Intrapatient, rigid transformation
39 patients with lymph node metastases in pelvic or para-aortic region, each patient underwent 5 SBRT treatment sessions except 1 patient who underwent 2 cycles, so 10 SBRT treatment sessions in total. MR images of the pelvic or abdominal region were obtained at the start, during and after each treatment session.
For each patient, typically one of the sequences below was used for each scan at start, mid and end of each treatment session (so monomodal image registration).
T1 MRI
Acquired on a 1.5T MR-linac (Elekta Unity).
Transverse 3D T1-weighted FFE scan with an acquisition time of 2 min (TR 11 ms, TE 4.6 ms, acquired voxel size 1.5 x 1.5 x 2.0 mm3, FOV 400 x 400 x 300 mm3)
T2 MRI
Acquired on a 1.5T MR-linac (Elekta Unity).
Transverse 3D T2-weighted TSE scan with an acquisition time of 3min40s (TR 1500 ms, TE 124 ms, acquired voxel size 1.3 x 1.3 x 2.0 mm3, FOV 400 x 400 x 300 mm3)
For each treatment session, the mid-treatment (PV) and end (post) scans were registered to the start (pre) scan of that treatment session.
For parameter files see the Elastix Model Zoo repository on GitHub.
Elastix version: 4.8
Description:
The rigid registration that was used to register cropped PV or post scans to the cropped pre scan. A fixed mask was used that encompassed 1 cm around the pre-scan bony anatomy in the region close to the target lymph node (bone mask cropped at 2 cm towards cranial and caudal from the planning target volume).
Command line call:
elastix -p par0061.txt –f pre_scan.mhd -m post_scan.mhd -out output_dir -fMask bone_mask.mhd
Werensteijn-Honingh et al. (2020), Impact of a vacuum cushion on intrafraction motion during online adaptive MR-guided SBRT, submitted.
© 2020 Viktor van der Valk