Voxel-wise dose rate calculation in clinical pencil beam scanning proton therapy.

OBJECTIVE: Clinical outcomes after proton therapy have shown some variability that is not fully understood. Different approaches have been suggested to explain the biological outcome, but none has yet provided a comprehensive and satisfactory rationale for observed toxicities. The relatively recent transition from passive scattering (PS) to pencil beam scanning (PBS) treatments has significantly increased the voxel-wise dose rate in proton therapy. In addition, the dose rate distribution is no longer uniform along the cross section of the target but rather highly heterogeneous, following the spot placement. We suggest investigating dose rate as potential contributor to a more complex proton RBE model. Approach. Due to the time structure of the PBS beam delivery the instantaneous dose rate is highly variable voxel by voxel. Several possible parameters to represent voxel-wise dose rate for a given clinical PBS treatment plan are detailed. These quantities were implemented in the scripting environment of our treatment planning system, and computations experimentally verified. Sample applications to treated patient plans are shown. Main Results. Computed dose rates we experimentally confirmed. Dose rate maps vary depending on which method is used to represent them. Mainly, the underlying time and dose intervals chosen determine the topography of the resultant distributions. The maximum dose rates experienced by any target voxel in a given PBS treatment plan in our system range from ~100 to ~450 Gy(RBE)/min, a factor of 10 - 100 increase compared to PS. These dose rate distributions are very heterogeneous, with distinct hot spots. Significance. Voxel-wise dose rates for current clinical PBS treatment plans vary greatly from clinically established practice with PS. The exploration of different dose rate measures to evaluate potential correlations with observed clinical outcomes is suggested, potentially adding a missing component in the understanding of proton RBE.

Modernization of safety environment for a dedicated beamline for proton ocular therapy.

BACKGROUND: Proton therapy is an effective treatment for ocular melanoma, and other tumors of the eye. The fixed horizontal beamline dedicated to ocular treatments at Massachusetts General Hospital was originally commissioned in 2002, with much of the equipment, safety features, and practices dating back to an earlier implementation at Harvard Cyclotron in the 1970s. PURPOSE: To describe the experience of reevaluation and enhancement of the safety environment for one of the longest continuously operating proton therapy programs. METHODS: Several enhancements in quality control had been introduced throughout the years of operation, as described in this manuscript, to better align the practice with the evolving standards of proton therapy and the demands of a modern hospital. We spotlight the design and results of the failure mode and effect analysis (FMEA), and subsequent actions introduced to mitigate the modes associated with elevated risk. The findings of the FMEA informed the specifications for the new software application, which facilitated the improved management of the treatment workflow and the image-guidance aspects of ocular treatments. RESULTS: Eleven failure modes identified as having the highest risk are described. Six of these were mitigated with the clinical roll-out of a new application for image-guided radiation therapy (IGRT). Others were addressed through task automation, the broader introduction of checklists, and enhancements in pre-treatment staff-led time-out. CONCLUSIONS: Throughout the task of modernizing the safety system of our dedicated ocular beamline, FMEA proved to be an effective instrument in soliciting inputs from the staff about safety and workflow concerns, helping to identify steps associated with elevated failure risks. Risks were reduced with the clinical introduction of a new IGRT application, which integrates quality management tools widely recognized for their role in risk mitigation: automation of the data transfer and workflow steps, and with the introduction of checklists and redundancy cross-checks.

Large anatomical changes in head-and-neck cancers - A dosimetric comparison of online and offline adaptive proton therapy.

