Par0059 - elastix

Registration Description

Intrapatient, B-spline transformation, mutual information, rigidity penalty

Subjects

CT and MR images in the pelvic region.

Image data

Acquired on a 3T Philips Ingenia.

In-phase image from a 3D T1-weighted multi-echo gradient echo sequence
Voxel size : 0.97x0.97x1 mm3
TE/TR = 2.1ms / 6.5ms
Voxel size: 1x1x1 mm3

Acquired on Brilliance Big Bore.

Pixel spacing : 0.3 to 0.7 mm
Slice spacing : 3 mm

Application

MR and CT scans were acquired for 27 male patient for radiotherapy purposes. Non-rigid registration was used to align and resample the MR and CT volumes to train a deep learning model for synthetic CT generation. MR was the fixed image and CT the moving image.

Registration Settings

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

Elastix version: 4.9

Description:

Aligns and resamples MR and CT images

Command line call:

elastix -p  par0059_rigid.txt -p par0059_bspline.txt –f MR_image.nii.gz -m CT_image.nii.gz -out output_dir -fMask MR_body_mask.nii.gz

Published in

Florkow et al. (2019), Deep learning-based MR-to-CT synthesis: the influence of varying gradient echo-based MR images as input channels, under submission