Photon beam relative dose validation of the DPM Monte Carlo code in lung-equivalent media

NeuralDAO: Incorporating neural network generated dose into direct aperture optimization for end-to-end IMRT planning

PURPOSE

Thecurrent practice in intensity-modulated radiation therapy (IMRT) planning almost always includes different dose calculation strategies for plan optimization and final dose verification. The accurate Monte Carlo (MC) dose algorithm is considered to be time-consuming for the optimization. Thus a fast, simplified dose algorithm is used in the optimization. The significant differences between the optimized dose and the delivered dose lead to tediously planning loops and potentially suboptimal solutions. This work aims to develop an IMRT optimization algorithm to minimize the dose discrepancy so that the delivered dose can be optimized in a holistic, end-to-end manner.

METHODS

The proposed algorithm, namely NeuralDAO, integrates a neural dose network into the column generation (CG) direct aperture optimization (DAO) formulation for step-and-shoot IMRT planning. The neural dose network is designed and trained to produce doses of MC-level accuracy within few milliseconds. Its differentiability is fully exploited to compute gradients for identifying potential aperture shapes. A prototype of NeuralDAO was developed in PyTorch and available to the public. Five lung patient cases have been studied. Dosimetric accuracy was compared with the MC dose. Plan quality and time were compared with a state-of-the-art (SoA) dose-correct algorithm. Statistical analysis was performed by Wilcoxon signed-rank test.

RESULTS

The average gamma passing rate at 2 mm/2% is 99.7% between the optimized and delivered doses. The convergence process produced by NeuralDAO is virtually identical to that produced by an MC-based DAO. The average dose calculation time is 12.1 ms for an aperture on GPU. One session of optimization took 10-36 min. Compared with the SoA, better conformity index and homogeneity index were observed for the target. The esophagus was significantly spared. Significant reductions were observed for the replanning number and the planning time.

CONCLUSIONS

A new DAO algorithm based on the neural dose network has been developed. The results suggest that this algorithm minimizes the discrepancy between the optimized and delivered doses, which offers a promising approach to reduce the time and effort required in IMRT planning. This work demonstrates the possibility of applying the neural network in IMRT optimization. It is of great potential to extend this algorithm to other treatment modalities.

Correlation of normal lung density changes with dose after stereotactic body radiotherapy (SBRT) for early stage lung cancer

BACKGROUND AND PURPOSE

To investigate the correlation between normal lung CT density changes with dose accuracy and outcome after stereotactic body radiation therapy (SBRT) for patients with early stage non-small-cell lung cancer (NSCLC).

MATERIALS AND METHODS

Thirty-one patients (with a total of 33 lesions) with non-small cell lung cancer were selected out of 270 patients treated with SBRT at a single institution between 2003 and 2009. Out of these 31 patients, 10 patients had developed radiation pneumonitis (RP). Dose distributions originally planned using a 1-D pencil beam-based dose algorithm were retrospectively recomputed using different algorithms. Prescription dose was 48 Gy in 4 fractions in most patients. Planning CT images were rigidly registered to follow-up CT datasets at 3-9 months after treatment. Corresponding dose distributions were mapped from planning to follow-up CT images. Hounsfield Unit (HU) changes in lung density in individual, 5 Gy, dose bins from 5 to 45 Gy were assessed in the peri-tumoral region. Correlations between HU changes in various normal lung regions, dose indices (V20, MLD, generalized equivalent uniform dose (gEUD)), and RP grade were investigated.

RESULTS

Strong positive correlation was found between HU changes in the peri-tumoral region and RP grade (Spearman's r = 0.760; p < 0.001). Positive correlation was also observed between RP and HU changes in the region covered by V20 for all algorithms (Spearman's r ≥ 0.738; p < 0.001). Additionally, V20, MLD, and gEUD were significantly correlated with RP grade (p < 0.01). MLD in the peri-tumoral region computed with model-based algorithms was 5-7% lower than the PB-based methods.

CONCLUSION

Changes of lung density in the peri-tumoral lung and in the region covered by V20 were strongly associated with RP grade. Relative to model-based methods, PB algorithms over-estimated mean peri-tumoral dose and showed displacement of the high-dose region, which correlated with HU changes on follow-up CT scans.

Beam modeling and beam model commissioning for Monte Carlo dose calculation-based radiation therapy treatment planning: Report of AAPM Task Group 157

Dose calculation plays an important role in the accuracy of radiotherapy treatment planning and beam delivery. The Monte Carlo (MC) method is capable of achieving the highest accuracy in radiotherapy dose calculation and has been implemented in many commercial systems for radiotherapy treatment planning. The objective of this task group was to assist clinical physicists with the potentially complex task of acceptance testing and commissioning MC-based treatment planning systems (TPS) for photon and electron beam dose calculations. This report provides an overview on the general approach of clinical implementation and testing of MC-based TPS with a specific focus on models of clinical photon and electron beams. Different types of beam models are described including those that utilize MC simulation of the treatment head and those that rely on analytical methods and measurements. The trade-off between accuracy and efficiency in the various source-modeling approaches is discussed together with guidelines for acceptance testing of MC-based TPS from the clinical standpoint. Specific recommendations are given on methods and practical procedures to commission clinical beam models for MC-based TPS.

DPM as a radiation transport engine for PRIMO

Background

PRIMO is a dose verification system based on the general-purpose Monte Carlo radiation transport code penelope, which implements an accurate physics model of the interaction cross sections and the radiation transport process but with low computational efficiency as compared with fast Monte Carlo codes. One of these fast Monte Carlo codes is the Dose Planning Method (DPM). The purpose of this work is to describe the adaptation of DPM as an alternative PRIMO computation engine, to validate its performance against penelope and to validate it for some specific cases.

Methods

DPM was parallelized and modified to perform radiation transport in quadric geometries, which are used to describe linacs, thus allowing the simulation of dynamic treatments. To benchmark the new code versus penelope, both in terms of accuracy of results and simulation time, several tests were performed, namely, irradiation of a multi-layer phantom, irradiation of a water phantom using a collimating pattern defined by the multileaf collimator (MLC), and four clinical cases. The gamma index, with passing criteria of 1 mm/1%, was used to compare the absorbed dose distributions. Clinical cases were compared using a 3-D gamma analysis.

Results

The percentage of voxels passing the gamma criteria always exceeded 99% for the phantom cases, with the exception of the transport through air, for which dose differences between DPM and penelope were as large as 24%. The corresponding percentage for the clinical cases was larger than 99%. The speedup factor between DPM and penelope ranged from 2.5 ×, for the simulation of the radiation transport through a MLC and the subsequent dose estimation in a water phantom, up to 11.8 × for a lung treatment. A further increase of the computational speed, up to 25 ×, can be obtained in the clinical cases when a voxel size of (2.5 mm)³ is used.

Conclusions

DPM has been incorporated as an efficient and accurate Monte Carlo engine for dose estimation in PRIMO. It allows the concatenated simulation of the patient-dependent part of the linac and the patient geometry in static and dynamic treatments. The discrepancy observed between DPM and penelope, which is due to an artifact of the cross section interpolation algorithm for low energy electrons in air, does not affect the results in other materials.

Novel Monte Carlo dose calculation algorithm for robotic radiosurgery with multi leaf collimator: Dosimetric evaluation

PURPOSE

At introduction in 2014, dose calculation for the first MLC on a robotic SRS/SBRT platform was limited to a correction-based Finite-Size Pencil Beam (FSPB) algorithm. We report on the dosimetric accuracy of a novel Monte Carlo (MC) dose calculation algorithm for this MLC, included in the Precision™ treatment planning system.

