Fred: a GPU-accelerated fast-Monte Carlo code for rapid treatment plan recalculation in ion beam therapy

IOeRT conventional and FLASH treatment planning system implementation exploiting fast GPU Monte Carlo: The case of breast cancer

Partial breast irradiation for the treatment of early-stage breast cancer patients can be performed by means of Intra Operative electron Radiation Therapy (IOeRT). One of the main limitations of this technique is the absence of a treatment planning system (TPS) that could greatly help in ensuring a proper coverage of the target volume during irradiation. An IOeRT TPS has been developed using a fast Monte Carlo (MC) and an ultrasound imaging system to provide the best irradiation strategy (electron beam energy, applicator position and bevel angle) and to facilitate the optimisation of dose prescription and delivery to the target volume while maximising the organs at risk sparing. The study has been performed in silico, exploiting MC simulations of a breast cancer treatment. Ultrasound-based input has been used to compute the absorbed dose maps in different irradiation strategies and a quantitative comparison between the different options was carried out using Dose Volume Histograms. The system was capable of exploring different beam energies and applicator positions in few minutes, identifying the best strategy with an overall computation time that was found to be completely compatible with clinical implementation. The systematic uncertainty related to tissue deformation during treatment delivery with respect to imaging acquisition was taken into account. The potential and feasibility of a GPU based full MC TPS implementation of IOeRT breast cancer treatments has been demonstrated in-silico. This long awaited tool will greatly improve the treatment safety and efficacy, overcoming the limits identified within the clinical trials carried out so far.

NRG Oncology and PTCOG Patterns of Practice Survey and Consensus Recommendations on Pencil-Beam Scanning Proton Stereotactic Body Radiation Therapy and Hypofractionated Radiation Therapy for Thoracic Malignancies

Stereotactic body radiation therapy (SBRT) and hypofractionation using pencil-beam scanning (PBS) proton therapy (PBSPT) is an attractive option for thoracic malignancies. Combining the advantages of target coverage conformity and critical organ sparing from both PBSPT and SBRT, this new delivery technique has great potential to improve the therapeutic ratio, particularly for tumors near critical organs. Safe and effective implementation of PBSPT SBRT/hypofractionation to treat thoracic malignancies is more challenging than the conventionally fractionated PBSPT because of concerns of amplified uncertainties at the larger dose per fraction. The NRG Oncology and Particle Therapy Cooperative Group Thoracic Subcommittee surveyed proton centers in the United States to identify practice patterns of thoracic PBSPT SBRT/hypofractionation. From these patterns, we present recommendations for future technical development of proton SBRT/hypofractionation for thoracic treatment. Among other points, the recommendations highlight the need for volumetric image guidance and multiple computed tomography-based robust optimization and robustness tools to minimize further the effect of uncertainties associated with respiratory motion. Advances in direct motion analysis techniques are urgently needed to supplement current motion management techniques.

Technical note: Beamlet-free optimization for Monte-Carlo-based treatment planning in proton therapy

BACKGROUND

Dose calculation and optimization algorithms in proton therapy treatment planning often have high computational requirements regarding time and memory. This can hinder the implementation of efficient workflows in clinics and prevent the use of new, elaborate treatment techniques aiming to improve clinical outcomes like robust optimization, arc, and adaptive proton therapy.

PURPOSE

A new method, namely, the beamlet-free algorithm, is presented to address the aforementioned issue by combining Monte Carlo dose calculation and optimization into a single algorithm and omitting the calculation of the time-consuming and costly dose influence matrix.

METHODS

The beamlet-free algorithm simulates the dose in proton batches of randomly chosen spots and evaluates their relative impact on the objective function at each iteration. Based on the approximated gradient, the spot weights are then updated and used to generate a new spot probability distribution. The beamlet-free method is compared against a conventional, beamlet-based treatment planning algorithm on a brain case and a prostate case.

RESULTS

The beamlet-free algorithm maintained a comparable plan quality while largely reducing the dependence of computation time and memory usage on the number of spots.

CONCLUSION

The implementation of a beamlet-free treatment planning algorithm for proton therapy is feasible and capable of achieving treatment plans of comparable quality to conventional methods. Its efficient usage of computational resources and low spot dependence makes it a promising method for large plans, robust optimization, and arc proton therapy.

A GPU-based fast Monte Carlo code that supports proton transport in magnetic field for radiation therapy

This paper presents the effort to extend a previously reported code ARCHER, a GPU-based Monte Carlo (MC) code for coupled photon and electron transport, into protons including the consideration of magnetic fields. The proton transport is modeled using a Class-II condensed-history algorithm with continuous slowing-down approximation. The model includes ionization, multiple scattering, energy straggling, elastic and inelastic nuclear interactions, as well as deflection due to the Lorentz force in magnetic fields. An additional direction change is added for protons at the end of each step in the presence of the magnetic field. Secondary charge particles, except for protons, are terminated depositing kinetic energies locally, whereas secondary neutral particles are ignored. Each proton is transported step by step until its energy drops to below 0.5 MeV or when the proton leaves the phantom. The code is implemented using the compute unified device architecture (CUDA) platform for optimized GPU thread-level parallelism and efficiency. The code is validated by comparing it against TOPAS. Comparisons of dose distributions between our code and TOPAS for several exposure scenarios, ranging from single square beams in water to patient plan with magnetic fields, show good agreement. The 3D-gamma pass rate with a 2 mm/2% criterion in the region with dose greater than 10% of the maximum dose is computed to be over 99% for all tested cases. Using a single NVIDIA TITAN V GPU card, the computational time of ARCHER is found to range from 0.82 to 4.54 seconds for 1 × 107 proton histories. Compared to a few hours running on TOPAS, this speed improvement is significant. This work presents, for the first time, the performance of a GPU-based MC code to simulate proton transportation magnetic fields, demonstrating the feasibility of accurate and efficient dose calculations in potential magnetic resonance imaging (MRI)-guided proton therapy.

