Par0043 - elastix

Registration Description

intra-subject; multi-resolution (4), rigid, Mutual Information metric (Mattes) with Adaptive Stochastic Gradient Descent optimizer

Image data

FOV was variable, but a typical value was about 500x500x300 mm^3 , in-plane matrix = 512x512, 3 mm slice thickness, 120 kV, exposure time = 923 ms, tube current = 121-183 mA.

it was obtained from 3D Cartesian dual Echo Dixon RF spoiled GRE, TE_1/TE_2 = 1.21/2.53 ms, TR =3.93 ms, Flip Angle = 10 deg , FOV = 477x477x300 mm^3, acquisition matrix = 248x281x120, reconstructed matrix = 480x480x120, BW = 1083 Hz and acquisition time of 2 min 13 s. The method used to generate pseudo-CT is a commercial solution called MR for Attenuation Correction (MRCAT, Philips Healthcare, Best, The Netherlands).

Application

The registration was used to align and resample CT to MR-based synbthetic-CT aiming at aligning the spatial coordinate of CT and MR-based synthetic-CT and also minimising positioning errors.

Registration Settings

For parameter files see the Elastix Model Zoo repository on GitHub.

elastix version: 4.700

Description:

Aligns the CT image to the pseudo-CT image, resampling it on the same image grid to enable voxelwise comparison.

Command line call:

elastix -f $gipl_fix -m $gipl_mov -out $dirReg -p $Para_File_Trans >/dev/null

where $gipl_fix is the fixed image (pseudo-CT), $gipl_mov is the moving image (CT), $dirReg is the output directory, and $Para_File_Trans is the file attached in this page.

Published in

Maspero M, et al. (2017), Quantification of confounding factors in MRI-based dose calculations as applied to prostate IMRT. Phys Med Biol. 62(3)

References

Maspero M, Seevinck PR, Schubert G, Hoesl MA, van Asselen B, Viergever MA, Lagendijk JJ, Meijer GJ, van den Berg CA. Quantification of confounding factors in MRI-based dose calculations as applied to prostate IMRT. Physics in Medicine & Biology. 2017 Jan 11;62(3):948. doi:https://doi.org/10.1088/1361-6560/aa4fe7;

Used also in

Maspero M, Savenije MH, Dinkla AM, Seevinck PR, Intven MP, Jurgenliemk-Schulz IM, Kerkmeijer LG, van den Berg CA. Dose evaluation of fast synthetic-CT generation using a generative adversarial network for general pelvis MR-only radiotherapy. Physics in Medicine & Biology. 2018 Sep 10;63(18):185001. doi:https://doi.org/10.1088/1361-6560/aada6d;

Maspero M, Tyyger MD, Tijssen RH, Seevinck PR, Intven MP, van den Berg CA. Feasibility of magnetic resonance imaging-only rectum radiotherapy with a commercial synthetic computed tomography generation solution. Physics and imaging in radiation oncology. 2018 Jul 1;7:58-64. doi:https://doi.org/10.1016/j.phro.2018.09.002.