METHODS

A phantom was built of one slab (5.0 cm) of lung-equivalent material (0.09…0.29 g/cc) enclosed by 3.5 cm (above) and 5 cm (below) slabs of solid water (1.045 g/cc). This was irradiated using rectangular (15.4 × 15.4 mm² to 53.8 × 53.7 mm²) and two irregular MLC-fields. Radiochromic film (EBT3) was positioned perpendicular to the slabs and parallel to the beam. Calculated dose distributions were compared to film measurements using line scans and 2D gamma analysis.

RESULTS

Measured and MC calculated percent depth dose curves showed a characteristic dose drop within the low-density region, which was not correctly reproduced by FSPB. Superior average gamma pass rates (2%/1 mm) were found for MC (91.2 ± 1.5%) compared to FSPB (55.4 ± 2.7%). However, MC calculations exhibited localized anomalies at mass density transitions around 0.15 g/cc, which were traced to a simplification in electron transport. Absence of these anomalies was confirmed in a modified build of the MC engine, which increased gamma pass rates to 96.6 ± 1.2%.

CONCLUSIONS

The novel MC algorithm greatly improves dosimetric accuracy in heterogeneous tissue, potentially expanding the clinical use of robotic radiosurgery with MLC. In-depth, independent validation is paramount to identify and reduce the residual uncertainties in any software solution.

Evaluation of AAA and XVMC Algorithms for Dose Calculation in Lung Equivalent Heterogeneity in Photon Fields: A Comparison of Calculated Results with Measurements

AIMS

The aims of the present work are (1) to evaluate dose calculation accuracy of two commonly used algorithms for 15 MV small photon fields in a medium encompassing heterogeneity and (2) to compare them with measured results obtained from gafchromic film EBT2.

MATERIALS AND METHODS

Authors employed kailwood (Pinus Wallichiana) to mimic lung. Briefly, seven Kailwood plates, each measuring 25x25 cm² of varying thicknesses totaling 13 cm equivalent to the mean thickness of an adult human lung, were sandwiched between 5 cm tissue equivalent material from top and 10 cm below. Physical measurements were performed using Radiochromic film EBT2. The field sizes of 1x1, 2x2, 5x5 and 10x10 cm² were selected at 100 cm SSD. Simulations were performed using EGSnrc/DOSRZnrc Monte Carlo code. The dose variation inside the inhomogeneity and near the interface was calculated using AAA & XVMC algorithm.

RESULTS

Preliminary results show that there is large variation of dose inside inhomogeneity. The maximum variation of dose inside the inhomogeneity for 1x1 cm² was found 40% by AAA and 4.5% by XVMC compared to measured/simulated results. For the field size of 2x2 cm², these figures were 27% by AAA & 3.5% by XVMC. For 5x5 cm² field size, the variation is small which becomes insignificant for larger fields.

CONCLUSION

The results presented in this work indicate that for smaller fields, XVMC algorithm gives more realistic prediction, while there is the need for caution on using AAA algorithm for dose calculations involving small area irradiation encompassing heterogeneities and low-density media.

Motion management strategies and technical issues associated with stereotactic body radiotherapy of thoracic and upper abdominal tumors: A review from NRG oncology

The efficacy of stereotactic body radiotherapy (SBRT) has been well demonstrated. However, it presents unique challenges for accurate planning and delivery especially in the lungs and upper abdomen where respiratory motion can be significantly confounding accurate targeting and avoidance of normal tissues. In this paper, we review the current literature on SBRT for lung and upper abdominal tumors with particular emphasis on addressing respiratory motion and its affects. We provide recommendations on strategies to manage motion for different, patient-specific situations. Some of the recommendations will potentially be adopted to guide clinical trial protocols.

Impact of inhomogeneity corrections on dose coverage in the treatment of lung cancer using stereotactic body radiation therapy

The purpose of this study is to assess the real target dose coverage when radiation treatments were delivered to lung cancer patients based on treatment planning according to the RTOG-0236 Protocol. We compare calculated dosimetric results between the more accurate anisotropic analytical algorithm (AAA) and the pencil beam algorithm for stereotactic body radiation therapy treatment planning in lung cancer. Ten patients with non-small cell lung cancer were given 60 Gy in three fractions using 6 and 10 MV beams with 8-10 fields. The patients were chosen in accordance with the lung RTOG-0236 protocol. The dose calculations were performed using the pencil beam algorithm with no heterogeneity corrections (PB-NC) and then recalculated with the pencil beam with modified Batho heterogeneity corrections (PB-MB) and the AAA using an identical beam setup and monitor units. The differences in calculated dose to 95% or 99% of the PTV, between using the PB-NC and the AAA, were within 10% of prescribed dose (60 Gy). However, the minimum dose to 95% and 99% of PTV calculated using the PB-MB were consistently overestimated by up to 40% and 36% of the prescribed dose, respectively, compared to that calculated by the AAA. Using the AAA as reference, the calculated maximum doses were underestimated by up to 27% using the PB-NC and overestimated by 19% using the PB-MB. The calculations of dose to lung from PB-NC generally agree with that of AAA except in the small high-dose region where PB-NC underestimates. The calculated dose distributions near the interface using the AAA agree with those from Monte Carlo calculations as well as measured values. This study indicates that the real minimum PTV dose coverage cannot be guaranteed when the PB-NC is used to calculate the monitor unit settings in dose prescriptions.

Fast CPU-based Monte Carlo simulation for radiotherapy dose calculation

Monte-Carlo (MC) simulations are considered to be the most accurate method for calculating dose distributions in radiotherapy. Its clinical application, however, still is limited by the long runtimes conventional implementations of MC algorithms require to deliver sufficiently accurate results on high resolution imaging data. In order to overcome this obstacle we developed the software-package PhiMC, which is capable of computing precise dose distributions in a sub-minute time-frame by leveraging the potential of modern many- and multi-core CPU-based computers. PhiMC is based on the well verified dose planning method (DPM). We could demonstrate that PhiMC delivers dose distributions which are in excellent agreement to DPM. The multi-core implementation of PhiMC scales well between different computer architectures and achieves a speed-up of up to 37[Formula: see text] compared to the original DPM code executed on a modern system. Furthermore, we could show that our CPU-based implementation on a modern workstation is between 1.25[Formula: see text] and 1.95[Formula: see text] faster than a well-known GPU implementation of the same simulation method on a NVIDIA Tesla C2050. Since CPUs work on several hundreds of GB RAM the typical GPU memory limitation does not apply for our implementation and high resolution clinical plans can be calculated.

Practical methods for improving dose distributions in Monte Carlo-based IMRT planning of lung wall-seated tumors treated with SBRT

Current commercially available planning systems with Monte Carlo (MC)‐based final dose calculation in IMRT planning employ pencil‐beam (PB) algorithms in the optimization process. Consequently, dose coverage for SBRT lung plans can feature cold‐spots at the interface between lung and tumor tissue. For lung wall (LW)‐seated tumors, there can also be hot spots within nearby normal organs (example: ribs). This study evaluated two different practical approaches to limiting cold spots within the target and reducing high doses to surrounding normal organs in MC‐based IMRT planning of LW‐seated tumors. First, “iterative reoptimization”, where the MC calculation (with PB‐based optimization) is initially performed. The resultant cold spot is then contoured and used as a simultaneous boost volume. The MC‐based dose is then recomputed. The second technique uses noncoplanar beam angles with limited path through lung tissue. Both techniques were evaluated against a conventional coplanar beam approach with a single MC calculation. In all techniques the prescription dose was normalized to cover 95% of the PTV. Fifteen SBRT lung cases with LW‐seated tumors were planned. The results from iterative reoptimization showed that conformity index (CI) and/or PTV dose uniformity (UPTV) improved in 12/15 plans. Average improvement was 13%, and 24%, respectively. Nonimproved plans had PTVs near the skin, trachea, and/or very small lung involvement. The maximum dose to 1cc volume (D1cc) of surrounding OARs decreased in 14/15 plans (average 10%). Using noncoplanar beams showed an average improvement of 7% in 10/15 cases and 11% in 5/15 cases for CI and UPTV, respectively. The D1cc was reduced by an average of 6% in 10/15 cases to surrounding OARs. Choice of treatment planning technique did not statistically significantly change lung V5. The results showed that the proposed practical approaches enhance dose conformity in MC‐based IMRT planning of lung tumors treated with SBRT, improving target dose coverage and potentially reducing toxicities to surrounding normal organs.

