Investigating the generalisation of an atlas-based synthetic-CT algorithm to another centre and MR scanner for prostate MR-only radiotherapy

Abstract

There is increasing interest in MR-only radiotherapy planning since it provides superb soft-tissue contrast without the registration uncertainties inherent in a CT-MR registration. However, MR images cannot readily provide the electron density information necessary for radiotherapy dose calculation. An algorithm which generates synthetic CTs for dose calculations from MR images of the prostate using an atlas of 3 T MR images has been previously reported by two of the authors. This paper aimed to evaluate this algorithm using MR data acquired at a different field strength and a different centre to the algorithm atlas. Twenty-one prostate patients received planning 1.5 T MR and CT scans with routine immobilisation devices on a flat-top couch set-up using external lasers. The MR receive coils were supported by a coil bridge. Synthetic CTs were generated from the planning MR images with ([Formula: see text]) and without (sCT) a one voxel body contour expansion included in the algorithm. This was to test whether this expansion was required for 1.5 T images. Both synthetic CTs were rigidly registered to the planning CT (pCT). A 6 MV volumetric modulated arc therapy plan was created on the pCT and recalculated on the sCT and [Formula: see text]. The synthetic CTs' dose distributions were compared to the dose distribution calculated on the pCT. The percentage dose difference at isocentre without the body contour expansion (sCT-pCT) was [Formula: see text] and with ([Formula: see text]-pCT) was [Formula: see text] (mean ± one standard deviation). The [Formula: see text] result was within one standard deviation of zero and agreed with the result reported previously using 3 T MR data. The sCT dose difference only agreed within two standard deviations. The mean ± one standard deviation gamma pass rate was [Formula: see text] for the sCT and [Formula: see text] for the [Formula: see text] (with [Formula: see text] global dose difference and [Formula: see text] distance to agreement gamma criteria). The one voxel body contour expansion improves the synthetic CT accuracy for MR images acquired at 1.5 T but requires the MR voxel size to be similar to the atlas MR voxel size. This study suggests that the atlas-based algorithm can be generalised to MR data acquired using a different field strength at a different centre.

A Systematic Review of the Clinical Implementation of Pelvic Magnetic Resonance Imaging–Only Planning for External Beam Radiation Therapy

“…Dosimetric accuracy of prostate synthetic CT solutions was investigated by the majority of studies and considered to be clinically acceptable in all cases. 6,16,[18][19][20][21] The similar results of these studies, despite significant differences in study design including various synthetic CT generation methods, shows that the dosimetric accuracy of synthetic CT techniques is broadly reproducible across a wide range of clinical systems and techniques, including multiple commercial options available for prostates. 10,11 The presented studies also provide a suitable blueprint for centers wishing to begin clinical implementation of MR-only planning themselves regarding dosimetric accuracy assessment.…”

“…In the case of Persson et al, 6 case, this was from a multicenter research-only study for commissioning purposes, whereas other experiences 14,17 resulted from treating their first MR-only patients. Wyatt et al 20 reported MR scan success rates from retrospectively assessed MR data, whereas Christiansen et al 16 reported synthetic CT generation success rate but did not discuss issues regarding MR acquisition. In describing their clinical workflow for MRonly planning, Tyagi et al 17 also reported a time savings when using an MR-only versus CT-MRIebased workflow.…”

“…Previous studies 6,16,[18][19][20][21] reported validating the dosimetric accuracy of their respectively chosen synthetic CT solution in a clinical environment, as required for commissioning MR-only planning.…”

Deep learning based synthetic‐CT generation in radiotherapy and PET: A review

“…Deep learning-based sCT generation in RT and PET August 18, 2021 page been envisioned to improve the quality of CBCT 29 . Cone-beam computed tomography plays a vital role in image-guided adaptive radiation therapy (IGART) for photon and proton therapy.…”

MRI-only treatment planning: benefits and challenges

“…This was followed by voxel based weighting of atlas HU values according to the acquired voxel MRI similarities to the registered atlas voxels(Dowling et al , 2015). The applicability of this method to 1.5T has recently been demonstrated(Wyatt et al , 2017). A similar approach termed ‘statistical decomposition algorithm’ uses atlas scans that are deformably registered to the acquired MRI with a first registration used to drive segmentation of tissue classes on the acquired MRI and then a second structure-guided registration using the segmentations.…”