Search

Background: The advantages of proton therapy can be further enhanced with online magnetic resonance imaging (MRI) guidance. One of the challenges in the realization of MRI‐guided proton therapy (MRPT) is accurately calculating the radiation dose in the presence of magnetic fields. Purpose: This study aims to develop an efficient and accurate proton dose calculation algorithm adapted to the presence of magnetic fields. Methods: An analytical‐numerical radiation dose calculation algorithm, Proton and Ion Dose Engine (PRIDE), was developed. The algorithm combines the pencil beam algorithm (PBA) with a novel iterative voxel‐based ray‐tracing algorithm. The new ray‐tracing method uses fewer assumptions and ensures broader applicability for proton beam trajectory prediction in magnetic fields, and has been compared to Wolf's method and Schellhammer's method. The accuracy of PRIDE algorithm was validated on three phantoms and two practical plans (one single‐field water plan and one prostate tumor plan) in different magnetic field strengths up to 3.0 T. The validation was performed by comparing the results against the Monte Carlo (MC) simulations, using the global gamma index criteria of 2%/2 mm and 3%/3 mm with a 10% threshold. Results: PRIDE showed good agreement with MC in homogeneous and slab heterogeneous phantom, achieving gamma passing rates (%GPs) above 99% for 2%/2 mm criteria when magnetic field strength is not greater than 1.5 T. Although the agreement decreased for scenarios involving high proton energy (240 MeV) and strong magnetic field (3.0 T), the 2%/2 mm %GPs still remained above 98%. In lateral heterogeneous phantom, the accuracy of PRIDE decreased due to the PBA's limitation. For the two practical plans in different magnetic fields, %GPs exceeded 98% and 99% for 2%/2 mm and 3%/3 mm criteria, respectively. Conclusions: PRIDE can perform efficient and accurate proton dose calculation in magnetic fields up to 3.0 T, and is expected to work as a useful tool for proton dose calculation in MRPT. [ABSTRACT FROM AUTHOR]

Background and Purpose: Poor image quality of cone-beam computed tomography (CBCT) images can hinder proton dose calculation to assess the influence of anatomy changes. The aim of this study was to evaluate image quality and proton dose calculation accuracy of synthetic CTs generated from CBCT using unsupervised 3D deep-learning networks. Materials and methods: A total of 102 head-and-neck cancer patients were used to train (N=82) and test (N=20) i) a cycle-consistent generative adversarial network, ii) a contrastive unpaired translation, and iii) a fusion of the two (CycleCUT). For patients in the test set, a repeat CT was deformably registered to a same-day CBCT to create a ground-truth CT for comparison. The proton plan was re-calculated on the ground-truth CT and synthetic CTs. The image quality of the synthetic CTs was evaluated using peak signal-to-noise ratio, structural similarity index measure, mean error, and mean absolute error (MAE). Proton dose calculation accuracy was assessed through 3D gamma analysis and dose-volume-histogram parameters. Results: All synthetic CTs accurately preserved the CBCT anatomy (verified by visual inspection) while improving the image quality. The CycleCUT network had slightly improved image quality compared to the other networks (MAE in body: 53 Hounsfield units (HU) vs. 54/55 HU). All networks had similar proton dose calculation accuracy with gamma passing rate above 97%. Conclusions: All three evaluated networks generated synthetic CT images with dose distributions comparable to those of conventional fan-beam CT. The synthetic CT generation was fast, making all networks feasible for adaptive proton therapy.

Purpose: Non‐Contrast Enhanced CT (NCECT) is normally required for proton dose calculation while Contrast Enhanced CT (CECT) is often scanned for tumor and organ delineation. Possible tissue motion between these two CTs raises dosimetry uncertainties, especially for moving tumors in the thorax and abdomen. Here we report a deep‐learning approach to generate NCECT directly from CECT. This method could be useful to avoid the NCECT scan, reduce CT simulation time and imaging dose, and decrease the uncertainties caused by tissue motion between otherwise two different CT scans. Methods: A deep network was developed to convert CECT to NCECT. The network receives a 3D image from CECT images as input and generates a corresponding contrast‐removed NCECT image patch. Abdominal CECT and NCECT image pairs of 20 patients were deformably registered and 8000 image patch pairs extracted from the registered image pairs were utilized to train and test the model. CTs of clinical proton patients and their treatment plans were employed to evaluate the dosimetric impact of using the generated NCECT for proton dose calculation. Results: Our approach achieved a Cosine Similarity score of 0.988 and an MSE value of 0.002. A quantitative comparison of clinical proton dose plans computed on the CECT and the generated NCECT for five proton patients revealed significant dose differences at the distal of beam paths. V100% of PTV and GTV changed by 3.5% and 5.5%, respectively. The mean HU difference for all five patients between the generated and the scanned NCECTs was ∼4.72, whereas the difference between CECT and the scanned NCECT was ∼64.52, indicating a ∼93% reduction in mean HU difference. Conclusions: A deep learning approach was developed to generate NCECTs from CECTs. This approach could be useful for the proton dose calculation to reduce uncertainties caused by tissue motion between CECT and NCECT. [ABSTRACT FROM AUTHOR]