PACS numbers: 87.55.de, 87.55.kh

On the quantification of the dosimetric accuracy of collapsed cone convolution superposition (CCCS) algorithm for small lung volumes using IMRT

Specialized techniques that make use of small field dosimetry are common practice in today's clinics. These new techniques represent a big challenge to the treatment planning systems due to the lack of lateral electronic equilibrium. Because of this, the necessity of planning systems to overcome such difficulties and provide an accurate representation of the true value is of significant importance. Pinnacle3 is one such planning system. During the IMRT optimization process, Pinnacle3 treatment planning system allows the user to specify a minimum segment size which results in multiple beams composed of several subsets of different widths. In this study, the accuracy of the engine dose calculation, collapsed cone convolution superposition algorithm (CCCS) used by Pinnacle3, was quantified by Monte Carlo simulations, ionization chamber, and Kodak extended dose range film (EDR2) measurements for 11 SBRT lung patients. Lesions were < 3.0 cm in maximal diameter and <27.0cm3 in volume. The Monte Carlo EGSnrc\BEAMnrc and EGS4\MCSIM were used in the comparison. The minimum segment size allowable during optimization had a direct impact on the number of monitor units calculated for each beam. Plans with the smallest minimum segment size (0.1 cm2 to 2.0 cm2) had the largest number of MUs. Although PTV coverage remained unaffected, the segment size did have an effect on the dose to the organs at risk. Pinnacle3-calculated PTV mean doses were in agreement with Monte Carlo-calculated mean doses to within 5.6% for all plans. On average, the mean dose difference between Monte Carlo and Pinnacle3 for all 88 plans was 1.38%. The largest discrepancy in maximum dose was 5.8%, and was noted for one of the plans using a minimum segment size of 1.0 cm2. For minimum dose to the PTV, a maximum discrepancy between Monte Carlo and Pinnacle3 was noted of 12.5% for a plan using a 6.0 cm2 minimum segment size. Agreement between point dose measurements and Pinnacle3-calculated doses were on average within 0.7% in both phantoms. The profiles show a good agreement between Pinnacle3, Monte Carlo, and EDR2 film. The gamma index and the isodose lines support the result.

GPU-based fast Monte Carlo simulation for radiotherapy dose calculation

Monte Carlo (MC) simulation is commonly considered to be the most accurate dose calculation method in radiotherapy. However, its efficiency still requires improvement for many routine clinical applications. In this paper, we present our recent progress toward the development of a graphics processing unit (GPU)-based MC dose calculation package, gDPM v2.0. It utilizes the parallel computation ability of a GPU to achieve high efficiency, while maintaining the same particle transport physics as in the original dose planning method (DPM) code and hence the same level of simulation accuracy. In GPU computing, divergence of execution paths between threads can considerably reduce the efficiency. Since photons and electrons undergo different physics and hence attain different execution paths, we use a simulation scheme where photon transport and electron transport are separated to partially relieve the thread divergence issue. A high-performance random number generator and a hardware linear interpolation are also utilized. We have also developed various components to handle the fluence map and linac geometry, so that gDPM can be used to compute dose distributions for realistic IMRT or VMAT treatment plans. Our gDPM package is tested for its accuracy and efficiency in both phantoms and realistic patient cases. In all cases, the average relative uncertainties are less than 1%. A statistical t-test is performed and the dose difference between the CPU and the GPU results is not found to be statistically significant in over 96% of the high dose region and over 97% of the entire region. Speed-up factors of 69.1 ∼ 87.2 have been observed using an NVIDIA Tesla C2050 GPU card against a 2.27 GHz Intel Xeon CPU processor. For realistic IMRT and VMAT plans, MC dose calculation can be completed with less than 1% standard deviation in 36.1 ∼ 39.6 s using gDPM.

Verification measurements and clinical evaluation of the iPlan RT Monte Carlo dose algorithm for 6 MV photon energy

This study presents data for verification of the iPlan RT Monte Carlo (MC) dose algorithm (BrainLAB, Feldkirchen, Germany). MC calculations were compared with pencil beam (PB) calculations and verification measurements in phantoms with lung-equivalent material, air cavities or bone-equivalent material to mimic head and neck and thorax and in an Alderson anthropomorphic phantom. Dosimetric accuracy of MC for the micro-multileaf collimator (MLC) simulation was tested in a homogeneous phantom. All measurements were performed using an ionization chamber and Kodak EDR2 films with Novalis 6 MV photon beams. Dose distributions measured with film and calculated with MC in the homogeneous phantom are in excellent agreement for oval, C and squiggle-shaped fields and for a clinical IMRT plan. For a field with completely closed MLC, MC is much closer to the experimental result than the PB calculations. For fields larger than the dimensions of the inhomogeneities the MC calculations show excellent agreement (within 3%/1 mm) with the experimental data. MC calculations in the anthropomorphic phantom show good agreement with measurements for conformal beam plans and reasonable agreement for dynamic conformal arc and IMRT plans. For 6 head and neck and 15 lung patients a comparison of the MC plan with the PB plan was performed. Our results demonstrate that MC is able to accurately predict the dose in the presence of inhomogeneities typical for head and neck and thorax regions with reasonable calculation times (5-20 min). Lateral electron transport was well reproduced in MC calculations. We are planning to implement MC calculations for head and neck and lung cancer patients.

[Comparison of RTPS and Monte Carlo dose distributions in heterogeneous phantoms for photon beams]

The purpose of this study was to compare dose distributions from three different RTPS with those from Monte Carlo (MC) calculations and measurements, in heterogeneous phantoms for photon beams. This study used four algorithms for RTPS: AAA (analytical anisotropic algorithm) implemented in the Eclipse (Varian Medical Systems) treatment planning system, CC (collapsed cone) superposition from the Pinnacle (Philips), and MGS (multigrid superposition) and FFT (fast Fourier transform) convolution from XiO (CMS). The dose distributions from these algorithms were compared with those from MC and measurements in a set of heterogeneous phantoms. Eclipse/AAA underestimated the dose inside the lung region for low energies of 4 and 6 MV. This is because Eclipse/AAA do not adequately account for a scaling of the spread of the pencil (lateral electron transport) based on changes in the electron density at low photon energies. The dose distributions from Pinnacle/CC and XiO/MGS almost agree with those of MC and measurements at low photon energies, but increase errors at high energy of 15 MV, especially for a small field of 3x3 cm(2). The FFT convolution extremely overestimated the dose inside the lung slab compared to MC. The dose distributions from the superposition algorithms almost agree with those from MC as well as measured values at 4 and 6 MV. The dose errors for Eclipse/AAA are lager in lung model phantoms for 4 and 6 MV. It is necessary to use the algorithms comparable to superposition for accuracy of dose calculations in heterogeneous regions.

