Dose evaluation of a deep learning-based synthetic CT generation in head and neck cancer for photon and proton therapy

Purpose: Pediatric craniopharyngioma, adult base-of-skull sarcoma and chordoma cases are all regarded as priority candidates for proton therapy. In this study, a dosimetric comparison between volumetric modulated arc therapy (VMAT) and intensity modulated proton therapy (IMPT) was first performed. We then investigated the impact of physical and biological uncertainties. We assessed whether IMPT plans remained dosimetrically superior when such uncertainty estimates were considered, especially with regards to sparing organs at risk (OARs).Methodology: We studied 10 cases: four chondrosarcoma, two chordoma and four pediatric craniopharyngioma. VMAT and IMPT plans were created according to modality-specific protocols. For IMPT, we considered (i) variable RBE modeling using the McNamara model for different values of (α/β)x, and (ii) robustness analysis with ±3 mm set-up and 3.5% range uncertainties.Results: When comparing the VMAT and IMPT plans, the dosimetric advantages of IMPT were clear: IMPT led to reduced integral dose and, typically, improved CTV coverage given our OAR constraints. When physical robustness analysis was performed for IMPT, some uncertainty scenarios worsened the CTV coverage but not usually beyond that achieved by VMAT. Certain scenarios caused OAR constraints to be exceeded, particularly for the brainstem and optical chiasm. However, variable RBE modeling predicted even more substantial hotspots, especially for low values of (α/β)x. Variable RBE modeling often prompted dose constraints to be exceeded for critical structures.Conclusion: For base-of-skull and pediatric craniopharyngioma cases, both physical and biological robustness analyses should be considered for IMPT: these analyses can substantially affect the sparing of OARs and comparisons against VMAT. All proton RBE modeling is subject to high levels of uncertainty, but the clinical community should remain cognizant possible RBE effects. Careful clinical and imaging follow-up, plus further research on end-of-range RBE mitigation strategies such as LET optimization, should be prioritized for these cohorts of proton patients.

BackgroundTo investigate the performance of a narrow-scope knowledge-based RapidPlan (RP) model for optimisation of intensity-modulated proton therapy (IMPT) and volumetric modulated arc therapy (VMAT) plans applied to patients with pleural mesothelioma. Second, estimate the potential benefit of IMPT versus VMAT for this class of patients.MethodsA cohort of 82 patients was retrospectively selected; 60 were used to "train" a dose-volume histogram predictive model; the remaining 22 provided independent validation. The performance of the RP models was benchmarked, comparing predicted versus achieved mean and near-to-maximum dose for all organs at risk (OARs) in the training set and by quantitative assessment of some dose-volume metrics in the comparison of the validation RP-based data versus the manually optimised training datasets. Treatment plans were designed for a prescription dose of 44 Gy in 22 fractions (proton doses account for a fixed relative biological effectiveness RBE = 1.1).ResultsTraining and validation RP-based plans resulted dosimetrically similar for both VMAT and IMPT groups, and the clinical planning aims were met for all structures. The IMPT plans outperformed the VMAT ones for all OARs for the contra-lateral and the mean and low dose regions for the ipsilateral OARs. Concerning the prediction performance of the RP models, the linear regression for the near-to-maximum dose resulted in Dachieved = 1.03Dpredicted + 0.58 and Dachieved = 1.02Dpredicted + 1.46 for VMAT and IMPT, respectively. For the mean dose it resulted: Dachieved = 0.99Dpredicted + 0.34 and Dachieved = 1.05Dpredicted + 0.27 respectively. In both cases, the linear correlation between prediction and achievement is granted with an angular coefficient deviating from unity for less than 5%. Concerning the dosimetric comparison between manual plans in the training cohort and RP-based plans in the validation cohort, no clinical differences were observed for the target volumes in both the VMAT and IMPT groups. Similar consistency was observed for the dose-volume metrics analysed for the OAR. This proves the possibility of achieving the same quality of plans with manual procedures (the training set) or with automated RP-based methods (the validation set).ConclusionTwo models were trained and validated for VMAT and IMPT plans for pleural mesothelioma. The RP model performance resulted satisfactory as measured by the agreement between predicted and achieved (after full optimisation) dose-volume metrics. The IMPT plans outperformed the VMAT plans for all the OARs (with different intensities for contra- or ipsilateral structures). RP-based planning enabled the automation of part of the optimisation and the harmonisation of the dose-volume results between training and validation. The IMPT data showed a systematic significant dosimetric advantage over VMAT. In general, using an RP-based approach can simplify the optimisation workflow in these complex treatment indications without impacting the quality of plans.

