Search

Background and purpose: Conventional elective neck irradiation (ENI) in head and neck cancer consists of radiotherapy (RT) to the regional lymph node (LN) levels contoured on computed tomography. Hybrid Magnetic Resonance (MR) - RT modalities, such as combined magnetic resonance imaging - linear accelerators (MRLs), might enable new ENI strategies in which individual non-suspect lymph nodes (i-LNs) are targeted. In this treatment planning study, new MR-based strategies targeting i-LNs (i-ENI) were compared to conventional treatment. Materials and methods: All i-LNs were delineated on MR images of ten retrospectively selected patients with T2-4aN0M0 laryngeal cancer. Three strategies were considered. Strategy A: Conventional ENI delivered with a conventional linear accelerator (35x 1.55 Gy). Strategy B: MRL-based i-ENI (35x 1.55 Gy) to the individual lymph nodes including a background dose to the conventional elective neck volumes (35x 1.03 Gy). Strategy C: Same as Strategy B, but without background dose. In all plans the dose prescription to the primary tumor was 35x 2 Gy. Mean dose (Dmean) reductions in the organs at risk (OAR) were compared using the Wilcoxon signed rank test. Results: Compared to conventional ENI (strategy A), significant Dmean reductions of 6.0 Gy and 8.0 Gy were observed in the submandibular glands, of 9.4 Gy and 13 Gy in the carotid arteries and of 9.9 Gy and 19.4 Gy in the thyroid for strategy B and C, respectively. Large inter-patient variations of Dmean reductions were observed in all OARs. Conclusion: MRL-based i-ENI is a new promising concept that could reduce the mean dose to OARs in the neck significantly for patients with laryngeal cancer.

Background and purpose Adaptive radiotherapy based on cone-beam computed tomography (CBCT) requires high CT number accuracy to ensure accurate dose calculations. Recently, deep learning has been proposed for fast CBCT artefact corrections on single anatomical sites. This study investigated the feasibility of applying a single convolutional network to facilitate dose calculation based on CBCT for head-and-neck, lung and breast cancer patients.Materials and Methods Ninety-nine patients diagnosed with head-and-neck, lung or breast cancer undergoing radiotherapy with CBCT-based position verification were included in this study. The CBCTs were registered to planning CT according to clinical procedures. Three cycle-consistent generative adversarial networks (cycle-GANs) were trained in an unpaired manner on 15 patients per anatomical site generating synthetic-CTs (sCTs). Another network was trained with all the anatomical sites together. Performances of all four networks were compared and evaluated for image similarity against rescan CT (rCT). Clinical plans were recalculated on rCT and sCT and analysed through voxel-based dose differences and γ-analysis.Results A sCT was generated in 10 s. Image similarity was comparable between models trained on different anatomical sites and a single model for all sites. Mean dose differences 95% were achieved for all sites.Conclusion Cycle-GAN reduced CBCT artefacts and increased similarity to CT, enabling sCT-based dose calculations. A single network achieved CBCT-based dose calculation generating synthetic CT for head-and-neck, lung, and breast cancer patients with similar performance to a network specifically trained for each anatomical site.