Comparison of dose calculation algorithms in slab phantoms with cortical bone equivalent heterogeneities

To evaluate the dose values predicted by several calculation algorithms in two treatment planning systems, Monte Carlo (MC) simulations and measurements by means of various detectors were performed in heterogeneous layer phantoms with water- and bone-equivalent materials. Percentage depth doses (PDDs) were measured with thermoluminescent dosimeters (TLDs), metal-oxide semiconductor field-effect transistors (MOSFETs), plane parallel and cylindrical ionization chambers, and beam profiles with films. The MC code used for the simulations was the PENELOPE code. Three different field sizes (10 x 10, 5 x 5, and 2 x 2 cm2) were studied in two phantom configurations and a bone equivalent material. These two phantom configurations contained heterogeneities of 5 and 2 cm of bone, respectively. We analyzed the performance of four correction-based algorithms and one based on convolution superposition. The correction-based algorithms were the Batho, the Modified Batho, the Equivalent TAR implemented in the Cadplan (Varian) treatment planning system (TPS), and the Helax-TMS Pencil Beam from the Helax-TMS (Nucletron) TPS. The convolution-superposition algorithm was the Collapsed Cone implemented in the Helax-TMS. All the correction-based calculation algorithms underestimated the dose inside the bone-equivalent material for 18 MV compared to MC simulations. The maximum underestimation, in terms of root-mean-square (RMS), was about 15% for the Helax-TMS Pencil Beam (Helax-TMS PB) for a 2 x 2 cm2 field inside the bone-equivalent material. In contrast, the Collapsed Cone algorithm yielded values around 3%. A more complex behavior was found for 6 MV where the Collapsed Cone performed less well, overestimating the dose inside the heterogeneity in 3%-5%. The rebuildup in the interface bone-water and the penumbra shrinking in high-density media were not predicted by any of the calculation algorithms except the Collapsed Cone, and only the MC simulations matched the experimental values within the estimated uncertainties. The TLD and MOSFET detectors were suitable for dose measurement inside bone-equivalent materials, while parallel ionization chambers, applying the same calibration and correction factors as in water, systematically underestimated dose by 3%-5%.

An anatomically realistic lung model for Monte Carlo-based dose calculations

Treatment planning for disease sites with large variations of electron density in neighboring tissues requires an accurate description of the geometry. This self-evident statement is especially true for the lung, a highly complex organ having structures with a wide range of sizes that range from about 10(-4) to 1 cm. In treatment planning, the lung is commonly modeled by a voxelized geometry obtained using computed tomography (CT) data at various resolutions. The simplest such model, which is often used for QA and validation work, is the atomic mix or mean density model, in which the entire lung is homogenized and given a mean (volume-averaged) density. The purpose of this paper is (i) to describe a new heterogeneous random lung model, which is based on morphological data of the human lung, and (ii) use this model to assess the differences in dose calculations between an actual lung (as represented by our model) and a mean density (homogenized) lung. Eventually, we plan to use the random lung model to assess the accuracy of CT-based treatment plans of the lung. For this paper, we have used Monte Carlo methods to make accurate comparisons between dose calculations for the random lung model and the mean density model. For four realizations of the random lung model, we used a single photon beam, with two different energies (6 and 18 MV) and four field sizes (1 x 1, 5 x 5, 10 x 10, and 20 x 20 cm2). We found a maximum difference of 34% of D(max) with the 1 x 1, 18 MV beam along the central axis (CAX). A "shadow" region distal to the lung, with dose reduction up to 7% of D(max), exists for the same realization. The dose perturbations decrease for larger field sizes, but the magnitude of the differences in the shadow region is nearly independent of the field size. We also observe that, compared to the mean density model, the random structures inside the heterogeneous lung can alter the shape of the isodose lines, leading to a broadening or shrinking of the penumbra region. For small field sizes, the mean lung doses significantly depend on the structures' relative locations to the beam. In addition to these comparisons between the random lung and mean density models, we also provide a preliminary comparison between dose calculations for the random lung model and a voxelized version of this model at 0.4 x 0.4 x 0.4 cm3 resolution. Overall, this study is relevant to treatment planning for lung tumors, especially in situations where small field sizes are used. Our results show that for such situations, the mean density model of the lung is inadequate, and a more accurate CT model of the lung is required. Future work with our model will involve patient motion, setup errors, and recommendations for the resolution of CT models.

The impact of breathing motion versus heterogeneity effects in lung cancer treatment planning

The purpose of this study is to investigate the effects of tissue heterogeneity and breathing-induced motion/deformation on conformal treatment planning for pulmonary tumors and to compare the magnitude and the clinical importance of changes induced by these effects. Treatment planning scans were acquired at normal exhale/inhale breathing states for fifteen patients. The internal target volume (ITV) was defined as the union of exhale and inhale gross tumor volumes uniformly expanded by 5 mm. Anterior/posterior opposed beams (AP/PA) and three-dimensional (3D)-conformal plans were designed using the unit-density exhale ("static") dataset. These plans were further used to calculate (a) density-corrected ("heterogeneous") static dose and (b) heterogeneous cumulative dose, including breathing deformations. The DPM Monte Carlo code was used for dose computations. For larger than coin-sized tumors, relative to unit-density plans, tumor and lung doses increased in the heterogeneity-corrected plans. In comparing cumulative and static plans, larger normal tissue complication probability changes were observed for tumors with larger motion amplitudes and uncompensated breathing-induced hot/cold spots in lung. Accounting for tissue heterogeneity resulted in average increases of 9% and 7% in mean lung dose (MLD) for the 6 MV and 15 MV photon beams, respectively. Breathing-induced effects resulted in approximately 1% and 2% average decreases in MLD from the static value, for the 6 and 15 MV photon beams, respectively. The magnitude of these effects was not found to correlate with the treatment plan technique, i.e., AP/PA versus 3D-CRT. Given a properly designed ITV, tissue heterogeneity effects are likely to have a larger clinical significance on tumor and normal lung treatment evaluation metrics than four-dimensional respiratory-induced changes.

Experimental verification of a Monte Carlo-based MLC simulation model for IMRT dose calculation

Inter- and intra-leaf transmission and head scatter can play significant roles in intensity modulated radiation therapy (IMRT)-based treatment deliveries. In order to accurately calculate the dose in the IMRT planning process, it is therefore important that the detailed geometry of the multi-leaf collimator (MLC), in addition to other components in the accelerator treatment head, be accurately modeled. In this paper, we have used the Monte Carlo method (MC) to develop a comprehensive model of the Varian 120 leaf MLC and have compared it against measurements in homogeneous phantom geometries under different IMRT delivery circumstances. We have developed a geometry module within the DPM MC code to simulate the detailed MLC design and the collimating jaws. Tests consisting of leakage, leaf positioning and static MLC shapes were performed to verify the accuracy of transport within the MLC model. The calculations show agreement within 2% in the high dose region for both film and ion-chamber measurements for these static shapes. Clinical IMRT treatment plans for the breast [both segmental MLC (SMLC) and dynamic MLC (DMLC)], prostate (SMLC) and head and neck split fields (SMLC) were also calculated and compared with film measurements. Such a range of cases were chosen to investigate the accuracy of the model as a function of modulation in the beamlet pattern, beamlet width, and field size. The overall agreement is within 2% /2 mm of the film data for all IMRT beams except the head and neck split field, which showed differences up to 5% in the high dose regions. Various sources of uncertainties in these comparisons are discussed.