Purpose: This work evaluates an online adaptive (OA) workflow for head-and-neck (H&N) intensity-modulated proton therapy (IMPT) and compares it with full offline replanning (FOR) in patients with large anatomical changes. Methods: IMPT treatment plans are created retrospectively for a cohort of eight H&N cancer patients that previously required replanning during the course of treatment due to large anatomical changes. Daily cone-beam CTs (CBCT) are acquired and corrected for scatter, resulting in 253 analyzed fractions. To simulate the FOR workflow, nominal plans are created on the planning-CT and delivered until a repeated-CT is acquired; at this point, a new plan is created on the repeated-CT. To simulate the OA workflow, nominal plans are created on the planning-CT and adapted at each fraction using a simple beamlet weight-tuning technique. Dose distributions are calculated on the CBCTs with Monte Carlo for both delivery methods. The total treatment dose is accumulated on the planning-CT. Results: Daily OA improved target coverage compared to FOR despite using smaller target margins. In the high-risk CTV, the median D98 degradation was 1.1 % and 2.1 % for OA and FOR, respectively. In the low-risk CTV, the same metrics yield 1.3 % and 5.2 % for OA and FOR, respectively. Smaller setup margins of OA reduced the dose to all OARs, which was most relevant for the parotid glands. Conclusion: Daily OA can maintain prescription doses and constraints over the course of fractionated treatment, even in cases of large anatomical changes, reducing the necessity for manual replanning in H&N IMPT.

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.

Comparison of weekly and daily online adaptation for head and neck intensity-modulated proton therapy.

The high conformality of intensity-modulated proton therapy (IMPT) dose distributions causes treatment plans to be sensitive to geometrical changes during the course of a fractionated treatment. This can be addressed using adaptive proton therapy (APT). One important question in APT is the frequency of adaptations performed during a fractionated treatment, which is related to the question whether plan adaptation has to be done online or offline. The purpose of this work is to investigate the impact of weekly and daily online IMPT plan adaptation on the treatment quality for head and neck patients. A cohort of ten head and neck patients with daily acquired cone-beam CT (CBCT) images was evaluated retrospectively. Dose tracking of the IMPT treatment was performed for three scenarios: base plan with no adaptation (BP), weekly online adaptation (OAW), and daily online adaptation (OAD). Both adaptation schemes used an in-house developed online APT workflow, performing Monte Carlo dose calculations on scatter-corrected CBCTs. IMPT plan adaptation was achieved by only tuning the weights of a subset of beamlets, based on deformable image registration from the planning CT to each CBCT. Although OADmitigated random delivery errors more effectively than OAWon a fraction per fraction basis, both OAWand OADachieved the clinical goals for all ten patients, while BP failed for six cases. In the high-risk CTV, accumulated values ofD98%ranged between 97.15% and 99.73% of the prescription dose for OAD, with a median of 98.07%. For OAW, values between 95.02% and 99.26% were obtained, with a median of 97.61% of the prescription dose. Otherwise, the dose to most organs at risk was similar for all three scenarios. Globally, our results suggest that OAWcould be used as an alternative approach to OADfor most patients in order to reduce the clinical workload.

A full-scale clinical prototype for proton range verification using prompt gamma-ray spectroscopy.

We present a full-scale clinical prototype system for in vivo range verification of proton pencil-beams using the prompt gamma-ray spectroscopy method. The detection system consists of eight LaBr3 scintillators and a tungsten collimator, mounted on a rotating frame. Custom electronics and calibration algorithms have been developed for the measurement of energy- and time-resolved gamma-ray spectra during proton irradiation at a clinical dose rate. Using experimentally determined nuclear reaction cross sections and a GPU-accelerated Monte Carlo simulation, a detailed model of the expected gamma-ray emissions is created for each individual pencil-beam. The absolute range of the proton pencil-beams is determined by minimizing the discrepancy between the measurement and this model, leaving the absolute range of the beam and the elemental concentrations of the irradiated matter as free parameters. The system was characterized in a clinical-like situation by irradiating different phantoms with a scanning pencil-beam. A dose of 0.9 Gy was delivered to a [Formula: see text] cm3 target with a beam current of 2 nA incident on the phantom. Different range shifters and materials were used to test the robustness of the verification method and to calculate the accuracy of the detected range. The absolute proton range was determined for each spot of the distal energy layer with a mean statistical precision of 1.1 mm at a 95% confidence level and a mean systematic deviation of 0.5 mm, when aggregating pencil-beam spots within a cylindrical region of 10 mm radius and 10 mm depth. Small range errors that we introduced were successfully detected and even large differences in the elemental composition do not affect the range verification accuracy. These results show that our system is suitable for range verification during patient treatments in our upcoming clinical study.