PurposeTo compare plan quality among photon volumetric modulated arc therapy (VMAT) and intensity‐modulated proton therapy (IMPT) with robustness using three different proton beam delivery systems with various spot size (σ) ranges: cyclotron‐generated proton beams (CPBs) (σ: 2.7–7.0 mm), linear accelerator proton beams (LPBs) (σ: 2.9–5.5 mm), and linear accelerator proton mini beams (LPMBs) (σ: 0.8–3.9 mm) for the treatment of head and neck (HN) cancer with bilateral neck irradiation.MethodsTen patients treated for oropharynx cancer with bilateral neck irradiation were planned using CPBs, LPBs, LPMBs, and VMAT. The homogeneity index (HI), mean body dose, and defined volumetric doses for selected critical organs‐at‐risk (OARs) were compared. Set‐up uncertainties of ±3 mm and ± 3.5% range uncertainties were included in robust evaluation using V95%Rx > 95% (Volume that covers 95% of the target volume at 95% of the prescription (Rx) dose) to high dose and low dose CTV volumes (CTV_70 Gy and CTV_56 Gy). VMAT and proton plans were compared in terms of OAR doses and mean body dose only. Homogeneity Indices were compared among IMPT plans in addition to OAR doses. The Wilcoxon signed‐rank test was used to evaluate statistical differences between evaluation metrics for VMAT plans and all proton plan types.ResultsOAR dose metrics were improved by 2% to 30% from CPB plans to LPB or LPMB plans. Compared to photon VMAT plans, all OAR doses except for mandible dose metrics were improved by 2% to 53% for all proton plans. The mean body dose was also improved by 7.5% from CPB to LPB and by 10.8% from CPB to LPMB. In addition, the mean body dose was also improved by 44% from VMAT to CPB, by 48% from VMAT to LPB, and by 50% from VMAT to LPMB plans. Compared to CPB plans, HI was significantly better (p < 0.05) for the LPB and LPMB plans. HI also improved considerably from VMAT to CPB, LPB, and LPMB. For both CTV_70 Gy and CTV_56 Gy, average robust evaluation across all worst‐case scenarios was slightly better for CPB plans, with an average of V95%Rx of the CTV_70 Gy of 97.6% ± 1.22%, followed by 97.2% ± 1.31% and 97.2% ± 1.35% for LPB and LPMB plans, respectively. Robustness for CTV_56 Gy showed comparable robustness across all proton plan types, with an average V95%Rx of 97.4% ± 0.87% for CPB, 97.4% ± 1.21%, and 97.5% ± 1.08% for CPB, LPB, and LPMB plans, respectively.ConclusionWith decreased spot size, the LPB and LPMB are excellent alternatives to VMAT and CPB therapy and can significantly reduce the dose to normal tissue.