An integrated Monte Carlo dosimetric verification system for radiotherapy treatment planning

An integrated Monte Carlo (MC) dose calculation system, MCRTV (Monte Carlo for radiotherapy treatment plan verification), has been developed for clinical treatment plan verification, especially for routine quality assurance (QA) of intensity-modulated radiotherapy (IMRT) plans. The MCRTV system consists of the EGS4/PRESTA MC codes originally written for particle transport through the accelerator, the multileaf collimator (MLC), and the patient/phantom, which run on a 28-CPU Linux cluster, and the associated software developed for the clinical implementation. MCRTV has an interface with a commercial treatment planning system (TPS) (Eclipse, Varian Medical Systems, Palo Alto, CA, USA) and reads the information needed for MC computation transferred in DICOM-RT format. The key features of MCRTV have been presented in detail in this paper. The phase-space data of our 15 MV photon beam from a Varian Clinac 2300C/D have been developed and several benchmarks have been performed under homogeneous and several inhomogeneous conditions (including water, aluminium, lung and bone media). The MC results agreed with the ionization chamber measurements to within 1% and 2% for homogeneous and inhomogeneous conditions, respectively. The MC calculation for a clinical prostate IMRT treatment plan validated the implementation of the beams and the patient/phantom configuration in MCRTV.

How extensive of a 4D dataset is needed to estimate cumulative dose distribution plan evaluation metrics in conformal lung therapy?

The purpose of this study was to investigate the number of intermediate states required to adequately approximate the clinically relevant cumulative dose to deforming/moving thoracic anatomy in four-dimensional (4D) conformal radiotherapy that uses 6 MV photons to target tumors. Four patients were involved in this study. For the first three patients, computed tomography images acquired at exhale and inhale were available; they were registered using B-spline deformation model and the computed transformation was further used to simulate intermediate states between exhale and inhale. For the fourth patient, 4D-acquired, phase-sorted datasets were available and each dataset was registered with the exhale dataset. The exhale-inhale transformation was also used to simulate intermediate states in order to compare the cumulative doses computed using the actual and the simulated datasets. Doses to each state were calculated using the Dose Planning Method (DPM) Monte Carlo code and dose was accumulated for scoring on the exhale anatomy via the transformation matrices for each state and time weighting factors. Cumulative doses were estimated using increasing numbers of intermediate states and compared to simpler scenarios such as a "2-state" model which used only the exhale and inhale datasets or the dose received during the average phase of the breathing cycle. Dose distributions for each modeled state as well as the cumulative doses were assessed using dose volume histograms and several treatment evaluation metrics such as mean lung dose, normal tissue complication probability, and generalized uniform dose. Although significant "point dose" differences can exist between each breathing state, the differences decrease when cumulative doses are considered, and can become less significant yet in terms of evaluation metrics depending upon the clinical end point. This study suggests that for certain "clinical" end points of importance for lung cancer, satisfactory predictions of accumulated total dose to be received by the distorting anatomy can be achieved by calculating the dose to but a few (or even simply the average) phases of the breathing cycle.

Method for Fast CT/SPECT-Based 3D Monte Carlo Absorbed Dose Computations in Internal Emitter Therapy

The DPM (Dose Planning Method) Monte Carlo electron and photon transport program, designed for fast computation of radiation absorbed dose in external beam radiotherapy, has been adapted to the calculation of absorbed dose in patient-specific internal emitter therapy. Because both its photon and electron transport mechanics algorithms have been optimized for fast computation in 3D voxelized geometries (in particular, those derived from CT scans), DPM is perfectly suited for performing patient-specific absorbed dose calculations in internal emitter therapy. In the updated version of DPM developed for the current work, the necessary inputs are a patient CT image, a registered SPECT image, and any number of registered masks defining regions of interest. DPM has been benchmarked for internal emitter therapy applications by comparing computed absorption fractions for a variety of organs using a Zubal phantom with reference results from the Medical Internal Radionuclide Dose (MIRD) Committee standards. In addition, the β decay source algorithm and the photon tracking algorithm of DPM have been further benchmarked by comparison to experimental data. This paper presents a description of the program, the results of the benchmark studies, and some sample computations using patient data from radioimmunotherapy studies using (131)I.

Synchronized dynamic dose reconstruction

Variations in target volume position between and during treatment fractions can lead to measurable differences in the dose distribution delivered to each patient. Current methods to estimate the ongoing cumulative delivered dose distribution make idealized assumptions about individual patient motion based on average motions observed in a population of patients. In the delivery of intensity modulated radiation therapy (IMRT) with a multi-leaf collimator (MLC), errors are introduced in both the implementation and delivery processes. In addition, target motion and MLC motion can lead to dosimetric errors from interplay effects. All of these effects may be of clinical importance. Here we present a method to compute delivered dose distributions for each treatment beam and fraction, which explicitly incorporates synchronized real-time patient motion data and real-time fluence and machine configuration data. This synchronized dynamic dose reconstruction method properly accounts for the two primary classes of errors that arise from delivering IMRT with an MLC: (a) Interplay errors between target volume motion and MLC motion, and (b) Implementation errors, such as dropped segments, dose over/under shoot, faulty leaf motors, tongue-and-groove effect, rounded leaf ends, and communications delays. These reconstructed dose fractions can then be combined to produce high-quality determinations of the dose distribution actually received to date, from which individualized adaptive treatment strategies can be determined.

Comparison of PENELOPE Monte Carlo dose calculations with Fricke dosimeter and ionization chamber measurements in heterogeneous phantoms (18 MeV electron and 12 MV photon beams)

Different measurements of depth-dose curves and dose profiles were performed in heterogeneous phantoms and compared to dose distributions calculated by a Monte Carlo code. These heterogeneous phantoms consisted of lung and/or bone heterogeneities. Irradiations and simulations were carried out for an 18 MeV electron beam and a 12 MV photon beam. Depth-dose curves were measured with Fricke dosimeters and with plane and cylindrical ionization chambers. Dose profiles were measured with a small cylindrical ionization chamber at different depths. The LINAC was modelled using the PENELOPE code and phase space files were used as input data for the calculations of the dose distributions in every simulation. The detectors (Fricke dosimeters and ionization chambers) were not modelled in the geometry. There is generally a good agreement between the measurements and PENELOPE. Some discrepancies exist, near interfaces, between the ionization chamber and PENELOPE due to the attenuation of the lower energy electrons by the wall of the ionization chamber.

Testing of the analytical anisotropic algorithm for photon dose calculation

The analytical anisotropic algorithm (AAA) was implemented in the Eclipse (Varian Medical Systems) treatment planning system to replace the single pencil beam (SPB) algorithm for the calculation of dose distributions for photon beams. AAA was developed to improve the dose calculation accuracy, especially in heterogeneous media. The total dose deposition is calculated as the superposition of the dose deposited by two photon sources (primary and secondary) and by an electron contamination source. The photon dose is calculated as a three-dimensional convolution of Monte-Carlo precalculated scatter kernels, scaled according to the electron density matrix. For the configuration of AAA, an optimization algorithm determines the parameters characterizing the multiple source model by optimizing the agreement between the calculated and measured depth dose curves and profiles for the basic beam data. We have combined the acceptance tests obtained in three different departments for 6, 15, and 18 MV photon beams. The accuracy of AAA was tested for different field sizes (symmetric and asymmetric) for open fields, wedged fields, and static and dynamic multileaf collimation fields. Depth dose behavior at different source-to-phantom distances was investigated. Measurements were performed on homogeneous, water equivalent phantoms, on simple phantoms containing cork inhomogeneities, and on the thorax of an anthropomorphic phantom. Comparisons were made among measurements, AAA, and SPB calculations. The optimization procedure for the configuration of the algorithm was successful in reproducing the basic beam data with an overall accuracy of 3%, 1 mm in the build-up region, and 1%, 1 mm elsewhere. Testing of the algorithm in more clinical setups showed comparable results for depth dose curves, profiles, and monitor units of symmetric open and wedged beams below dmax. The electron contamination model was found to be suboptimal to model the dose around dmax, especially for physical wedges at smaller source to phantom distances. For the asymmetric field verification, absolute dose difference of up to 4% were observed for the most extreme asymmetries. Compared to the SPB, the penumbra modeling is considerably improved (1%, 1 mm). At the interface between solid water and cork, profiles show a better agreement with AAA. Depth dose curves in the cork are substantially better with AAA than with SPB. Improvements are more pronounced for 18 MV than for 6 MV. Point dose measurements in the thoracic phantom are mostly within 5%. In general, we can conclude that, compared to SPB, AAA improves the accuracy of dose calculations. Particular progress was made with respect to the penumbra and low dose regions. In heterogeneous materials, improvements are substantial and more pronounced for high (18 MV) than for low (6 MV) energies.