A fast analytical dose calculation approach for MRI-guided proton therapy

Objective.Magnetic resonance (MR) is an innovative technology for online image guidance in conventional radiotherapy and is also starting to be considered for proton therapy as well. For MR-guided therapy, particularly for online plan adaptations, fast dose calculation is essential. Monte Carlo (MC) simulations, however, which are considered the gold standard for proton dose calculations, are very time-consuming. To address the need for an efficient dose calculation approach for MRI-guided proton therapy, we have developed a fast GPU-based modification of an analytical dose calculation algorithm incorporating beam deflections caused by magnetic fields.Approach.Proton beams (70-229 MeV) in orthogonal magnetic fields (0.5/1.5 T) were simulated using TOPAS-MC and central beam trajectories were extracted to generate look-up tables (LUTs) of incremental rotation angles as a function of water-equivalent depth. Beam trajectories are then reconstructed using these LUTs for the modified ray casting dose calculation. The algorithm was validated against MC in water, different materials and for four example patient cases, whereby it has also been fully incorporated into a treatment plan optimisation regime.Main results.Excellent agreement between analytical and MC dose distributions could be observed with sub-millimetre range deviations and differences in lateral shifts <2 mm even for high densities (1000 HU). 2%/2 mm gamma pass rates were comparable to the 0 T scenario and above 94.5% apart for the lung case. Further, comparable treatment plan quality could be achieved regardless of magnetic field strength.Significance.A new method for accurate and fast proton dose calculation in magnetic fields has been developed and successfully implemented for treatment plan optimisation.

Accurate and Fast Deep Learning Dose Prediction for a Preclinical Microbeam Radiation Therapy Study Using Low-Statistics Monte Carlo Simulations

Microbeam radiation therapy (MRT) utilizes coplanar synchrotron radiation beamlets and is a proposed treatment approach for several tumor diagnoses that currently have poor clinical treatment outcomes, such as gliosarcomas. Monte Carlo (MC) simulations are one of the most used methods at the Imaging and Medical Beamline, Australian Synchrotron to calculate the dose in MRT preclinical studies. The steep dose gradients associated with the 50μm-wide coplanar beamlets present a significant challenge for precise MC simulation of the dose deposition of an MRT irradiation treatment field in a short time frame. The long computation times inhibit the ability to perform dose optimization in treatment planning or apply online image-adaptive radiotherapy techniques to MRT. Much research has been conducted on fast dose estimation methods for clinically available treatments. However, such methods, including GPU Monte Carlo implementations and machine learning (ML) models, are unavailable for novel and emerging cancer radiotherapy options such as MRT. In this work, the successful application of a fast and accurate ML dose prediction model for a preclinical MRT rodent study is presented for the first time. The ML model predicts the peak doses in the path of the microbeams and the valley doses between them, delivered to the tumor target in rat patients. A CT imaging dataset is used to generate digital phantoms for each patient. Augmented variations of the digital phantoms are used to simulate with Geant4 the energy depositions of an MRT beam inside the phantoms with 15% (high-noise) and 2% (low-noise) statistical uncertainty. The high-noise MC simulation data are used to train the ML model to predict the energy depositions in the digital phantoms. The low-noise MC simulations data are used to test the predictive power of the ML model. The predictions of the ML model show an agreement within 3% with low-noise MC simulations for at least 77.6% of all predicted voxels (at least 95.9% of voxels containing tumor) in the case of the valley dose prediction and for at least 93.9% of all predicted voxels (100.0% of voxels containing tumor) in the case of the peak dose prediction. The successful use of high-noise MC simulations for the training, which are much faster to produce, accelerates the production of the training data of the ML model and encourages transfer of the ML model to different treatment modalities for other future applications in novel radiation cancer therapies.

GPU-accelerated Monte Carlo simulation of electron and photon interactions for radiotherapy applications

Objective. The Monte Carlo simulation software is a valuable tool in radiation therapy, in particular to achieve the needed accuracy in the dose evaluation for the treatment plans optimisation. The current challenge in this field is the time reduction to open the way to many clinical applications for which the computational time is an issue. In this manuscript we present an innovative GPU-accelerated Monte Carlo software for dose valuation in electron and photon based radiotherapy, developed as an update of the FRED (Fast paRticle thErapy Dose evaluator) software.Approach. The code transports particles through a 3D voxel grid, while scoring their energy deposition along their trajectory. The models of electromagnetic interactions in the energy region between 1 MeV-1 GeV available in literature have been implemented to efficiently run on GPUs, allowing to combine a fast tracking while keeping high accuracy in dose assessment. The FRED software has been bench-marked against state-of-art full MC (FLUKA, GEANT4) in the realm of two different radiotherapy applications: Intra-Operative Radio Therapy and Very High Electron Energy radiotherapy applications.Results. The single pencil beam dose-depth profiles in water as well as the dose map computed on non-homogeneous phantom agree with full-MCs at 2% level, observing a gain in processing time from 200 to 5000.Significance. Such performance allows for computing a plan with electron beams in few minutes with an accuracy of ∼%, demonstrating the FRED potential to be adopted for fast plan re-calculation in photon or electron radiotherapy applications.