ObjectiveThe purpose of this study is to compare the dosimetric and biological evaluation differences between photon and proton radiation therapy.MethodsThirty esophageal squamous cell carcinoma (ESCC) patients were generated for volumetric modulated arc therapy (VMAT) planning and intensity-modulated proton therapy (IMPT) planning to compare with intensity-modulated radiation therapy (IMRT) planning. According to dose–volume histogram (DVH), dose–volume parameters of the plan target volume (PTV) and homogeneity index (HI), conformity index (CI), and gradient index (GI) were used to analyze the differences between the various plans. For the organs at risk (OARS), dosimetric parameters were compared. Tumor control probability (TCP) and normal tissue complication probability (NTCP) was also used to evaluate the biological effectiveness of different plannings.ResultsCI, HI, and GI of IMPT planning were significantly superior in the three types of planning (p < 0.001, p < 0.001, and p < 0.001, respectively). Compared to IMRT and VMAT planning, IMPT planning improved the TCP (p<0.001, p<0.001, respectively). As for OARs, IMPT reduced the bilateral lung and heart accepted irradiation dose and volume. The dosimetric parameters, such as mean lung dose (MLD), mean heart dose (MHD), V5, V10, and V20, were significantly lower than IMRT or VMAT. IMPT afforded a lower maximum dose (Dmax) of the spinal cord than the other two-photon plans. What’s more, the radiation pneumonia of the left lung, which was caused by IMPT, was lower than IMRT and VMAT. IMPT achieved the pericarditis probability of heart is only 1.73% ± 0.24%. For spinal cord myelitis necrosis, there was no significant difference between the three different technologies.ConclusionProton radiotherapy is an effective technology to relieve esophageal cancer, which could improve the TCP and spare the heart, lungs, and spinal cord. Our study provides a prediction of radiotherapy outcomes and further guides the individual treatment.

Purpose: Fractionated stereotactic radiosurgery (FSRS) can be given with at least three modalities: Gamma Knife, with the noninvasive frameless extend system (GKE); linear accelerator-based volumetric modulated arc therapy (VMAT); and intensity-modulated proton therapy (IMPT). We extracted treatment plans for patients who had received FSRS with GKE for recurrent skull base or intracranial tumors, created corresponding plans for VMAT and IMPT, and compared the quality of the three sets of plans. Methods and materials: Plans were extracted for 9 patients with recurrent malignant skull-base tumors (n = 6) or large intracranial tumors (n = 3) who had received FSRS with GKE (median dose 24 Gy in 3 fractions) in 2013 through 2015. Plans for VMAT were generated with a TrueBeam STx LINAC machine using 6-MV photons, and plans for IMPT were generated with multi-field optimization. The optimized VMAT and IMPT plans were normalized to achieve the best possible target coverage while meeting the same dose-volume constraints on organs at risk (OARs) as the GKE plans. Plans were evaluated on the basis of target coverage, conformity index, homogeneity index, gradient index, and treatment efficiency. Results: The median target volume was 10.2 cm3 (range 1.9 - 33.8 cm3). The VMAT and IMPT plans met all OAR constraints, and target coverage and conformity were comparable among all plans. VMAT and IMPT plans showed significantly better target uniformity and treatment delivery efficiency (P P

Intensity-based dual model method for generation of synthetic CT images from standard T2-weighted MR images – Generalized technique for four different MR scanners

Single Arc VMAT of H&N Cancer

MRI-only radiotherapy treatment planning (RTP) relies on synthetic-CT (sCT) images for dose calculation. This study evaluates the clinical feasibility of using a commercial sCT solution in brain RTP, MRCAT Brain, focusing on dosimetric accuracy and patient setup verification. For dosimetric evaluation, 93 patients with brain cancer who were treated with volumetric modulated arc therapy (VMAT) using a CT/MRI fusion workflow were included. sCT images were generated using MRCAT Brain. The sCT images were rigidly co-registered to the CT images. The clinical plan produced on the CT was recalculated on the sCT. Dosimetric accuracy was assessed by comparing dose differences in dose volume histogram (DVH) statistics for the planning target volume (PTV) and organs at risk (OARs).For patient setup verification, 70 patients were included, and total of 572 cone beam CT (CBCT) registrations were performed with sCT and CT as reference images. The sCT matching accuracy was validated by comparing the translational and rotational differences between sCT-CBCT and CT-CBCT registrations. The PTV mean dose difference between CT and sCT were 0.3%, 0.4%, and 0.2% for D50%, D2%, and D98%, respectively. The OAR mean dose differences were less than 0.3% for all OARs. 4 of 93 patients (4.3%) showed gross dosimetric errors of greater than±2%. 3/4 were caused by sCT error. For positioning verification, all results were between±1 mm and±1°. This study demonstrates the clinical feasibility of the MRCAT solution for brain MRI-only RTP, with dosimetric differences being clinically acceptable, along with submillimetre and sub-degree accuracy in patient setup verification.