Reporting and analyzing statistical uncertainties in Monte Carlo-based treatment planning

PURPOSE

To investigate methods of reporting and analyzing statistical uncertainties in doses to targets and normal tissues in Monte Carlo (MC)-based treatment planning.

METHODS AND MATERIALS

Methods for quantifying statistical uncertainties in dose, such as uncertainty specification to specific dose points, or to volume-based regions, were analyzed in MC-based treatment planning for 5 lung cancer patients. The effect of statistical uncertainties on target and normal tissue dose indices was evaluated. The concept of uncertainty volume histograms for targets and organs at risk was examined, along with its utility, in conjunction with dose volume histograms, in assessing the acceptability of the statistical precision in dose distributions. The uncertainty evaluation tools were extended to four-dimensional planning for application on multiple instances of the patient geometry. All calculations were performed using the Dose Planning Method MC code.

RESULTS

For targets, generalized equivalent uniform doses and mean target doses converged at 150 million simulated histories, corresponding to relative uncertainties of less than 2% in the mean target doses. For the normal lung tissue (a volume-effect organ), mean lung dose and normal tissue complication probability converged at 150 million histories despite the large range in the relative organ uncertainty volume histograms. For "serial" normal tissues such as the spinal cord, large fluctuations exist in point dose relative uncertainties.

CONCLUSIONS

The tools presented here provide useful means for evaluating statistical precision in MC-based dose distributions. Tradeoffs between uncertainties in doses to targets, volume-effect organs, and "serial" normal tissues must be considered carefully in determining acceptable levels of statistical precision in MC-computed dose distributions.

Monte Carlo dose calculations for a 6-MV photon beam in a thorax phantom

PURPOSE

In this study we evaluated the accuracy of the Monte Carlo (MC) and effective path length (EPL) methods for dose calculations in the inhomogeneous thorax phantom.

MATERIALS AND METHODS

The Philips SL 75/5 linear accelerator head was modeled using the MCNP4C Monte Carlo code. An anatomic inhomogeneous thorax phantom was irradiated with a 6-MV photon beam, and the doses along points of the central axis of the beam were measured by a small ionization chamber. The central axis relative dose was calculated by the MCNP4C code and the EPL method in a conventional treatment planning system. The results of calculations and measurements were compared.

RESULTS

For all measured points on the thorax phantom the results of the MC method were in agreement with the actual measurement (local difference was less than 2%). For the EPL method, the amount of error was dependent on the field size and the point location in the phantom. The maximum error was +19.5 and +26.8 for field sizes of 10 x 10 and 5 x 5 cm2 for lateral irradiation.

CONCLUSION

Our study showed large, unacceptable errors for EPL calculations in the lung for both field sizes. The accuracy of the MC method was better than the recommended value of 3%. Thus, application of this method is strongly recommended for lung dose calculations, especially for small field sizes.

Multiple fields may offer better esophagus sparing without increased probability of lung toxicity in optimized IMRT of lung tumors

PURPOSE

To evaluate whether increasing numbers of intensity-modulated radiation therapy (IMRT) fields enhance lung-tumor dose without additional predicted toxicity for difficult planning geometries.

METHODS AND MATERIALS

Data from 8 previous three dimensional conformal radiation therapy (3D-CRT) patients with tumors located in various regions of each lung, but with planning target volumes (PTVs) overlapping part of the esophagus, were used as input. Four optimized-beamlet IMRT plans (1 plan that used the 3D-CRT beam arrangement and 3 plans with 3, 5, or 7 axial, but predominantly one-sided, fields) were compared. For IMRT, the equivalent uniform dose (EUD) in the whole PTV was optimized simultaneously with that in a reduced PTV exclusive of the esophagus. Normal-tissue complication probability-based costlets were used for the esophagus, heart, and lung.

RESULTS

Overall, IMRT plans (optimized by use of EUD to judiciously allow relaxed PTV dose homogeneity) result in better minimum PTV isodose surface coverage and better average EUD values than does conformal planning; dose generally increases with the number of fields. Even 7-field plans do not significantly alter normal-lung mean-dose values or lung volumes that receive more than 13, 20, or 30 Gy.

CONCLUSION

Optimized many-field IMRT plans can lead to escalated lung-tumor dose in the special case of esophagus overlapping PTV, without unacceptable alteration in the dose distribution to normal lung.

A proposed alternative to phase-space recycling using the adaptive kernel density estimator method

We have implemented a nonparametric density estimation technique, the adaptive kernel density estimator (AKDE), to generate additional phase space (PS) variables in the vicinity of simulated PS points in Monte Carlo linear accelerator simulation. The method involves the placement of kernels at simulated PS points that have a "window width" that depends on the density of simulated PS points. This method has been tested on known one-dimensional (1-D) and two-dimensional (2-D) probability density functions (PDFs) and has been used to sample (photons only) from PS files generated from accelerator simulations. The original simulated PS vector (x, y, u, v, E) was reduced to a rotationally invariant PS vector (r, theta, alpha, E) that takes advantage of the azimuthal symmetry (phi) above the collimating jaws. The new PS vector (r', theta', alpha', E') is sampled in the vicinity of the sampled PS vector (r, theta, alpha, E). The first step in assessing the accuracy of the method was a correlation analysis among the AKDE generated PS variables compared with correlations among the original PS variables. "In-air" particle fluence distributions between AKDE samples and the original PS distribution showed agreement within 2% (-8.8% to 6.8%) across the entire phase space plane. Central axis energy distributions and angular distributions agreed on average to within 1.5% (range = -1.5% to 6.6%) and 0.1% (range = 0 to 3.0%), respectively. Dose profiles were calculated for field sizes 3 x 3 cm2, 10 x 10 cm2, and 30 x 30 cm2 for AKDE and compared against calculations performed with PS recycling. AKDE calculated depth doses and profiles were within 2% and 2%/1 mm, respectively, of those computed using PS recycling.

Dose evaluation in lung-equivalent media in high-energy photon external radiotherapy

In high-energy photon external radiotherapy treatment planning systems (TPSs) are used to calculate the dose to the target volume and the dose distribution around it. Commonly used TPSs include algorithms based on measurements in water and often fail in the estimate of dose in the presence of heterogeneities. In this study TL detectors were used to study the reliability of the Cadplan (Varian) TPS in the presence of low-density heterogeneities such as the lung for 6 and 18 MV photon beams at different field sizes. TL measurements were compared with TPS calculations and Monte Carlo simulations performed with the PENELOPE MC code. In a phantom with lung heterogeneity, TL measurements and MC simulations agreed, with an average deviation inside the lung of 2%. In contrast, TPS results overestimated the dose inside the lung, with a maximum deviation of 39% for the 18 MV photon beam and a field size of 2 x 2 cm(2).