GPU accelerated Monte Carlo scoring of positron emitting isotopes produced during proton therapy for PET verification

Objective.Verification of delivered proton therapy treatments is essential for reaping the many benefits of the modality, with the most widely proposedin vivoverification technique being the imaging of positron emitting isotopes generated in the patient during treatment using positron emission tomography (PET). The purpose of this work is to reduce the computational resources and time required for simulation of patient activation during proton therapy using the GPU accelerated Monte Carlo code FRED, and to validate the predicted activity against the widely used Monte Carlo code GATE.Approach.We implement a continuous scoring approach for the production of positron emitting isotopes within FRED version 5.59.9. We simulate treatment plans delivered to 95 head and neck patients at Centrum Cyklotronowe Bronowice using this GPU implementation, and verify the accuracy using the Monte Carlo toolkit GATE version 9.0.Main results.We report an average reduction in computational time by a factor of 50 when using a local system with 2 GPUs as opposed to a large compute cluster utilising between 200 to 700 CPU threads, enabling simulation of patient activity within an average of 2.9 min as opposed to 146 min. All simulated plans are in good agreement across the two Monte Carlo codes. The two codes agree within a maximum of 0.95σon a voxel-by-voxel basis for the prediction of 7 different isotopes across 472 simulated fields delivered to 95 patients, with the average deviation over all fields being 6.4 × 10^-3σ.Significance.The implementation of activation calculations in the GPU accelerated Monte Carlo code FRED provides fast and reliable simulation of patient activation following proton therapy, allowing for research and development of clinical applications of range verification for this treatment modality using PET to proceed at a rapid pace.

Virtual particle Monte Carlo: A new concept to avoid simulating secondary particles in proton therapy dose calculation

BACKGROUND

In proton therapy dose calculation, Monte Carlo (MC) simulations are superior in accuracy but more time consuming, compared to analytical calculations. Graphic processing units (GPUs) are effective in accelerating MC simulations but may suffer thread divergence and racing condition in GPU threads that degrades the computing performance due to the generation of secondary particles during nuclear reactions.

PURPOSE

A novel concept of virtual particle (VP) MC (VPMC) is proposed to avoid simulating secondary particles in GPU-accelerated proton MC dose calculation and take full advantage of the computing power of GPU.

METHODS

Neutrons and gamma rays were ignored as escaping from the human body; doses of electrons, heavy ions, and nuclear fragments were locally deposited; the tracks of deuterons were converted into tracks of protons. These particles, together with primary and secondary protons, are considered to be the realistic particles. Histories of primary and secondary protons were replaced by histories of multiple VPs. Each VP corresponded to one proton (either primary or secondary). A continuous-slowing-down-approximation model, an ionization model, and a large angle scattering event model corresponding to nuclear interactions were developed for VPs by generating probability distribution functions (PDFs) based on simulation results of realistic particles using MCsquare. For efficient calculations, these PDFs were stored in the Compute Unified Device Architecture textures. VPMC was benchmarked with TOPAS and MCsquare in phantoms and with MCsquare in 13 representative patient geometries. Comparisons between the VPMC calculated dose and dose measured in water during patient-specific quality assurance (PSQA) of the selected 13 patients were also carried out. Gamma analysis was used to compare the doses derived from different methods and calculation efficiencies were also compared.

RESULTS

Integrated depth dose and lateral dose profiles in both homogeneous and inhomogeneous phantoms all matched well among VPMC, TOPAS, and MCsquare calculations. The 3D-3D gamma passing rates with a criterion of 2%/2 mm and a threshold of 10% was 98.49% between MCsquare and TOPAS and 98.31% between VPMC and TOPAS in homogeneous phantoms, and 99.18% between MCsquare and TOPAS and 98.49% between VPMC and TOPAS in inhomogeneous phantoms, respectively. In patient geometries, the 3D-3D gamma passing rates with 2%/2 mm/10% between dose distributions from VPMC and MCsquare were 98.56 ± 1.09% in patient geometries. The 2D-3D gamma analysis with 3%/2 mm/10% between the VPMC calculated dose distributions and the 2D measured planar dose distributions during PSQA was 98.91 ± 0.88%. VPMC calculation was highly efficient and took 2.84 ± 2.44 s to finish for the selected 13 patients running on four NVIDIA Ampere GPUs in patient geometries.

CONCLUSION

VPMC was found to achieve high accuracy and efficiency in proton therapy dose calculation.

ProTheRaMon - a GATE simulation framework for proton therapy range monitoring using PET imaging

OBJECTIVE

This paper reports on the implementation and shows examples of the use of the ProTheRaMon framework for simulating the delivery of proton therapy treatment plans and range monitoring using positron emission tomography (PET). ProTheRaMon offers complete processing of proton therapy treatment plans, patient CT geometries, and intra-treatment PET imaging, taking into account therapy and imaging coordinate systems and activity decay during the PET imaging protocol specific to a given proton therapy facility. We present the ProTheRaMon framework and illustrate its potential use case and data processing steps for a patient treated at the Cyclotron Centre Bronowice (CCB) proton therapy center in Krakow, Poland.

APPROACH

The ProTheRaMon framework is based on GATE Monte Carlo software, the CASToR reconstruction package and in-house developed Python and bash scripts. The framework consists of five separated simulation and data processing steps, that can be further optimized according to the user's needs and specific settings of a given proton therapy facility and PET scanner design.

MAIN RESULTS

ProTheRaMon is presented using example data from a patient treated at CCB and the J-PET scanner to demonstrate the application of the framework for proton therapy range monitoring. The output of each simulation and data processing stage is described and visualized.

SIGNIFICANCE

We demonstrate that the ProTheRaMon simulation platform is a high-performance tool, capable of running on a computational cluster and suitable for multi-parameter studies, with databases consisting of large number of patients, as well as different PET scanner geometries and settings for range monitoring in a clinical environment. Due to its modular structure, the ProTheRaMon framework can be adjusted for different proton therapy centers and/or different PET detector geometries. It is available to the community via github.

