SLICE TO VOLUME RECONSTRUCTION (SVR)

During MRI acquisition fetus moves. Fast acquisition of 2D slices freezes motion in time. Between-slice motion is the corrected with slice-to-volume registration and the 3D volume is reconstructed using super-resolution reconstruction. Watch this youtube video to learn more about fetal diffusion MRI.

IRTK SIMPLE is a software package has been created to distribute software for motion correction and super-resolution reconstruction of structural and diffusion MRI of the fetal brain. The reconstruction algorithms were developed and implemented by Dr Maria Deprez (Kuklisova Murgasova) within IRTK sofware package.

• installation: Install and compile on UBUNTU
• reconstruction of structural MRI
• distortion correction for EPI
• spherical harmonics reconstruction of diffusion MRI
• references: Read detailed description of reconstruction algorithms
• about irtk-simple: why the package was created
• about irtk
• disclaimer

References

The IRTK SIMPLE package contains three main algorithms:

Motion correction and super-resolution reconstruction of fetal structural MRI

Reconstruction of fetal brain MRI with intensity matching and complete outlier removal. M Kuklisova-Murgasova, G Quaghebeur, MA Rutherford, JV Hajnal, JA Schanbel. Medical Image Analysis 16(8), 2012. doi: 10.1016/j.media.2012.07.004

Distortion correction of fetal EPI

Distortion Correction in Fetal EPI Using Non-Rigid Registration With a Laplacian Constraint. M Kuklisova-Murgasova, G Lockwood Estrin, RG Nunes, SJ Malik, MA Rutherford, D Rueckert, JV Hajnal IEEE Transactions on Medical Imaging 37(1), 2018. doi: 10.1109/TMI.2017.2667227

Spherical Harmonics reconstruction of fetal diffusion MRI

Higher Order Spherical Harmonics Reconstruction of Fetal Diffusion MRI With Intensity Correction. M Deprez, A Price, D Christiaens, G Lockwood Estrin, L Cordero-Grande, J Hutter, A Daducci, JD Tournier, M Rutherford, SJ Counsell, M Bach Cuadra, JV Hajnal IEEE Transactions on Medical Imaging 39(4), 2020. doi: 10.1109/TMI.2019.2943565

About IRTK SIMPLE

The IRTK SIMPLE package is based on a lightweight and easy to compile subset of Image Registration Toolkit (IRTK).

The package is parallelized using Threading Building Blocks (TBB), which is included to ensure compatibility.

Installation

Installation instructions are available for UBUNTU systems. The dependences on the version of ubuntu have been minimised. It has not been tested on other systems.

• Install libraries:

  sudo apt-get install git cmake cmake-curses-gui g++ zlib1g-dev

• Download package from the Gitlab repository

  git clone https://github.com/mariadeprez/irtk-simple.git

  
  (you will need to enter your username and the personal token)

• Build

  cd irtk-simple
  mkdir build
  cd build
  cmake ..
  make

• run the executables to display input arguments and options

  irtk-simple/build/bin/reconstruction
  irtk-simple/build/bin/estimate-distortion
  irtk-simple/build/bin/reconstructionDWI
  

Structural MRI

The algorithm performs slice-to-volume registration and super-resolution reconstruction of structural MRI of fetal brain. It also includes intensity matching and robust statistics to reject misaligned and corrupted slices. Input data typically consists of several stacks of 2D slices of the fetal brain acquired in several different orientations. Typically, these are T2w HASTE, but can be other structural sequences (e.g. T2w or T1w bSSFP).

The example command:

reconstruction reconstructed_volume.nii.gz 3 axial_stack.nii.gz coronal_stack.nii.gz sagital_stack.nii.gz -mask axial_mask.nii.gz -packages 2 2 2 -thickness 2.5 2.5 2.5

It is essential to imput the brain (or head) mask. Packages refer to number of loops in acquisition order of the slices, e.g. if the slices are acquired as first odd then even, we have 2 packages. It is essential to input correct slice thickness for the super-resolution algorithm to work correctly.

Distortion correction

This command estimates distortion of echo-planar imaging (EPI) of the fetal brain, by registering the EPI stacks to the reconstructed stuctural volume, which is not affected by geometric distortion.