Monte Carlo-based lung cancer treatment planning incorporating PET-defined target volumes

Despite the well-known benefits of positron emission tomography (PET) imaging in lung cancer diagnosis and staging, the poor spatial resolution of PET has limited its use in radiotherapy planning. Methods used for segmenting tumor from normal tissue, such as threshold boundaries using a fraction of the standardized uptake value (SUV), are subject to uncertainties. The issue of respiratory motion in the thorax confounds the problem of accurate target definition. In this work, we evaluate how changing the PET-defined target volume by varying the threshold value in the segmentation process impacts target and normal lung tissue doses. For each of eight lung cancer patients we retrospectively generated multiple PET-target volumes; each target volume corresponds to those voxels with intensities above a given threshold level, defined by a percentage of the maximum voxel intensity. PET-defined targets were compared to those from CT; CT targets comprise a composite volume generated from breath-hold inhale and exhale datasets; the CT dataset therefore also includes the extents of tumor motion. Treatment plans using Monte Carlo dose calculation were generated for all targets; the dose uniformity was approximately 100+/-5% within the internal target volume (ITV) (formed by a uniform 8-mm expansion of the composite gross target volume (GTV)). In all cases differences were observed in the generalized equivalent uniform doses (gEUDs) to the targets and in the mean lung doses (MLDs) and normal tissue complication probabilities (NTCPs) to the normal lung tissues. The magnitudes of the dose differences were found to depend on the target volume, location, and amount of irradiated normal lung tissue, and in many instances were clinically meaningful (greater than a single 2 Gy fraction). For those patients studied, results indicate that accurate dosimetry using PET volumes is highly dependent on accurate target segmentation. Further study with correlation to clinical outcome will be helpful in determining how to apply these various PET and CT volumes in treatment planning, to potentially improve local tumor control and reduce normal tissue toxicities.

Esophagus sparing with IMRT in lung tumor irradiation: An EUD-based optimization technique

PURPOSE

The aim of this study was to evaluate (1) the use of generalized equivalent uniform dose (gEUD) to optimize dose escalation of lung tumors when the esophagus overlaps the planning target volume (PTV) and (2) the potential benefit of further dose escalation in only the part of the PTV that does not overlap the esophagus.

METHODS AND MATERIALS

The treatment-planning computed tomography (CT) scans of patients with primary lung tumors located in different regions of the left and right lung were used for the optimization of beamlet intensity modulated radiation therapy (IMRT) plans. In all cases, the PTV overlapped part of the esophagus. The dose in the PTV was maximized according to 7 different primary cost functions: 2 plans that made use of mean dose (MD) (the reference plan, in which the 95% isodose surface covered the PTV and a second plan that had no constraint on the minimum isodose), 3 plans based on maximizing gEUD for the whole PTV with ever increasing assumptions for tumor aggressiveness, and 2 plans that used different gEUD values in 2 simultaneous, overlapping target volumes (the whole PTV and the PTV minus esophagus). Beam arrangements and NTCP-based costlets for the organs at risk (OARs) were kept identical to the original conformal plan for each case. Regardless of optimization method, the relative ranking of the resulting plans was evaluated in terms of the absence of cold spots within the PTV and the final gEUD computed for the whole PTV.

RESULTS

Because the MD-optimized plans lacked a constraint on minimum PTV coverage, they resulted in cold spots that affected approximately 5% of the PTV volume. When optimizing over the whole PTV volume, gEUD-optimized plans resulted in higher equivalent uniform PTV doses than did the reference plan while still maintaining normal-tissue constraints. However, only under the assumption of extremely aggressive tumors could cold spots in the PTV be avoided. Generally, high-level overall results are obtained when optimization in the whole PTV is also associated with a second simultaneous optimization in the PTV minus overlapping portions of the esophagus.

CONCLUSIONS

Intensity modulated radiation therapy optimizations that utilize gEUD-based cost functions for the PTV and NTCP-based constraints for the OARs result in increased doses to large portions of the PTV in cases where the PTV overlaps the esophagus, while still maintaining (and confining to the overlap region) minimum dose coverage equivalent to the homogeneous PTV optimization cases.

Investigation of geometrical and scoring grid resolution for Monte Carlo dose calculations for IMRT

Monte Carlo based treatment planning of two different patient groups treated with step-and-shoot IMRT (head-and-neck and lung treatments) with different CT resolutions and scoring methods is performed to determine the effect of geometrical and scoring voxel sizes on DVHs and calculation times. Dose scoring is performed in two different ways: directly into geometrical voxels (or in a number of grouped geometrical voxels) or into scoring voxels defined by a separate scoring grid superimposed on the geometrical grid. For the head-and-neck cancer patients, more than 2% difference is noted in the right optical nerve when using voxel dimensions of 4 x 4 x 4 mm3 compared to the reference calculation with 1 x 1 x 2 mm3 voxel dimensions. For the lung cancer patients, 2% difference is noted in the spinal cord when using voxel dimensions of 4 x 4 x 10 mm3 compared to the 1 x 1 x 5 mm3 calculation. An independent scoring grid introduces several advantages. In cases where a relatively high geometrical resolution is required and where the scoring resolution is less important, the number of scoring voxels can be limited while maintaining a high geometrical resolution. This can be achieved either by grouping several geometrical voxels together into scoring voxels or by superimposing a separate scoring grid of spherical voxels with a user-defined radius on the geometrical grid. For the studied lung cancer cases, both methods produce accurate results and introduce a speed increase by a factor of 10-36. In cases where a low geometrical resolution is allowed, but where a high scoring resolution is required, superimposing a separate scoring grid on the geometrical grid allows a reduction in geometrical voxels while maintaining a high scoring resolution. For the studied head-and-neck cancer cases, calculations performed with a geometrical resolution of 2 x 2 x 2 mm3 and a separate scoring grid containing spherical scoring voxels with a radius of 2 mm produce accurate results and introduce a speed increase by a factor of 13. The scoring grid provides an additional degree of freedom for limiting calculation time and memory requirements by selecting optimized scoring and geometrical voxel dimensions in an independent way.

Dose reconstruction in deforming lung anatomy: Dose grid size effects and clinical implications

In this study we investigated the accumulation of dose to a deforming anatomy (such as lung) based on voxel tracking and by using time weighting factors derived from a breathing probability distribution function (p.d.f.). A mutual information registration scheme (using thin-plate spline warping) provided a transformation that allows the tracking of points between exhale and inhale treatment planning datasets (and/or intermediate state scans). The dose distributions were computed at the same resolution on each dataset using the Dose Planning Method (DPM) Monte Carlo code. Two accumulation/interpolation approaches were assessed. The first maps exhale dose grid points onto the inhale scan, estimates the doses at the "tracked" locations by trilinear interpolation and scores the accumulated doses (via the p.d.f.) on the original exhale data set. In the second approach, the "volume" associated with each exhale dose grid point (exhale dose voxel) is first subdivided into octants, the center of each octant is mapped to locations on the inhale dose grid and doses are estimated by trilinear interpolation. The octant doses are then averaged to form the inhale voxel dose and scored at the original exhale dose grid point location. Differences between the interpolation schemes are voxel size and tissue density dependent, but in general appear primarily only in regions with steep dose gradients (e.g., penumbra). Their magnitude (small regions of few percent differences) is less than the alterations in dose due to positional and shape changes from breathing in the first place. Thus, for sufficiently small dose grid point spacing, and relative to organ motion and deformation, differences due solely to the interpolation are unlikely to result in clinically significant differences to volume-based evaluation metrics such as mean lung dose (MLD) and tumor equivalent uniform dose (gEUD). The overall effects of deformation vary among patients. They depend on the tumor location, field size, volume expansion, tissue heterogeneity, and direction of tumor displacement with respect to the beam, and are more likely to have an impact on serial organs (such as esophagus), rather than on large parallel organs (such as lung).