MOQUI: an open-source GPU-based Monte Carlo code for proton dose calculation with efficient data structure

Objective.Monte Carlo (MC) codes are increasingly used for accurate radiotherapy dose calculation. In proton therapy, the accuracy of the dose calculation algorithm is expected to have a more significant impact than in photon therapy due to the depth-dose characteristics of proton beams. However, MC simulations come at a considerable computational cost to achieve statistically sufficient accuracy. There have been efforts to improve computational efficiency while maintaining sufficient accuracy. Among those, parallelizing particle transportation using graphic processing units (GPU) achieved significant improvements. Contrary to the central processing unit, a GPU has limited memory capacity and is not expandable. It is therefore challenging to score quantities with large dimensions requiring extensive memory. The objective of this study is to develop an open-source GPU-based MC package capable of scoring those quantities.Approach.We employed a hash-table, one of the key-value pair data structures, to efficiently utilize the limited memory of the GPU and score the quantities requiring a large amount of memory. With the hash table, only voxels interacting with particles will occupy memory, and we can search the data efficiently to determine their address. The hash-table was integrated with a novel GPU-based MC code, moqui.Main results.The developed code was validated against an MC code widely used in proton therapy, TOPAS, with homogeneous and heterogeneous phantoms. We also compared the dose calculation results of clinical treatment plans. The developed code agreed with TOPAS within 2%, except for the fall-off and regions, and the gamma pass rates of the results were >99% for all cases with a 2 mm/2% criteria.Significance.We can score dose-influence matrix and dose-rate on a GPU for a 3-field H&N case with 10 GB of memory using moqui, which would require more than 100 GB of memory with the conventionally used array data structure.

A Data-Driven Fragmentation Model for Carbon Therapy GPU-Accelerated Monte-Carlo Dose Recalculation

The advent of Graphics Processing Units (GPU) has prompted the development of Monte Carlo (MC) algorithms that can significantly reduce the simulation time with respect to standard MC algorithms based on Central Processing Unit (CPU) hardware. The possibility to evaluate a complete treatment plan within minutes, instead of hours, paves the way for many clinical applications where the time-factor is important. FRED (Fast paRticle thErapy Dose evaluator) is a software that exploits the GPU power to recalculate and optimise ion beam treatment plans. The main goal when developing the FRED physics model was to balance accuracy, calculation time and GPU execution guidelines. Nowadays, FRED is already used as a quality assurance tool in Maastricht and Krakow proton clinical centers and as a research tool in several clinical and research centers across Europe. Lately the core software has been updated including a model of carbon ions interactions with matter. The implementation is phenomenological and based on carbon fragmentation data currently available. The model has been tested against the MC FLUKA software, commonly used in particle therapy, and a good agreement was found. In this paper, the new FRED data-driven model for carbon ion fragmentation will be presented together with the validation tests against the FLUKA MC software. The results will be discussed in the context of FRED clinical applications to ¹²C ions treatment planning.

Quantification of biological range uncertainties in patients treated at the Krakow proton therapy centre

Background

Variable relative biological effectiveness (vRBE) in proton therapy might significantly modify the prediction of RBE-weighted dose delivered to a patient during proton therapy. In this study we will present a method to quantify the biological range extension of the proton beam, which results from the application of vRBE approach in RBE-weighted dose calculation.

Methods and materials

The treatment plans of 95 patients (brain and skull base patients) were used for RBE-weighted dose calculation with constant and the McNamara RBE model. For this purpose the Monte Carlo tool FRED was used. The RBE-weighted dose distributions were analysed using indices from dose-volume histograms. We used the volumes receiving at least 95% of the prescribed dose (V95) to estimate the biological range extension resulting from vRBE approach.

Results

The vRBE model shows higher median value of relative deposited dose and D95 in the planning target volume by around 1% for brain patients and 4% for skull base patients. The maximum doses in organs at risk calculated with vRBE was up to 14 Gy above dose limit. The mean biological range extension was greater than 0.4 cm.

Discussion

Our method of estimation of biological range extension is insensitive for dose inhomogeneities and can be easily used for different proton plans with intensity-modulated proton therapy (IMPT) optimization. Using volumes instead of dose profiles, which is the common method, is more universal. However it was tested only for IMPT plans on fields arranged around the tumor area.

Conclusions

Adopting a vRBE model results in an increase in dose and an extension of the beam range, which is especially disadvantageous in cancers close to organs at risk. Our results support the need to re-optimization of proton treatment plans when considering vRBE.

Supplementary Information

The online version contains supplementary material available at 10.1186/s13014-022-02022-5.

Adaptive proton therapy

Radiation therapy treatments are typically planned based on a single image set, assuming that the patient's anatomy and its position relative to the delivery system remains constant during the course of treatment. Similarly, the prescription dose assumes constant biological dose-response over the treatment course. However, variations can and do occur on multiple time scales. For treatment sites with significant intra-fractional motion, geometric changes happen over seconds or minutes, while biological considerations change over days or weeks. At an intermediate timescale, geometric changes occur between daily treatment fractions. Adaptive radiation therapy is applied to consider changes in patient anatomy during the course of fractionated treatment delivery. While traditionally adaptation has been done off-line with replanning based on new CT images, online treatment adaptation based on on-board imaging has gained momentum in recent years due to advanced imaging techniques combined with treatment delivery systems. Adaptation is particularly important in proton therapy where small changes in patient anatomy can lead to significant dose perturbations due to the dose conformality and finite range of proton beams. This review summarizes the current state-of-the-art of on-line adaptive proton therapy and identifies areas requiring further research.

Study of relationship between dose, LET and the risk of brain necrosis after proton therapy for skull base tumors

PURPOSE

We investigated the relationship between RBE-weighted dose (DRBE) calculated with constant (cRBE) and variable RBE (vRBE), dose-averaged linear energy transfer (LETd) and the risk of radiographic changes in skull base patients treated with protons.

METHODS

Clinical treatment plans of 45 patients were recalculated with Monte Carlo tool FRED. Radiographic changes (i.e. edema and/or necrosis) were identified by MRI. Dosimetric parameters for cRBE and vRBE were computed. Biological margin extension and voxel-based analysis were employed looking for association of DRBE(vRBE) and LETd with brain edema and/or necrosis.

RESULTS

When using vRBE, Dmax in the brain was above the highest dose limits for 38% of patients, while such limit was never exceeded assuming cRBE. Similar values of Dmax were observed in necrotic regions, brain and temporal lobes. Most of the brain necrosis was in proximity to the PTV. The voxel-based analysis did not show evidence of an association with high LETd values.

CONCLUSIONS

When looking at standard dosimetric parameters, the higher dose associated with vRBE seems to be responsible for an enhanced risk of radiographic changes. However, as revealed by a voxel-based analysis, the large inter-patient variability hinders the identification of a clear effect for high LETd.

Dose Calculation Algorithms for External Radiation Therapy: An Overview for Practitioners

Radiation therapy (RT) is a constantly evolving therapeutic technique; improvements are continuously being introduced for both methodological and practical aspects. Among the features that have undergone a huge evolution in recent decades, dose calculation algorithms are still rapidly changing. This process is propelled by the awareness that the agreement between the delivered and calculated doses is of paramount relevance in RT, since it could largely affect clinical outcomes. The aim of this work is to provide an overall picture of the main dose calculation algorithms currently used in RT, summarizing their underlying physical models and mathematical bases, and highlighting their strengths and weaknesses, referring to the most recent studies on algorithm comparisons. This handy guide is meant to provide a clear and concise overview of the topic, which will prove useful in helping clinical medical physicists to perform their responsibilities more effectively and efficiently, increasing patient benefits and improving the overall quality of the management of radiation treatment.

Clinical validation of a GPU-based Monte Carlo dose engine of a commercial treatment planning system for pencil beam scanning proton therapy

PURPOSE

To perform the validation of the GPU-based (Graphical Processing Unit based) proton Monte Carlo (MC) dose engine implemented in a commercial TPS (RayStation 10B) and to report final dose calculation times for clinical cases.

MATERIALS AND METHODS

440 patients treated at the Proton Therapy Center of Trento, Italy, between 2018 and 2019 were selected for this study. 636 approved plans with 3361 beams computed with the clinically implemented CPU-MC dose engine (version 4.2 and 4.5), were used for the validation of the new algorithm. For each beam, the dose was recalculated using the new GPU-MC dose engine with the initial CPU computation settings and compared to the original CPU-MC dose. Beam dose difference distributions were studied to ensure that the two dose distributions were equal within the expected fluctuations of the MC statistical uncertainty (s) of each computation. Plan dose distributions were compared with respect to the dosimetric indices D₉₈, D₅₀ and D₁ of all ROIs defined as targets. A complete assessment of the computation time as a function of s and dose grid voxel size was done.

RESULTS

The median over all mean beam dose differences between CPU- and GPU-MC was -0.01% and the median of the corresponding standard deviations was close to (√2s) both for simulations with an s of 0.5% and 1.0% per beam. This shows that the two dose distributions can be considered equal. All the DVH indices showed an average difference below 0.04%. About half of the plans were computed with 1.0% statistical uncertainty on a 2 mm dose calculation grid, for which the median computation time was 5.2 s. The median computational speed for all plans in the study was 8.4 million protons/second.

CONCLUSION

A validation of a clinical MC algorithm running on GPU was performed on a large pool of patients treated with pencil beam scanning proton therapy. We demonstrated that the differences with the previous CPU-based MC were only due to the intrinsic statistical fluctuations of the MC method, which translated to insignificant differences on plan dose level. The significant increase in dose calculation speed is expected to facilitate new clinical workflows.

A Critical Review of LET-Based Intensity-Modulated Proton Therapy Plan Evaluation and Optimization for Head and Neck Cancer Management

In this review article, we review the 3 important aspects of linear-energy-transfer (LET) in intensity-modulated proton therapy (IMPT) for head and neck (H&N) cancer management. Accurate LET calculation methods are essential for LET-guided plan evaluation and optimization, which can be calculated either by analytical methods or by Monte Carlo (MC) simulations. Recently, some new 3D analytical approaches to calculate LET accurately and efficiently have been proposed. On the other hand, several fast MC codes have also been developed to speed up the MC simulation by simplifying nonessential physics models and/or using the graphics processor unit (GPU)–acceleration approach. Some concepts related to LET are also briefly summarized including (1) dose-weighted versus fluence-weighted LET; (2) restricted versus unrestricted LET; and (3) microdosimetry versus macrodosimetry. LET-guided plan evaluation has been clinically done in some proton centers. Recently, more and more studies using patient outcomes as the biological endpoint have shown a positive correlation between high LET and adverse events sites, indicating the importance of LET-guided plan evaluation in proton clinics. Various LET-guided plan optimization methods have been proposed to generate proton plans to achieve biologically optimized IMPT plans. Different optimization frameworks were used, including 2-step optimization, 1-step optimization, and worst-case robust optimization. They either indirectly or directly optimize the LET distribution in patients while trying to maintain the same dose distribution and plan robustness. It is important to consider the impact of uncertainties in LET-guided optimization (ie, LET-guided robust optimization) in IMPT, since IMPT is sensitive to uncertainties including both the dose and LET distributions. We believe that the advancement of the LET-guided plan evaluation and optimization will help us exploit the unique biological characteristics of proton beams to improve the therapeutic ratio of IMPT to treat H&N and other cancers.

Dosimetric influence of deformable image registration uncertainties on propagated structures for online daily adaptive proton therapy of lung cancer patients

PURPOSE

A major burden of introducing an online daily adaptive proton therapy (DAPT) workflow is the time and resources needed to correct the daily propagated contours. In this study, we evaluated the dosimetric impact of neglecting the online correction of the propagated contours in a DAPT workflow.

MATERIAL AND METHODS

For five NSCLC patients with nine repeated deep-inspiration breath-hold CTs, proton therapy plans were optimised on the planning CT to deliver 60 Gy-RBE in 30 fractions. All repeated CTs were registered with six different clinically used deformable image registration (DIR) algorithms to the corresponding planning CT. Structures were propagated rigidly and with each DIR algorithm and reference structures were contoured on each repeated CT. DAPT plans were optimised with the uncorrected, propagated structures (propagated DAPT doses) and on the reference structures (ideal DAPT doses), non-adapted doses were recalculated on all repeated CTs.

RESULTS

Due to anatomical changes occurring during the therapy, the clinical target volume (CTV) coverage of the non-adapted doses reduces on average by 9.7% (V95) compared to an ideal DAPT doses. For the propagated DAPT doses, the CTV coverage was always restored (average differences in the CTV V95 < 1% compared to the ideal DAPT doses). Hotspots were always reduced with any DAPT approach.

CONCLUSION

For the patients presented here, a benefit of online DAPT was shown, even if the daily optimisation is based on propagated structures with some residual uncertainties. However, a careful (offline) structure review is necessary and corrections can be included in an offline adaption.

Rapid effective dose calculation for raster-scanning 4He ion therapy with the modified microdosimetric kinetic model (mMKM)

PURPOSE

To develop and verify effective dose (D_RBE) calculation in ⁴He ion beam therapy based on the modified microdosimetric kinetic model (mMKM) and evaluate the bio-sensitivity of mMKM-based plans to clinical parameters using a fast analytical dose engine.

METHODS

Mixed radiation field particle spectra (MRFS) databases have been generated with Monte-Carlo (MC) simulations for ⁴He-ion beams. Relative biological effectiveness (RBE) and D_RBE calculation using MRFS were established within a fast analytical engine. Spread-out Bragg-Peaks (SOBPs) in water were optimized for two dose levels and two tissue types with photon linear-quadratic model parameters α_ph, β_ph, and (α/β)_ph to verify MRFS-derived database implementation against computations with MC-generated mixed-field α and β databases. Bio-sensitivity of the SOBPs was investigated by varying absolute values of β_ph, while keeping (α/β)_ph constant. Additionally, dose, dose-averaged linear energy transfer, and bio-sensitivity were investigated for two patient cases.

RESULTS

Using MRFS-derived databases, dose differences ≲2% in the plateau and SOBP are observed compared to computations with MC-generated databases. Bio-sensitivity studies show larger deviations when altering the absolute β_ph value, with maximum D_50% changes of ~5%, with similar results for patient cases. Bio-sensitivity analysis indicates a greater impact on D_RBE varying (α/β)_ph than β_ph in mMKM.

CONCLUSIONS

The MRSF approach yielded negligible differences in the target and small differences in the plateau compared to MC-generated databases. The presented analyses provide guidance for proper implementation of RBE-weighted ⁴He ion dose prescription and planning with mMKM. The MRFS-D_RBE calculation approach using mMKM will be implemented in a clinical treatment planning system.

The practical radius of a pencil beam in proton therapy

The central Gaussian shaped high dose region of a pencil beam (PB) in light ion beam therapy (LIBT) is enveloped by a low dose region causing non-negligible field size effects and impairs the dose calculation accuracy considerably if the low dose envelope is not well modeled. The purpose of this study was to calculate the practical radius, R_c, at which a PB does not influence a field more than a certain accuracy level. Lateral dose profiles of proton beams in water were simulated using GATE/Geant4. Those lateral dose profiles were integrated numerically and used to calculate field size factors (FSFs). The R_c was then determined such, that the lateral dose at radii exceeding R_c can be neglected without compromising the FSF of a 20cm×20cm field more than a desired accuracy level c. The practical radius R_c yielding c=0.5% was compared to the frequently applied concept of full width at a ratio x of the maximum (FWxM). The sensitivity to variations of the beam width was tested by increasing the initial beam width σ_C of the clinical beam model by 0.5 and 1mm, respectively. Neglecting the dose at radii exceeding R_c resulted in the desired FSF accuracy, whereas using the FW0.01%M cut resulted in varying accuracy. In order to yield a constant FSF accuracy, the ratio x in FWxM ranged from 0.003% to 0.065% of the maximum. In contrast to R_c, FWxM was sensitive to variations of the initial beam width. The maximum R_c over all depths was less than 7cm for the low(62.4MeV) and medium(148.2MeV) proton energy beam, which suggests that a plane parallel ionization chamber exceeding that radius is sufficient to acquire laterally integrated depth dose distributions for those energies. However, this holds not true for the highest energy (252.7MeV) or when including a range shifter (RaShi). The values of R_c are specific to our beam line configuration as the maximum R_c was depending on both, the scattering material in the Nozzle as well as the distance of the air-gap between Nozzle and phantom.

Preliminary results in using Deep Learning to emulate BLOB, a nuclear interaction model

PURPOSE

A reliable model to simulate nuclear interactions is fundamental for Ion-therapy. We already showed how BLOB ("Boltzmann-Langevin One Body"), a model developed to simulate heavy ion interactions up to few hundreds of MeV/u, could simulate also ¹²C reactions in the same energy domain. However, its computation time is too long for any medical application. For this reason we present the possibility of emulating it with a Deep Learning algorithm.

METHODS

The BLOB final state is a Probability Density Function (PDF) of finding a nucleon in a position of the phase space. We discretised this PDF and trained a Variational Auto-Encoder (VAE) to reproduce such a discrete PDF. As a proof of concept, we developed and trained a VAE to emulate BLOB in simulating the interactions of ¹²C with ¹²C at 62 MeV/u. To have more control on the generation, we forced the VAE latent space to be organised with respect to the impact parameter (b) training a classifier of b jointly with the VAE.

RESULTS

The distributions obtained from the VAE are similar to the input ones and the computation time needed to use the VAE as a generator is negligible.

CONCLUSIONS

We show that it is possible to use a Deep Learning approach to emulate a model developed to simulate nuclear reactions in the energy range of interest for Ion-therapy. We foresee the implementation of the generation part in C++ and to interface it with the most used Monte Carlo toolkit: Geant4.

Technical Note: Integrating an open source Monte Carlo code "MCsquare" for clinical use in intensity-modulated proton therapy

PURPOSE

To commission an open source Monte Carlo (MC) dose engine, "MCsquare" for a synchrotron-based proton machine, integrate it into our in-house C++-based I/O user interface and our web-based software platform, expand its functionalities, and improve calculation efficiency for intensity-modulated proton therapy (IMPT).

METHODS

We commissioned MCsquare using a double Gaussian beam model based on in-air lateral profiles, integrated depth dose of 97 beam energies, and measurements of various spread-out Bragg peaks (SOBPs). Then we integrated MCsquare into our C++-based dose calculation code and web-based second check platform "DOSeCHECK." We validated the commissioned MCsquare based on 12 different patient geometries and compared the dose calculation with a well-benchmarked GPU-accelerated MC (gMC) dose engine. We further improved the MCsquare efficiency by employing the computed tomography (CT) resampling approach. We also expanded its functionality by adding a linear energy transfer (LET)-related model-dependent biological dose calculation.

RESULTS

Differences between MCsquare calculations and SOBP measurements were <2.5% (<1.5% for ~85% of measurements) in water. The dose distributions calculated using MCsquare agreed well with the results calculated using gMC in patient geometries. The average 3D gamma analysis (2%/2 mm) passing rates comparing MCsquare and gMC calculations in the 12 patient geometries were 98.0 ± 1.0%. The computation time to calculate one IMPT plan in patients' geometries using an inexpensive CPU workstation (Intel Xeon E5-2680 2.50 GHz) was 2.3 ± 1.8 min after the variable resolution technique was adopted. All calculations except for one craniospinal patient were finished within 3.5 min.

CONCLUSIONS

MCsquare was successfully commissioned for a synchrotron-based proton beam therapy delivery system and integrated into our web-based second check platform. After adopting CT resampling and implementing LET model-dependent biological dose calculation capabilities, MCsquare will be sufficiently efficient and powerful to achieve Monte Carlo-based and LET-guided robust optimization in IMPT, which will be done in the future studies.

A track repeating algorithm for intensity modulated carbon ion therapy

The fast dose calculator (FDC), a track repeating Monte Carlo (MC) algorithm was initially developed for proton therapy. The validation for proton therapy has been demonstrated in a previous work. In this work we presented the extension of FDC to the calculation of dose distributions for ions, particularly for carbon. Moreover the code algorithm is validated by comparing 3D dose distributions and dose volume histograms (DVH) calculated by FDC with Geant4. A total of 19 patients were employed, including three patients of prostate, five of brain, three of head and neck, four of lung and four of spine. We used a gamma-index technique to analyze dose distributions and we performed a dosimetric analysis for DVHs, a more direct and informative quantity for planning system assessment. The gamma-index passing rates of all patients discussed in this paper are above 90% with the criterion 1%/1 mm, above 98% with the criterion 2%/2 mm and over 99.9% with the criterion 3%/3 mm. The root mean square (RMS) of percent difference of dosimetric indices D ₀₂, D ₀₅, D ₅₀, D ₉₅ and D ₉₈ are 0.75%, 0.70%, 0.79%, 0.83% and 0.76%. All the differences are within clinically accepted norms.

RIDOS: A new system for online computation of the delivered dose distributions in scanning ion beam therapy

PURPOSE

To describe a new system for scanned ion beam therapy, named RIDOS (Real-time Ion DOse planning and delivery System), which performs real time delivered dose verification integrating the information from a clinical beam monitoring system with a Graphic Processing Unit (GPU) based dose calculation in patient Computed Tomography.

METHODS

A benchmarked dose computation algorithm for scanned ion beams has been parallelized and adapted to run on a GPU architecture. A workstation equipped with a NVIDIA GPU has been interfaced through a National Instruments PXI-crate with the dose delivery system of the Italian National Center of Oncological Hadrontherapy (CNAO) to receive in real-time the measured beam parameters. Data from a patient monitoring system are also collected to associate the respiratory phases with each spot during the delivery of the dose. Using both measured and planned spot properties, RIDOS evaluates during the few seconds of inter-spill time the cumulative delivered and prescribed dose distributions and compares them through a fast γ-index algorithm.

RESULTS

The accuracy of the GPU-based algorithms was assessed against the CPU-based ones and the differences were found below 1‰. The cumulative planned and delivered doses are computed at the end of each spill in about 300 ms, while the dose comparison takes approximatively 400 ms. The whole operation provides the results before the next spill starts.

CONCLUSIONS

RIDOS system is able to provide a fast computation of the delivered dose in the inter-spill time of the CNAO facility and allows to monitor online the dose deposition accuracy all along the treatment.

Evaluation of electromagnetic and nuclear scattering models in GATE/Geant4 for proton therapy

Purpose

The dose core of a proton pencil beam (PB) is enveloped by a low dose area reaching several centimeters off the central axis and containing a considerable amount of the dose. Adequate modeling of the different components of the PB profile is, therefore, required for accurate dose calculation. In this study, we experimentally validated one electromagnetic and two nuclear scattering models in GATE/Geant4 for dose calculation of proton beams in the therapeutic energy window (62–252 MeV) with and without range shifter (RaShi).

Methods

The multiple Coulomb scattering (MCS) model was validated by lateral dose core profiles measured for five energies at up to four depths from beam plateau to Bragg peak region. Nuclear halo profiles of single PBs were evaluated for three (62.4, 148.2, and 252.7 MeV) and two (97.4 and 124.7 MeV) energies, without and with RaShi, respectively. The influence of the dose core and nuclear halo on field sizes varying from 2–20 cm was evaluated by means of output factors (OFs), namely frame factors (FFs) and field size factors (FSFs), to quantify the relative increase of dose when increasing the field size.

Results

The relative increase in the dose core width in the simulations deviated negligibly from measurements for depths until 80% of the beam range, but was overestimated by up to 0.2 mm in σ toward the end of range for all energies. The dose halo region of the lateral dose profile agreed well with measurements in the open beam configuration, but was notably overestimated in the deepest measurement plane of the highest energy or when the beam passed through the RaShi. The root‐mean‐square deviations (RMSDs) between the simulated and the measured FSFs were less than 1% at all depths, but were higher in the second half of the beam range as compared to the first half or when traversing the RaShi. The deviations in one of the two tested hadron physics lists originated mostly in elastic scattering. The RMSDs could be reduced by approximately a factor of two by exchanging the default elastic scattering cross sections for protons.

Conclusions

GATE/Geant4 agreed satisfyingly with most measured quantities. MCS was systematically overestimated toward the end of the beam range. Contributions from nuclear scattering were overestimated when the beam traversed the RaShi or at the depths close to the end of the beam range without RaShi. Both, field size effects and calculation uncertainties, increased when the beam traversed the RaShi. Measured field size effects were almost negligible for beams up to medium energy and were highest for the highest energy beam without RaShi, but vice versa when traversing the RaShi.

Regularised patient-specific stopping power calibration for proton therapy planning based on proton radiographic images

Proton transmission imaging has been proposed and investigated as imaging modality complementary to x-ray based techniques in proton beam therapy. In particular, it addresses the issue of range uncertainties due to the conversion of an x-ray patient computed tomography (CT) image expressed in Hounsfield Units (HU) to relative stopping power (RSP) needed as input to the treatment planning system. One approach to exploit a single proton radiographic projection is to perform a patient-specific calibration of the CT to RSP conversion curve by optimising the match between a measured and a numerically integrated proton radiography. In this work, we develop the mathematical tools needed to perform such an optimisation in an efficient and robust way. Our main focus lies on set-ups which combine pencil beam scanning with a range telescope detector, although most of our methods can be employed in combination with other set-ups as well. Proton radiographies are simulated in Monte Carlo using an idealised detector and applying the same data processing chain used with experimental data. This approach allows us to have a ground truth CT-RSP curve to compare the optimisation results with. Our results show that the parameters of the CT-RSP curve are strongly correlated when using a pencil beam based set-up, which leads to unrealistic variation in the optimised CT-RSP curves. To address this issue, we introduce a regularisation procedure which guarantees a plausible degree of smoothness in the optimised CT-RSP curves. We investigate three different methods to perform the numerical projection operation needed to generate a proton digitally reconstructed radiography. We find that the approximate and computationally faster method performs as well as the more accurate but more demanding method. We perform a Monte Carlo experiment based on a head and neck patient to evaluate the range accuracy achievable with the optimised CT-RSP curves and find an agreement with the ground truth expectation of better than [Formula: see text]. Our results further indicate that the region in the patient in which the proton radiography is acquired does not necessarily have to correspond to the treatment volume to achieve this accuracy. This is important as the imaged region could be freely chosen, e.g. in order to spare organs at risk.

Advanced Treatment Planning

Treatment planning for protons and heavier ions is adapting technologies originally developed for photon dose optimization, but also has to meet its particular challenges. Since the quality of the applied dose is more sensitive to geometric uncertainties, treatment plan robust optimization has a much more prominent role in particle therapy. This has led to specific planning tools, approaches, and research into new formulations of the robust optimization problems. Tools for solution space navigation and automatic planning are also being adapted to particle therapy. These challenges become even greater when detailed models of relative biological effectiveness (RBE) are included into dose optimization, as is required for heavier ions.

Fast robust dose calculation on GPU for high-precision 1H, 4He, 12C and 16O ion therapy: the FRoG platform

Radiotherapy with protons and heavier ions landmarks a novel era in the field of high-precision cancer therapy. To identify patients most benefiting from this technologically demanding therapy, fast assessment of comparative treatment plans utilizing different ion species is urgently needed. Moreover, to overcome uncertainties of actual in-vivo physical dose distribution and biological effects elicited by different radiation qualities, development of a reliable high-throughput algorithm is required. To this end, we engineered a unique graphics processing unit (GPU) based software architecture allowing rapid and robust dose calculation. FRoG, Fast Recalculation on GPU, currently operates with four particle beams available at Heidelberg Ion Beam Therapy center, i.e., raster-scanning proton (¹H), helium (⁴He), carbon (¹²C) and oxygen ions (¹⁶O). FRoG enables comparative analysis of different models for estimation of physical and biological effective dose in 3D within minutes and in excellent agreement with the gold standard Monte Carlo (MC) simulation. This is a crucial step towards development of next-generation patient specific radiotherapy.

Clinical implementations of 4D pencil beam scanned particle therapy: Report on the 4D treatment planning workshop 2016 and 2017

In 2016 and 2017, the 8th and 9th 4D treatment planning workshop took place in Groningen (the Netherlands) and Vienna (Austria), respectively. This annual workshop brings together international experts to discuss research, advances in clinical implementation as well as problems and challenges in 4D treatment planning, mainly in spot scanned proton therapy. In the last two years several aspects like treatment planning, beam delivery, Monte Carlo simulations, motion modeling and monitoring, QA phantoms as well as 4D imaging were thoroughly discussed. This report provides an overview of discussed topics, recent findings and literature review from the last two years. Its main focus is to highlight translation of 4D research into clinical practice and to discuss remaining challenges and pitfalls that still need to be addressed and to be overcome.