Range verification of passively scattered proton beams using prompt gamma-ray detection.

We performed an experimental study to verify the range of passively scattered proton beams by detecting prompt gamma-rays emitted from proton-nuclear interactions. A method is proposed using a single scintillation detector positioned near the distal end of the irradiated target. Lead shielding was used to attenuate gamma-rays emitted along most of the entrance path of the beam. By synchronizing the prompt gamma-ray detector to the rotation of the range modulation wheel, the relation between the gamma emission from the distal part of the target and the range of the incident proton beam was determined. In experiments with a water phantom and an anthropomorphic head phantom, this relation was found to be sensitive to range shifts that were introduced. The wide opening angle of the detector enabled a sufficient signal-to-background ratio to be achieved in the presence of neutron-induced background from the scattering and collimating devices. Uniform range shifts were detected with a standard deviation of 0.1 mm to 0.2 mm at a dose level of 30 cGy to 50 cGy (RBE). The detectable magnitude of a range shift limited to a part of the treatment field area was approximately proportional to the ratio between the field area and the area affected by the range shift. We conclude that it is feasible to detect changes in the range of passively scattered proton beams using a relatively simple prompt gamma-ray detection system. The method can be employed for in vivo verification of the consistency of the delivered range in fractionated treatments.

Proton range verification through prompt gamma-ray spectroscopy.

We present an experimental study of a novel method to verify the range of proton therapy beams. Differential cross sections were measured for 15 prompt gamma-ray lines from proton-nuclear interactions with (12)C and (16)O at proton energies up to 150 MeV. These cross sections were used to model discrete prompt gamma-ray emissions along proton pencil-beams. By fitting detected prompt gamma-ray counts to these models, we simultaneously determined the beam range and the oxygen and carbon concentration of the irradiated matter. The performance of the method was assessed in two phantoms with different elemental concentrations, using a small scale prototype detector. Based on five pencil-beams with different ranges delivering 5 × 10(8) protons and without prior knowledge of the elemental composition at the measurement point, the absolute range was determined with a standard deviation of 1.0-1.4 mm. Relative range shifts at the same dose level were detected with a standard deviation of 0.3-0.5 mm. The determined oxygen and carbon concentrations also agreed well with the actual values. These results show that quantitative prompt gamma-ray measurements enable knowledge of nuclear reaction cross sections to be used for precise proton range verification in the presence of tissue with an unknown composition.

Range verification of passively scattered proton beams based on prompt gamma time patterns.

We propose a proton range verification technique for passive scattering proton therapy systems where spread out Bragg peak (SOBP) fields are produced with rotating range modulator wheels. The technique is based on the correlation of time patterns of the prompt gamma ray emission with the range of protons delivering the SOBP. The main feature of the technique is the ability to verify the proton range with a single point of measurement and a simple detector configuration. We performed four-dimensional (time-dependent) Monte Carlo simulations using TOPAS to show the validity and accuracy of the technique. First, we validated the hadronic models used in TOPAS by comparing simulations and prompt gamma spectrometry measurements published in the literature. Second, prompt gamma simulations for proton range verification were performed for the case of a water phantom and a prostate cancer patient. In the water phantom, the proton range was determined with 2 mm accuracy with a full ring detector configuration for a dose of ~2.5 cGy. For the prostate cancer patient, 4 mm accuracy on range determination was achieved for a dose of ~15 cGy. The results presented in this paper are encouraging in view of a potential clinical application of the technique.

Proton radiography and proton computed tomography based on time-resolved dose measurements.