Experimental verification of lung dose with radiochromic film: comparison with Monte Carlo simulations and commercially available treatment planning systems

The purpose of this study was to assess the absorbed dose in and around lung tissue by performing radiochromic film measurements, Monte Carlo simulations and calculations with superposition convolution algorithms. We considered a layered polystyrene phantom of 12 x 12 x 12 cm3 containing a central cavity of 6 x 6 x 6 cm3 filled with Gammex RMI lung-equivalent material. Two field configurations were investigated, a small 1 x 10 cm2 field and a larger 10 x 10 cm2 field. First, we performed Monte Carlo simulations to investigate the influence of radiochromic film itself on the measured dose distribution when the film intersects a lung-equivalent region and is oriented parallel to the central beam axis. To that end, the film and the lung-equivalent materials were modelled in detail, taking into account their specific composition. Next, measurements were performed with the film oriented both parallel and perpendicular to the central beam axis to verify the results of our Monte Carlo simulations. Finally, we digitized the phantom in two commercially available treatment planning systems, Helax-TMS version 6.1A and Pinnacle version 6.2b, and calculated the absorbed dose in the phantom with their incorporated superposition convolution algorithms to compare with the Monte Carlo simulations. Comparing Monte Carlo simulations with measurements reveals that radiochromic film is a reliable dosimeter in and around lung-equivalent regions when the film is positioned perpendicular to the central beam axis. Radiochromic film is also able to predict the absorbed dose accurately when the film is positioned parallel to the central beam axis through the lung-equivalent region. However, attention must be paid when the film is not positioned along the central beam axis, in which case the film gradually attenuates the beam and decreases the dose measured behind the cavity. This underdosage disappears by offsetting the film a few centimetres. We find deviations of about 3.6% between Monte Carlo and the superposition convolution algorithm of Pinnacle behind the lung region, for both field configurations. Pinnacle is quite accurate in the lung region. Deviations up to 5.6% for the small field are found in the lung region between Monte Carlo and the superposition convolution algorithm of Helax-TMS. Behind the lung region, Helax-TMS is in better agreement with Monte Carlo. Radiochromic film measurements or Monte Carlo simulations are reliable methods to establish the dose in and around lung tissue.

The influence of beam model differences in the comparison of dose calculation algorithms for lung cancer treatment planning

In this study, we show that beam model differences play an important role in the comparison of does calculated with various algorithms for lung cancer treatment planning. These differences may impact the accurate correlation of dose with clinical outcome. To accomplish this, we modified the beam model penumbral parameters in an equivalent path length (EPL) algorithm and subsequently compared the EPL doses with those generated with Monte Carlo (MC). A single AP beam was used for beam fitting. Two different beam models were generated for EPL calculations: (1) initial beam model (init_fit) and (2) optimized beam model (best_fit) , with parameters optimized to produce the best agreement with MC calculated profiles at several depths in a water phantom. For the 6 MV, AP beam, EPL(init_fit) calculations were on average within 2%/2 mm (1.4 mm max.) agreement with MC; the agreement for EPL(best_fit) was 2%/1.0 mm (1.3 mm max.) for EPL(best_fit). Treatment planning was performed using a realistic lung phantom using 6 and 15 MV photons. In all homogeneous phantom plans, EPL(best_fit) calculations were in better agreement with MC. In the heterogeneous 6 MV plan, differences between EPL(best_fit and init_fit) and MC were significant for the tumour. The EPL(init_fit), unlike the EPL(best_fit) calculation, showed large differences in the lung relative to MC. For the 15 MV heterogeneous plan, clinically important differences were found between EPL(best_fit or init_fit) and MC for tumour and lung, suggesting that the algorithmic difference in inhomogeneous cases, differences between EPL(best_fit) and MC for lung tissues were smaller compared to those between EPL(init_fit) and MC. Although the extent to which beam model differences impact the dose comparisons will be dependent upon beam parameters (orientation, field size and energy), and the size and location of the tumour, this study shows that failing to correctly account for beam model differences will lead to biased comparisons between dose algorithms. This may ultimately hinder our ability to accurately correlate dose with clinical outcome.

MCDE: a new Monte Carlo dose engine for IMRT

A new accurate Monte Carlo code for IMRT dose computations, MCDE (Monte Carlo dose engine), is introduced. MCDE is based on BEAMnrc/DOSXYZnrc and consequently the accurate EGSnrc electron transport. DOSXYZnrc is reprogrammed as a component module for BEAMnrc. In this way both codes are interconnected elegantly, while maintaining the BEAM structure and only minimal changes to BEAMnrc.mortran are necessary. The treatment head of the Elekta SLiplus linear accelerator is modelled in detail. CT grids consisting of up to 200 slices of 512 x 512 voxels can be introduced and up to 100 beams can be handled simultaneously. The beams and CT data are imported from the treatment planning system GRATIS via a DICOM interface. To enable the handling of up to 50 x 10(6) voxels the system was programmed in Fortran95 to enable dynamic memory management. All region-dependent arrays (dose, statistics, transport arrays) were redefined. A scoring grid was introduced and superimposed on the geometry grid, to be able to limit the number of scoring voxels. The whole system uses approximately 200 MB of RAM and runs on a PC cluster consisting of 38 1.0 GHz processors. A set of in-house made scripts handle the parallellization and the centralization of the Monte Carlo calculations on a server. As an illustration of MCDE, a clinical example is discussed and compared with collapsed cone convolution calculations. At present, the system is still rather slow and is intended to be a tool for reliable verification of IMRT treatment planning in the case of the presence of tissue inhomogeneities such as air cavities.

ORANGE: a Monte Carlo dose engine for radiotherapy

This study presents data for the verification of ORANGE, a fast MCNP-based dose engine for radiotherapy treatment planning. In order to verify the new algorithm, it has been benchmarked against DOSXYZ and against measurements. For the benchmarking, first calculations have been done using the ICCR-XIII benchmark. Next, calculations have been done with DOSXYZ and ORANGE in five different phantoms (one homogeneous, two with bone equivalent inserts and two with lung equivalent inserts). The calculations have been done with two mono-energetic photon beams (2 MeV and 6 MeV) and two mono-energetic electron beams (10 MeV and 20 MeV). Comparison of the calculated data (from DOSXYZ and ORANGE) against measurements was possible for a realistic 10 MV photon beam and a realistic 15 MeV electron beam in a homogeneous phantom only. For the comparison of the calculated dose distributions and dose distributions against measurements, the concept of the confidence limit (CL) has been used. This concept reduces the difference between two data sets to a single number, which gives the deviation for 90% of the dose distributions. Using this concept, it was found that ORANGE was always within the statistical bandwidth with DOSXYZ and the measurements. The ICCR-XIII benchmark showed that ORANGE is seven times faster than DOSXYZ, a result comparable with other accelerated Monte Carlo dose systems when no variance reduction is used. As shown for XVMC, using variance reduction techniques has the potential for further acceleration. Using modern computer hardware, this brings the total calculation time for a dose distribution with 1.5% (statistical) accuracy within the clinical range (less then 10 min). This means that ORANGE can be a candidate for a dose engine in radiotherapy treatment planning.