The example command:

estimate-distortion corrected_EPI_stacks.nii.gz T2aligned.nii.gz EPI_stacks.nii.gz -thickness 3.5 -mask mask.nii.gz  -phase y  -packages 2 -save_fieldmap fieldmap.nii.gz  

EPI stacks are in one nifti file which can be 3D or 4D if there are multiple stacks. The assumption is that the geometric distortion is stationary for these stacks. It is important to ensure that phase-encoding direction (in which the distortion occurs) entered correctly (can be x or y, z is always slice direction), otherwise the algorithm will not work correctly. The command will save corrected EPI stacks, and as an opion also estimated distortion field (fieldmap).

The distortion should be estimated from the B0 diffusion stacks (non-diffusion waited, e.g. b=0) or fMRI stacks. For diffusion weighted (DWI) stacks (b>0) the fieldmap estimated from the B0 stacks can be applied to correct them too. Example command:

fieldmapcorrect DWI_stacks.nii.gz corrected_DWI_stacks.nii.gz fieldmap.nii.gz -y

To ensure that fieldmap is correctly applied, input the coordinate on which the correction should be performed, e.g. -x or -y. If the direction in which the fieldmap needs to be applied is descending, use option -minus.

Diffusion MRI

This command reconstruct diffusion weighed MRI (DWI) of the fetal brain on the regular grid using Sherical Harmonics representation. Note that input signal DWI_signal.nii.gz should correspond to a single b-value (e.g. b=1000). Gradient table grad.b needs to be in MRtrix3 format.

An example command:

reconstructionDWI DWI_recon.nii.gz DWI_signal.nii.gz grad.b T2-template.nii.gz orient.dof -mask $mask -packages 2 -resolution 2

The fetal brain in DWI signal is usually in random orientation, as it was positioned in the scanner space. The command asks for a T2 weighted structural brain MRI image in standard orientation T2-template.nii.gz and a rigid transformation orient.dof that approximatelly reorients the diffusion data to the T2 template space. To check that your files are correct, download IRTK viewer rview here and check the alignment.

rview T2-template.nii.gz DWI_signal.nii.gz orient.dof

You can also create the transformation orient.dof manually in rview. It might be helpful to make sure that the origins of both images are set to zero to facilitate this, this can be achieved by IRTK command headertool also provided in this package.

The reconstructed signal DWI_recon.nii.gz can be viewed in rview. However, visualisation of of the fibre orientation distribution needs to be performed in MRtrix3:

dwi2response tournier DWI_recon.nii.gz response.txt -lmax 6 -grad grad.b -mask mask.nii.gz
dwi2fod csd DWI_recon.nii.gz response.txt csd.mif -lmax 6 -grad grad.b -mask mask.nii.gz
mrview T2.nii.gz -odf.load_sh  csd.mif

About IRTK

IRTK stands for Image Registration Toolkit

Authors: Daniel Rueckert and Julia Schnabel

The registration algorithms implemented in the software are described by the following publications:

J. A. Schnabel, D. Rueckert, M. Quist, J. M. Blackall, A. D. Castellano Smith, T. Hartkens, G. P. Penney, W. A. Hall, H. Liu, C. L. Truwit, F. A. Gerritsen, D. L. G. Hill, and D. J. Hawkes. A generic framework for non-rigid registration based on non-uniform multi-level free-form deformations. In Fourth Int. Conf. on Medical Image Computing and Computer-Assisted Intervention (MICCAI '01), pages 573-581, Utrecht, NL, October 2001

D. Rueckert, L. I. Sonoda, C. Hayes, D. L. G. Hill, M. O. Leach, and D. J. Hawkes. Non-rigid registration using free-form deformations: Application to breast MR images. IEEE Transactions on Medical Imaging, 18(8):712-721, 1999.

E. R. E. Denton, L. I. Sonoda, D. Rueckert, S. C. Rankin, C. Hayes, M. Leach, D. L. G. Hill, and D. J. Hawkes. Comparison and evaluation of rigid and non-rigid registration of breast MR images. Journal of Computer Assisted Tomography, 23:800-805, 1999.

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.