We present a proof of principle study of proton radiography and proton computed tomography (pCT) based on time-resolved dose measurements. We used a prototype, two-dimensional, diode-array detector capable of fast dose rate measurements, to acquire proton radiographic images expressed directly in water equivalent path length (WEPL). The technique is based on the time dependence of the dose distribution delivered by a proton beam traversing a range modulator wheel in passive scattering proton therapy systems. The dose rate produced in the medium by such a system is periodic and has a unique pattern in time at each point along the beam path and thus encodes the WEPL. By measuring the time dose pattern at the point of interest, the WEPL to this point can be decoded. If one measures the time­dose patterns at points on a plane behind the patient for a beam with sufficient energy to penetrate the patient, the obtained 2D distribution of the WEPL forms an image. The technique requires only a 2D dosimeter array and it uses only the clinical beam for a fraction of second with negligible dose to patient. We first evaluated the accuracy of the technique in determining the WEPL for static phantoms aiming at beam range verification of the brain fields of medulloblastoma patients. Accurate beam ranges for these fields can significantly reduce the dose to the cranial skin of the patient and thus the risk of permanent alopecia. Second, we investigated the potential features of the technique for real-time imaging of a moving phantom. Real-time tumor tracking by proton radiography could provide more accurate validations of tumor motion models due to the more sensitive dependence of proton beam on tissue density compared to x-rays. Our radiographic technique is rapid (~100 ms) and simultaneous over the whole field, it can image mobile tumors without the problem of interplay effect inherently challenging for methods based on pencil beams. Third, we present the reconstructed pCT images of a cylindrical phantom containing inserts of different materials. As for all conventional pCT systems, the method illustrated in this work produces tomographic images that are potentially more accurate than x-ray CT in providing maps of proton relative stopping power (RSP) in the patient without the need for converting x-ray Hounsfield units to proton RSP. All phantom tests produced reasonable results, given the currently limited spatial and time resolution of the prototype detector. The dose required to produce one radiographic image, with the current settings, is ~0.7 cGy. Finally, we discuss a series of techniques to improve the resolution and accuracy of radiographic and tomographic images for the future development of a full-scale detector.

Energy- and time-resolved detection of prompt gamma-rays for proton range verification.

In this work, we present experimental results of a novel prompt gamma-ray detector for proton beam range verification. The detection system features an actively shielded cerium-doped lanthanum(III) bromide scintillator, coupled to a digital data acquisition system. The acquisition was synchronized to the cyclotron radio frequency to separate the prompt gamma-ray signals from the later-arriving neutron-induced background. We designed the detector to provide a high energy resolution and an effective reduction of background events, enabling discrete proton-induced prompt gamma lines to be resolved. Measuring discrete prompt gamma lines has several benefits for range verification. As the discrete energies correspond to specific nuclear transitions, the magnitudes of the different gamma lines have unique correlations with the proton energy and can be directly related to nuclear reaction cross sections. The quantification of discrete gamma lines also enables elemental analysis of tissue in the beam path, providing a better prediction of prompt gamma-ray yields. We present the results of experiments in which a water phantom was irradiated with proton pencil-beams in a clinical proton therapy gantry. A slit collimator was used to collimate the prompt gamma-rays, and measurements were performed at 27 positions along the path of proton beams with ranges of 9, 16 and 23 g cm(-2) in water. The magnitudes of discrete gamma lines at 4.44, 5.2 and 6.13 MeV were quantified. The prompt gamma lines were found to be clearly resolved in dimensions of energy and time, and had a reproducible correlation with the proton depth-dose curve. We conclude that the measurement of discrete prompt gamma-rays for in vivo range verification of clinical proton beams is feasible, and plan to further study methods and detector designs for clinical use.

Dosimetric accuracy of proton therapy for chordoma patients with titanium implants.