Deep Inspiration Breath-Hold (DIBH) for Locoregional Breast Irradiation using Volumetric Modulated Arc Therapy (VMAT) and Intensity Modulated Proton Therapy (IMPT): A Dosimetric Comparison to Cardiac Substructure

We aimed to assess the dosimetric impact of advanced delivery radiotherapy techniques using either intensity modulated x-ray beams (IMXT), volumetric modulated arc therapy (VMAT), or intensity modulated proton therapy (IMPT), for high-risk prostate cancer patients with sentinel nodes in the pararectal region. Twenty high-risk prostate cancer patients were included in a prospective trial evaluating sentinel nodes on pelvic SPECT acquisition. To be eligible for the dosimetric study, patients had to present with pararectal sentinel nodes usually not included in the clinical target volume encompassing the pelvic lymph nodes. Radiotherapy-plans including the prostate, the seminal vesicles, and the pelvic lymph nodes with the pararectal sentinel nodes were optimized for 6 eligible patients. IMXT and IMPT were delivered with 7 and 3 beams respectively and VMAT with 2 arcs. Results were assessed with Dose-Volume Histograms and predictive normal tissue complication probabilities (NTCPs) models between the three competing treatment modalities aiming to deliver a total dose of 50.4 Gy in 1.8 Gy daily fractions. Target coverage was optimized with IMPT when compared to IMXT and VMAT. Coverage of the sentinel node was slightly better with IMXT (D98% 5 57.3 ± 5.1 Gy) when compared with VMAT (D98% 5 56.2 ± 4.1 Gy). The irradiation of rectal, bladder, small bowel, and femoral heads volumes was significantly reduced with IMPT when compared to IMXT and VMAT. NTCPs rates for rectal and bladder ≥ grade-3 late toxicity were better with IMPT (0.4 ± 0.0% and 0.0 ± 0.0%) compared with IMXT (4.6 ± 3.3% and 1.4 ± 1.1%), and VMAT (4.5 ± 4.0% and 1.6 ± 1.6%), respectively. Acceptable dose-volume distributions and low rectal and urinary NTCPs were estimated to geometrically complex pelvic volumes such as the ones proposed in this study using IMXT, VMAT and IMPT. IMPT succeeded, however, to propose the best physical and biological treatment plans compared to both X-ray derived plans.

Feasibility of magnetic resonance imaging-only rectum radiotherapy with a commercial synthetic computed tomography generation solution

BackgroundThe incorporation of magnetic resonance (MR) imaging in radiotherapy (RT) workflows improves contouring precision, yet it introduces geometrical uncertainties when registered with computed tomography (CT) scans. Synthetic CT (sCT) images could minimize these uncertainties and streamline the RT workflow. This study aims to compare the contouring capabilities of sCT images with conventional CT-based/MR-assisted RT workflows, with an emphasis on managing artefacts caused by surgical fixation devices (SFDs).MethodsThe study comprised a commissioning cohort of 100 patients with cranial tumors treated using a conventional CT-based/MR-assisted RT workflow and a validation cohort of 30 patients with grade IV glioblastomas treated using an MR-only workflow. A CE-marked artificial-intelligence-based sCT product was utilized. The delineation accuracy comparison was performed using dice similarity coefficient (DSC) and average Hausdorff distance (AHD). Artefacts within the commissioning cohort were visually inspected, classified and an estimation of thickness was derived using Hausdorff distance (HD). For the validation cohort, boolean operators were used to extract artefact volumes adjacent to the target and contrasted to the planning treatment volume.ResultsThe combination of high DSC (0.94) and low AHD (0.04 mm) indicates equal target delineation capacity between sCT images and conventional CT scans. However, the results for organs at risk delineation were less consistent, likely because of voxel size differences between sCT images and CT scans and absence of standardized delineation routines. Artefacts observed in sCT images appeared as enhancements of cranial bone. When close to the target, they could affect its definition. Therefore, in the validation cohort the clinical target volume (CTV) was expanded towards the bone by 3.5 mm, as estimated by HD analysis. Subsequent analysis on cone-beam CT scans showed that the CTV adjustment was enough to provide acceptable target coverage.ConclusionThe tested sCT product performed on par with conventional CT in terms of contouring capability. Additionally, this study provides both the first comprehensive classification of metal artefacts on a sCT product and a novel method to assess the clinical impact of artefacts caused by SFDs on target delineation. This methodology encourages similar analysis for other sCT products.