Purpose: Proton pencil beam energy spectrum is an essential parameter for calculations of dose and linear energy transfer. We extract the energy spectrum from measured integral depth dose (IDD). Methods: A measured IDD (measIDD) in water is decomposed into many IDDs of mono‐energetic pencil beams (monoIDDs) in water. A simultaneous iterative technique is used to do the decomposition that extracts the energy spectrum of protons from the measIDD. The monoIDDs are generated by our analytic random walk model‐based proton dose calculation algorithm. The linear independence of the monoIDDs is considered and high spatial resolution monoIDDs are used to improve their linear independence. To validate the extraction, first we use synthesized IDDs (synIDD) with Gaussian energy spectrum and compare the extracted energy spectrum with the Gaussian; second, for the energy spectrum extracted from measIDDs, the accuracy of the extraction is validated by comparing the calculated IDD from the energy spectrum with the measIDD. The measIDDs are derived from commissioning of a cyclotron proton pencil beam system with a Bragg peak ionization chamber. The nominal energy of the pencil beams is from 70 to 245 MeV. The monoIDDs are generated for energies from 0.05 to 275 MeV in steps of 0.05 MeV with a spatial resolution of 1 mm. Results: The difference of the extracted and original Gaussian energy spectrum peaked at 75 and 80 MeV was <1%. As the energy decreased, the difference increased but was reduced by using 0.1‐mm monoIDDs. The difference was not sensitive to the energy interval of monoIDDs when the interval increased from 0.05 to 1 MeV. For the energy spectrum extraction from measIDDs, there was a main peak near the nominal energy but the spectrum was not in Gaussian distribution. In three example cases (70, 160, and 245 MeV), the relative differences of the measIDDs and calculated IDDs were within 3.4%, 2.9%, and 4.7% of the Bragg peak value, respectively. Fitting the spectrum by Gaussian distribution, we had σ = 0.87, 1.51, and 0.86 MeV, respectively, for these three examples, and the relative differences of the resultant calculated IDDs from the measIDDs were within 4.7%, 7.4%, and 4.5%, respectively. Conclusions: Our algorithm accurately extracted the energy spectrum from the synIDDs and measIDDs. High‐resolution monoIDDs are necessary to extract low‐energy spectrum. Energy spectrum extraction from measIDD reveals important information for beam modeling. [ABSTRACT FROM AUTHOR]

Pencil beam (PB) dose calculation is fast but inaccurate due to the approximations when dealing with inhomogeneities. Monte Carlo (MC) dose calculation is the most accurate method but it is time consuming. The aim of this study was to develop a deep learning model that can boost the accuracy of PB dose calculation to the level of MC dose by converting PB dose to MC dose for different tumor sites. The proposed model uses the PB dose and computed tomography image as inputs to generate the MC dose. We used 290 patients (90 head and neck, 93 liver, 75 prostate and 32 lung) to train, validate, and test the model. For each tumor site, we performed four numerical experiments to explore various combinations of training datasets. Training the model on data from all tumor sites together and using the dose distribution of each individual beam as input yielded the best performance for all four tumor sites. The average gamma passing rate (1 mm/1%) between the converted and the MC dose was 92.8%, 92.7%, 89.7% and 99.6% for head and neck, liver, lung, and prostate test patients, respectively. The average dose conversion time for a single field was less than 4 s. The trained model can be adapted to new datasets through transfer learning. Our deep learning-based approach can quickly boost the accuracy of PB dose to that of MC dose. The developed model can be added to the clinical workflow of proton treatment planning to improve dose calculation accuracy.

Purpose: Pencil beam algorithms are commonly used for proton therapy dose calculations. Szymanowski and Oelfke ['Two-dimensional pencil beam scaling: An improved proton dose algorithm for heterogeneous media,' Phys. Med. Biol. 47, 3313-3330 (2002)] developed a two-dimensional (2D) scaling algorithm which accurately models the radial pencil beam width as a function of depth in heterogeneous slab geometries using a scaled expression for the radial kernel width in water as a function of depth and kinetic energy. However, an assumption made in the derivation of the technique limits its range of validity to cases where the input expression for the radial kernel width in water is derived from a local scattering power model. The goal of this work is to derive a generalized form of 2D pencil beam scaling that is independent of the scattering power model and appropriate for use with any expression for the radial kernel width in water as a function of depth. Methods: Using Fermi-Eyges transport theory, the authors derive an expression for the radial pencil beam width in heterogeneous slab geometries which is independent of the proton scattering power and related quantities. The authors then perform test calculations in homogeneous and heterogeneous slab phantoms using both the original 2D scaling model and the new model with expressions for the radial kernel width in water computed from both local and nonlocal scattering power models, as well as a nonlocal parameterization of Molière scattering theory. In addition to kernel width calculations, dose calculations are also performed for a narrow Gaussian proton beam. Results: Pencil beam width calculations indicate that both 2D scaling formalisms perform well when the radial kernel width in water is derived from a local scattering power model. Computing the radial kernel width from a nonlocal scattering model results in the local 2D scaling formula under-predicting the pencil beam width by as much as 1.4 mm (21%) at the depth of the Bragg peak for a 220 MeV proton beam in homogeneous water. This translates into a 32% dose discrepancy for a 5 mm Gaussian proton beam. Similar trends were observed for calculations made in heterogeneous slab phantoms where it was also noted that errors tend to increase with greater beam penetration. The generalized 2D scaling model performs well in all situations, with a maximum dose error of 0.3% at the Bragg peak in a heterogeneous phantom containing 3 cm of hard bone. Conclusions: The authors have derived a generalized form of 2D pencil beam scaling which is independent of the proton scattering power model and robust to the functional form of the radial kernel width in water used for the calculations. Sample calculations made with this model show excellent agreement with expected values in both homogeneous water and heterogeneous phantoms. [ABSTRACT FROM AUTHOR]