PURPOSE: To investigate dosimetric errors in proton therapy treatment planning due to titanium implants, and to determine how these affect postoperative passively scattered proton therapy for chordoma patients with orthopedic hardware. METHODS: The presence of titanium hardware near the tumor may affect the dosimetric accuracy of proton therapy. Artifacts in the computed tomography (CT) scan can cause errors in the proton stopping powers used for dose calculation, which are derived from CT numbers. Also, clinical dose calculation algorithms may not accurately simulate proton beam transport through the implants, which have very different properties as compared to human tissue. The authors first evaluated the impact of these two main issues. Dose errors introduced by metal artifacts were studied using phantoms with and without titanium inserts, and patient scans on which a metal artifact reduction method was applied. Pencil-beam dose calculations were compared to models of nuclear interactions in titanium and Monte Carlo simulations. Then, to assess the overall impact on treatment plans for chordoma, the authors compared the original clinical treatment plans to recalculated dose distributions employing both metal artifact reduction and Monte Carlo methods. RESULTS: Dose recalculations of clinical proton fields showed that metal artifacts cause range errors up to 6 mm distal to regions affected by CT artifacts. Monte Carlo simulations revealed dose differences >10% in the high-dose area, and range differences up to 10 mm. Since these errors are mostly local in nature, the large number of fields limits the impact on target coverage in the chordoma treatment plans to a small decrease of dose homogeneity. CONCLUSIONS: In the presence of titanium implants, CT metal artifacts and the approximations of pencil-beam dose calculations cause considerable errors in proton dose calculation. The spatial distribution of the errors however limits the overall impact on passively scattered proton therapy for chordoma.

Simulation of prompt gamma-ray emission during proton radiotherapy.

The measurement of prompt gamma rays emitted from proton-induced nuclear reactions has been proposed as a method to verify in vivo the range of a clinical proton radiotherapy beam. A good understanding of the prompt gamma-ray emission during proton therapy is key to develop a clinically feasible technique, as it can facilitate accurate simulations and uncertainty analysis of gamma detector designs. Also, the gamma production cross-sections may be incorporated as prior knowledge in the reconstruction of the proton range from the measurements. In this work, we performed simulations of proton-induced nuclear reactions with the main elements of human tissue, carbon-12, oxygen-16 and nitrogen-14, using the nuclear reaction models of the GEANT4 and MCNP6 Monte Carlo codes and the dedicated nuclear reaction codes TALYS and EMPIRE. For each code, we made an effort to optimize the input parameters and model selection. The results of the models were compared to available experimental data of discrete gamma line cross-sections. Overall, the dedicated nuclear reaction codes reproduced the experimental data more consistently, while the Monte Carlo codes showed larger discrepancies for a number of gamma lines. The model differences lead to a variation of the total gamma production near the end of the proton range by a factor of about 2. These results indicate a need for additional theoretical and experimental study of proton-induced gamma emission in human tissue.

CT metal artifact reduction method correcting for beam hardening and missing projections.

We present and validate a computed tomography (CT) metal artifact reduction method that is effective for a wide spectrum of clinical implant materials. Projections through low-Z implants such as titanium were corrected using a novel physics correction algorithm that reduces beam hardening errors. In the case of high-Z implants (dental fillings, gold, platinum), projections through the implant were considered missing and regularized iterative reconstruction was performed. Both algorithms were combined if multiple implant materials were present. For comparison, a conventional projection interpolation method was implemented. In a blinded and randomized evaluation, ten radiation oncologists ranked the quality of patient scans on which the different methods were applied. For scans that included low-Z implants, the proposed method was ranked as the best method in 90% of the reviews. It was ranked superior to the original reconstruction (p = 0.0008), conventional projection interpolation (p < 0.0001) and regularized limited data reconstruction (p = 0.0002). All reviewers ranked the method first for scans with high-Z implants, and better as compared to the original reconstruction (p < 0.0001) and projection interpolation (p = 0.004). We conclude that effective reduction of CT metal artifacts can be achieved by combining algorithms tailored to specific types of implant materials.