Dosimetric Comparison of the Newest Version Tomotherapy, Volumetric-Modulated Arc Therapy, and Intensity-Modulated Proton Therapy Plans for Angiosarcoma of the Scalp

The aim of this study was to compare various dosimetric parameters of dynamic mlc intensity modulated radiotherapy (IMRT) plans with volumetric modulated arc therapy (VMAT) plans for sino-nasal cancers, which are rare and complex tumors to treat with radiotherapy. IMRT using five fields, coplanar in the sagittal plane and VMAT employing two coplanar arc plans were created for five patients. The plans were assessed by comparing Conformity Index and Sigma Index (dose homogeneity) in the Planning Target Volume (PTV) and through comparison of dose-volume characteristics to the following organs at risk (OARs): Spinal cord, brainstem, eye, ipsilateral and contralateral optic nerve and the volume of brain receiving 10% of the prescribed dose (V10%). The total monitor units required to deliver the plan were also compared. Conformity Index was found to be superior in VMAT plans for three patients and in IMRT plans for two patients. Dose homogeneity within the PTV was better with VMAT plans for all five cases. The mean difference in Sigma Index was 0.68%. There was no significant difference in dose between IMRT and VMAT plans for any of the OARs assessed in these patients. The monitor units were significantly reduced in the VMAT plan in comparison to the IMRT plan for four out of five patients, with mean reduction of 66%. It was found in this study that for the treatment of sino-nasal cancer, VMAT produced minimal, and statistically insignificant improvement in dose homogeneity within the PTV when compared with IMRT. VMAT plans were delivered using significantly fewer monitor units. We conclude in this study that VMAT does not offer significant improvement of treatment for sino-nasal cancer over the existing IMRT techniques, but the findings may change with a larger sample of patients in this rare condition.

Simple SummaryThe research highlights the application of synthetic CT images, which are created by deforming planning CT scans to match daily CBCT anatomy in interfractional proton therapy for pediatric patients with pelvic tumors. The objective of the study is to identify changes in the proton path length and determine the impact of anatomical changes on the treatment plan’s quality. The findings reveal that the water equivalent path length method can effectively estimate proton range deviations on synthetic-CT images. The daily synthetic CT images can also be utilized as a surrogate to calculate dose and predict dosimetric changes in the plan of the day. This approach eliminates the need for frequent rescanning, thereby making the adaptive therapy process more streamlined and less burdensome for young patients. The results have the potential to improve the precision of proton therapy, hence paving the way for more effective treatments.(1) Background: Synthetic CT images of the pelvis were generated from daily CBCT images to monitor changes in water equivalent path length (WEPL) and determine the dosimetric impact of anatomy changes along the proton beam’s path; (2) Methods: Ten pediatric patients with pelvic tumors treated using proton therapy with daily CBCT were included. The original planning CT was deformed to the same-day CBCT to generate synthetic CT images for WEPL comparison and dosimetric evaluation; (3) Results: WEPL changes of 20 proton fields at the distal edge of the CTV ranged from 0.1 to 12 mm with a median of 2.5 mm, and 75th percentile of 5.1 mm for (the original CT—rescanned CT) and ranged from 0.3 to 10.1 mm with a median of 2.45 mm and 75th percentile of 4.8 mm for (the original CT—synthetic CT). The dosimetric impact was due to proton range pullback or overshoot, which led to reduced coverage in CTV Dmin averaging 12.1% and 11.3% in the rescanned and synthetic CT verification plans, respectively; (4) Conclusions: The study demonstrated that synthetic CT generated by deforming the original planning CT to daily CBCT can be used to quantify proton range changes and predict adverse dosimetric scenarios without the need for excessive rescanned CT scans during large interfractional variations in adaptive proton therapy of pediatric pelvic tumors.