK/SVRTK⁠

The SVRTK docker repository was implemented by Dr Alena Uus (KCL) and Dr Tom Roberts.

In case of any questions please contact: alena.uus (at) kcl.ac.uk

⁠Instructions:

Pull the docker image:

docker pull fetalsvrtk/svrtk:latest

Run the docker mounted to the folder with .nii.gz files for processing:

docker run -it --rm --mount type=bind,source=location_on_your_machine,target=/home/data fetalsvrtk/svrtk /bin/bash

Run SVR reconstruction for fetal brain (structural 3D):

Please note that it requires a 3D brain mask created (e.g., in ITK-SNAP) for the selected template stack.

cd /home/data

mkdir proc_out
cd proc_out
mirtk reconstruct ../outputSVR.nii.gz  5 ../stack1.nii.gz ../stack2.nii.gz ../stack3.nii.gz ../stack4.nii.gz ../stack5.nii.gz -mask ../mask.nii.gz  -template ../stack3.nii.gz -thickness 2.5 2.5 2.5 2.5 2.5 -svr_only -resolution 0.75 -iterations 3 

Run DSVR reconstruction for fetal trunk/body (structural 3D):

Please note that it requires a 3D trunk mask created (e.g., in ITK-SNAP) for the selected template stack.

cd /home/data

mkdir proc_out
cd proc_out
mirtk reconstructFFD ../outputDSVR.nii.gz 6 ../stack1.nii.gz ../stack2.nii.gz ../stack3.nii.gz ../stack4.nii.gz ../stack5.nii.gz ../stack6.nii.gz -mask ../mask.nii.gz -template ../stack2.nii.gz -default_thickness 2.5 -structural -resolution 0.85

For more examples see: https://github.com/SVRTK/SVRTK⁠

In case of any question - please contact: alena.uus (at) kcl.ac.uk

⁠CNN-based fully automated reconstruction fetal brain and thorax reconstruction in atlas space(without required masking / template stack selection) - general_auto_amd tag:

The source code repository: https://github.com/SVRTK/auto-proc-svrtk⁠

docker pull fetalsvrtk/svrtk:general_auto_amd

docker run -it --rm --mount type=bind,source=location_on_your_machine,target=/home/data fetalsvrtk/svrtk:general_auto_amd /bin/bash

• automated SVR brain reconstruction in the atlas space (SS TSE, 0.5T, 1.5T and 3T, TE=80 ... 180 ms) (Uus et al., 2024; Kuklisova et al., 2012):

bash /home/auto-proc-svrtk/scripts/auto-brain-reconstruction.sh /home/data/[path_to_folder_with_nii_files]  /home/data/[path_to_output_folder] OPTIONAL: [motion correction mode: (0 or 1): 0 - minor, 1 - >180 degree rotations / default: 0];  [slice thickness / default: 3.0];  [output recon resolution / default: 0.8];  [number of packages / default: 1 ]

• automated DSVR body reconstruction in the atlas space (SS TSE, 1.5T and 3T, TE=80 ms) see (Uus et al., 2022):

bash /home/auto-proc-svrtk/scripts/auto-body-reconstruction.sh /home/data/[path_to_folder_with_nii_files]  /home/data/[path_to_output_folder] OPTIONAL: [motion correction mode: (0 or 1): 0 - minor, 1 - >180 degree rotations / default: 1];  [slice thickness / default: 3.0];  [output recon resolution / default: 0.8];  [number of packages / default: 1]

Input data requirements:

• more than 5-6 stacks
• full ROI coverage in all stacks
• 21-40 weeks GA
• no extreme shading artifacts
• singleton pregnancy
• 0.55 / 1.5 / 3T
• 80 – 180ms TE
• sufficient SNR and image quality

Output:

• 0.8mm resolution
• Standard radiological space

or use direct docker run option:

 docker run --rm  --mount type=bind,source=LOCATION_ON_YOUR_MACHINE,target=/home/data  fetalsvrtk/svrtk:general_auto_amd sh -c ' bash /home/auto-proc-svrtk/scripts/auto-body-reconstruction.sh /home/data/folder-with-files /home/data/out-thorax-recon-results 1 3.0 0.8 1; chmod 1777 -R /home/data/out-thorax-recon-results ; '

 docker run --rm  --mount type=bind,source=LOCATION_ON_YOUR_MACHINE,target=/home/data  fetalsvrtk/svrtk:general_auto_amd sh -c ' bash /home/auto-proc-svrtk/scripts/auto-brain-reconstruction.sh /home/data/folder-with-files /home/data/out-brain-recon-results 1 3.0 0.8 1; chmod 1777 -R /home/data/out-brain-recon-results ; ' 

0.55T reconstruction options:

#0.55T auto brain reconstruction
docker run --rm  --mount type=bind,source=LOCATION_ON_YOUR_MACHINE,target=/home/data  fetalsvrtk/svrtk:general_auto_amd sh -c ' bash /home/auto-proc-svrtk/scripts/auto-brain-055t-reconstruction.sh /home/data/folder-with-files /home/data/out-brain-recon-results 1 4.5 1.0 1 ; ' 

#0.55T auto body reconstruction
docker run --rm  --mount type=bind,source=LOCATION_ON_YOUR_MACHINE,target=/home/data  fetalsvrtk/svrtk:general_auto_amd sh -c ' bash /home/auto-proc-svrtk/scripts/auto-body-055t-reconstruction.sh /home/data/folder-with-files /home/data/out-body-recon-results 1 4.5 1.0 1 ; '

⁠Recommendations:

In order to make sure that reconstruction is fast enough - please select a sufficient number of CPUs (e.g., > 8) and amount of RAM (e.g., > 20 GB) in the docker settings.

For GUI-enabled version run:

• for Ubuntu:

docker run -it -rm -e DISPLAY=$DISPLAY --mount type=bind,source=location_on_your_machine,target=/home/data -v /tmp/.X11-unix:/tmp/.X11-unix fetalsvrtk/svrtk:latest /bin/bash

• for Windows:

docker run --env DISPLAY="host.docker.internal:0.0" -it -rm -v %inputFolder%:/home/data fetalsvrtk/svrtk:latest  /bin/bash

• for OSX:

docker run -it  --rm --mount type=bind,source=location_on_your_machine,target=/home/data --env=“DISPLAY” --ipc="host" --privileged  --volume="/tmp/.X11-unix:/tmp/.X11-unix:rw”  fetalsvrtk/svrtk:latest  /bin/bash

⁠License:

The SVRTK dockers are distributed under the terms of the GNU General Public License v3.0⁠. This program is free software: you can redistribute it and/or modify it under the terms of the GNU General Public License as published by the Free Software Foundation. This software is distributed in the hope that it will be useful, but WITHOUT ANY WARRANTY; without even the implied warranty of MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the GNU General Public License for more details.

⁠Disclaimer:

This software has been developed for research purposes only, and hence should not be used as a diagnostic tool. In no event shall the authors or distributors be liable to any direct, indirect, special, incidental, or consequential damages arising of the use of this software, its documentation, or any derivatives thereof, even if the authors have been advised of the possibility of such damage.

⁠Citation and acknowledgements:

Please give appropriate credit to the SVRTK package by citing Uus et al., 2020 (original work where SVRTK was implemented).

I.e., this is the official citation for SVRTK:

And use additional references for individual reconstruction functions:

3D automated fetal brain reconstruction in the atlas space:

Original reconstruction pipeline for 3D fetal brain (original software using IRTK: https://github.com/BioMedIA/IRTK⁠):

3D DSVR fetal body / placenta reconstruction:

3D automated fetal thorax reconstruction in the atlas space:

4D cardiac magnitude reconstruction (original software using IRTK: https://github.com/jfpva/fetal_cmr_4d⁠):

4D cardiac velocity reconstruction:

SH brain diffusion reconstruction (HARDI):

3D and 4D T2* placenta reconstruction: