Pinnacle Calculations Research Articles

Abstract ID: 200 Evaluation of key parameters for non-small cell lung cancer treatments using Geant4 as benchmark dose calculation algorithm

Significant variations in tumor control probability (TCP) have been reported according to differences found at dose calculations between treatment planning system (TPS) and Monte Carlo (MC) method [1] . Our purpose is to quantify dosimetric and radiobiological effects due to divergences between dose distributions calculated by Pinnacle3 TPS (version 9.8) and Geant4 toolkit (version 10.01.p01) [2] , [3] for lung cancer cases treated with photons. Three clinical cases with 6 MV photon beams and conventional fractionation were analysed. These cases are distinguished by their tumoral size and localization. From dose distributions calculated through Geant4 and TPS we have obtained conformity index (CI) and homogeneity index (HI) [4] . The TCP was achieved using LQ model [5] with an alpha–beta ratio of 10 Gy and a density of clonogenic cells of 107 cells/cm3. For each clinical case the TCP values according to Geant4 calculations against Pinnacle calculations were 93.77 vs 97.74, 92.95 vs 92.79 and 98.31 vs 99.03 respectively. These deviations achieve a maximum value close to 4% and they are in agreement with the CI and HI values obtained in every single treatment. We have verified lung cancer treatments through Geant4 toolkit and we have found more heterogeneous dose distributions and variations in TCP up to 4% with respect to TPS calculations. These discrepancies are due to incomplete modeling in the TPS algorithm of patient tissue heterogeneities and photon fluences used in modulated beam techniques. A further study including more sophisticated radiobiology models and hypofractionated radiotherapy schemes will be subject of further work.

SU‐F‐T‐623: Dosimetric and Radiobiological Comparison of Conventional and Monte Carlo Treatment Plans of Lung Tumors Using SBRT

Purpose:This study aims at estimating the dosimetric differences of Pinnacle and Monte Carlo (MC) calculation algorithms in computing the dose distributions in lung cancer SBRT cases. Furthermore, to interpret the dosimetric findings into expected tissue response rates in order to estimate their clinical impact.Methods:17 patients were studied and for each one, two treatment plans (one from Pinnacle and one MC) were created. The structures that were used for the optimization were: PTV, ipsilateral and contralateral lungs, spinal cord, heart, esophagus and stomach. The same dosimetric constraints were used for both pairs of plans. These dosimetric indices were used for the comparison of the two calculation modalities. Additionally, the TCP and NTCP values of the different targets and organs at risk (OAR) were calculated and compared together with the complicationfree tumor control probability.Results:The proposed plan evaluation shows that the plans by Pinnacle deliver higher doses to the OAR than MC. The same is observed for the PTV (minimum dose and BEUD: 58.3 vs. 55.1Gy, respectively). However, these dosimetric differences are translated to negligible differences in TCP and NTCP.Conclusion:The results of this work indicate that for the fractionation scheme of 18Gy in 3 fractions, the treatment plans in Pinnacle and Monte Carlo show minor differences in the doses to the organs at risk (in all the cases the doses by MC were lower than those calculated by Pinnacle). Those doses are predicted to cause 0.0% of side effects. On the other hand, the findings of the analysis indicate that Pinnacle calculation is shown to deliver higher dose to the PTV than MC. However, the dosimetric differences do not appear to have a measurable impact in the expected TCP rates of the targets.

SU-F-T-614: Comparison of Pinnacle and Monaco Dose Calculations of SBRT Treatments

Purpose: The purpose of this study was to evaluate, compare and investigate the differences of SBRT treatment plans on dose calculation and plan metrics between Pinnacle and Monaco treatment planning systems (TPS). Methods: 14 previously treated patients’ files were exported from the Pinnacle treatment planning system to Monaco. All of the treatment plans were created in Pinnacle and delivered on a Varian Novalis with the HD120 multi-leaf collimator. The plans were recalculated within Monaco using a Monte-Carlo algorithm. From the calculated dose distributions, the PITV, R50, conformity and homogeneity indices were computed and used for SBRT plan evaluation. Dose volume histogram comparison and OAR dose statistics were also used for the comparison. Results: The mean PITV for Monaco increased 7% over Pinnacle, while the standard deviation increased from 0.047 to 0.076. The Homogeneity Index increased from 1.32 to 1.35. The Target Coverage decreased slightly from 0.98 to 0.97. The standard deviations for Homogeneity Index and Target coverage varied insignificantly by less than 0.2%. Conclusion: The PITV increasing shows that the prescription dose volume increases, since the PTV will not change between treatment planning systems. This indicates that the more accurate dose near density change interfaces from Monaco shows the original planning has the dose shifted away from the isocenter. The homogeneity index increases shows that there is more of a change between the minimum and maximum doses within the PTV so the target coverage decreases.

SU‐E‐T‐35: An Investigation of the Accuracy of Cervical IMRT Dose Distribution Using 2D/3D Ionization Chamber Arrays System and Monte Carlo Simulation

Purpose:The purpose of this work is to compare the verification results of three solutions (2D/3D ionization chamber arrays measurement and Monte Carlo simulation), the results will help make a clinical decision as how to do our cervical IMRT verification.Methods:Seven cervical cases were planned with Pinnacle 8.0m to meet the clinical acceptance criteria. The plans were recalculated in the Matrixx and Delta4 phantom with the accurate plans parameters. The plans were also recalculated by Monte Carlo using leaf sequences and MUs for individual plans of every patient, Matrixx and Delta4 phantom. All plans of Matrixx and Delta4 phantom were delivered and measured. The dose distribution of iso slice, dose profiles, gamma maps of every beam were used to evaluate the agreement. Dose‐volume histograms were also compared.Results:The dose distribution of iso slice and dose profiles from Pinnacle calculation were in agreement with the Monte Carlo simulation, Matrixx and Delta4 measurement. A 95.2%/91.3% gamma pass ratio was obtained between the Matrixx/Delta4 measurement and Pinnacle distributions within 3mm/3% gamma criteria. A 96.4%/95.6% gamma pass ratio was obtained between the Matrixx/Delta4 measurement and Monte Carlo simulation within 2mm/2% gamma criteria, almost 100% gamma pass ratio within 3mm/3% gamma criteria. The DVH plot have slightly differences between Pinnacle and Delta4 measurement as well as Pinnacle and Monte Carlo simulation, but have excellent agreement between Delta4 measurement and Monte Carlo simulation.Conclusion:It was shown that Matrixx/Delta4 and Monte Carlo simulation can be used very efficiently to verify cervical IMRT delivery. In terms of Gamma value the pass ratio of Matrixx was little higher, however, Delta4 showed more problem fields. The primary advantage of Delta4 is the fact it can measure true 3D dosimetry while Monte Carlo can simulate in patients CT images but not in phantom.

Characterization of a novel 2D array dosimeter for patient‐specific quality assurance with volumetric arc therapy

In this study, the authors are evaluating a new, commercially available 2D array that offers 3D dose reconstruction for patient specific intensity modulated radiation therapy quality assurance (IMRT QA). The OCTAVIUS 4D system and its accompanying software (VERISOFT) by PTW were evaluated for the accuracy of the dose reconstruction for patient specific pretreatment IMRT QA. OCTAVIUS 4D measures the dose plane at the linac isocenter as the phantom rotates synchronously with the gantry, maintaining perpendicularity with the beam, by means of an inclinometer and a motor. The measurements collected during a volumetric modulated arc therapy delivery (VMAT) are reconstructed into a 3D dose volume. The VERISOFT application is used to perform the analysis, by comparing the reconstructed dose against the 3D dose matrix from the treatment planning system (TPS) that is computed for the same geometry and beam arrangement as that of the measurement. In this study, the authors evaluated the 3D dose reconstruction algorithm of this new system using a series of tests. Using the Octavius 4D phantom as the patient, dose distributions for various field sizes, beam orientations, shapes, and combination of fields were calculated using the Pinnacle3, TPS, and the respective DICOMRT dose was exported to the VERISOFT analysis software. Measurements were obtained by delivering the test treatment plans and comparisons were made based on gamma index, dose profiles, and isodose distribution analysis. In addition, output factors were measured and the dose linearity of the array was assessed. Those measurements were compared against measurements in water using a single, calibrated ionization chamber as well as calculations from Pinnacle for the same delivery geometries. The number of voxels that met the 3%/3 mm criteria for the volumetric 3D gamma index analysis ranged from 92.3% to 98.9% for all the patient plans that the authors evaluated. 2D gamma analysis in the axial, sagittal, and coronal planes produced similar results to those in the 3D gamma analysis. The new detector system does not require an angular dependence correction because it rotates in synchrony with the gantry and the detector array maintains a constant SAD while always perpendicular to the beam axis. Output factors were within 2% when compared to ionization chamber measurements and Pinnacle calculations. Similar agreement was observed when testing the MU linearity (for MU values above 2) as well as dose rate effect. The OCTAVIUS 4D system has some unique characteristics that can potentially improve the patient specific pretreatment IMRT QA data collection and analysis. The ability of the software to reconstruct from the measurements the true 3D dose distribution in the phantom, provides a unique perspective for the medical physicist that evaluates a patient's QA plan.

Independent calculation-based verification of IMRT plans using a 3D dose-calculation engine

Independent monitor unit verification of intensity-modulated radiation therapy (IMRT) plans requires detailed 3-dimensional (3D) dose verification. The aim of this study was to investigate using a 3D dose engine in a second commercial treatment planning system (TPS) for this task, facilitated by in-house software. Our department has XiO and Pinnacle TPSs, both with IMRT planning capability and modeled for an Elekta-Synergy 6MV photon beam. These systems allow the transfer of computed tomography (CT) data and RT structures between them but do not allow IMRT plans to be transferred. To provide this connectivity, an in-house computer programme was developed to convert radiation therapy prescription (RTP) files as generated by many planning systems into either XiO or Pinnacle IMRT file formats. Utilization of the technique and software was assessed by transferring 14 IMRT plans from XiO and Pinnacle onto the other system and performing 3D dose verification. The accuracy of the conversion process was checked by comparing the 3D dose matrices and dose volume histograms (DVHs) of structures for the recalculated plan on the same system. The developed software successfully transferred IMRT plans generated by 1 planning system into the other. Comparison of planning target volume (TV) DVHs for the original and recalculated plans showed good agreement; a maximum difference of 2% in mean dose, − 2.5% in D95, and 2.9% in V95 was observed. Similarly, a DVH comparison of organs at risk showed a maximum difference of +7.7% between the original and recalculated plans for structures in both high- and medium-dose regions. However, for structures in low-dose regions (less than 15% of prescription dose) a difference in mean dose up to +21.1% was observed between XiO and Pinnacle calculations. A dose matrix comparison of original and recalculated plans in XiO and Pinnacle TPSs was performed using gamma analysis with 3%/3mm criteria. The mean and standard deviation of pixels passing gamma tolerance for XiO-generated IMRT plans was 96.1 ± 1.3, 96.6 ± 1.2, and 96.0 ± 1.5 in axial, coronal, and sagittal planes respectively. Corresponding results for Pinnacle-generated IMRT plans were 97.1 ± 1.5, 96.4 ± 1.2, and 96.5 ± 1.3 in axial, coronal, and sagittal planes respectively.

MO-D-108-02: A Secondary Monitor Unit Calculation Algorithm Using Superposition of Symmetric, Open Fields for IMRT and VMAT Treatment Plans

Purpose: To perform a secondary dose calculation for an IMRT or VMAT plan to a point on or off axis within 2% using open field data. Methods: The calculation is performed by subdividing dose into the contributions from each opposing leaf pair for a given MLC configuration. Leaf pair dose is determined by drawing four rectangular fields based on leaf positions that are symmetric about the point of calculation. Field dose is determined using isocentric monitor unit calculation formalism with open field data. Superposition of these fields yields the dose from the leaf pair to the point. VMAT plans are approximated by a static MLC configuration at four degree intervals. The algorithm requires standard open field data (e.g., Scps, TPRs), and the MLC control point information. These calculations were done with additional measured small field output factors down to a 1.2 cm2. Algorithm doses were compared with homogeneous and heterogeneous Pinnacle calculations of a series of prostate, head and neck, and chest wall treatment plans. For each treatment plan the calculation point was taken as the isocenter and/or the center of the PTV. Delivery techniques included SMLC‐IMRT and VMAT. Results: Good agreement was obtained between doses calculated by the algorithm and the Pinnacle Treatment Planning System. For homogeneous cases, the total calculated doses were within 2% for all patient plans. Individual beam errors were as high as 7.8% with an average of 2.8%. Larger errors were associated for beam segments with leaf positions very near the calculation point. Larger errors were also found for heterogeneous calculations, where a simple ratio of TPR correction factor was utilized in the algorithm. Conclusion: Results demonstrate that clinically acceptable agreement is obtained using this method. Further improvement could be made with more accurate heterogeneity correction factors and/or incorporation of MLC leaf end effects. This work was supported in part by a research agreement with Oncology Data Systems, Inc.

SU-E-T-527: Assessment of Field-Junction Dosimetry: Pinnacle V9.0 versus Film (EBT2).

To investigate dose agreement of Pinnacle V9.0 with measurements across the junction between abutting photon and electron fields. Photon-photon, photon-electron, and electron-electron field junctions of energies 6 MV and 9 MeV were evaluated with 0, 5, and 10 mm gaps between the field edges at depths of 2, 15, and 30 mm with Gafchromic film (EBT2). Photon and electron fields (10×10 cm) were setup to deliver 300 cGy in their field centers to 30 mm depth. Film optical density was measured every millimeter across the field-junction with a densitometer (light beam 2 mm). Optical Density to dose conversion was accomplished using a calibration curve specific to the batch of this film. For comparison, all field combinations were generated in Pinnacle, and doses were calculated with collapsed-cone convolution algorithm employing calculation-grid sizes 2, 3, and 4 mm. Measured doses at junction are as much as 20% different than Pinnacle calculation. 3 profiles with zero gap, 1 profile with 0.5 cm gap, and 0 profiles with 1.0 cm gap showed a disagreement greater than 12%. Measurement showed a bias towards higher dose in the junction when compared to Pinnacle calculation. With dose-calculation grid reduced below 4 mm, Pinnacle calculations showed improved agreements, but only a few percent at most. Pinnacle versus measured dose disagreements do not show clear trends with field separation or depth. Doses calculated by Pinnacle V9.0 in the field-junction region may not be accurate for abutting photon and electron fields, particularly when a coarse dose-calculation grid is utilized. Lack of clear trends in dose-agreement with field separation and depth suggests that uncertainties in other factors such as jaw positions and phantom set-ups may be additional contributors. Significant inaccuracies in doses reported by the treatment-planning system in field-junction region should be considered while making clinical decisions.

SU-E-T-514: Implementation and Commissioning of a Micro Multileaf Collimator in Pinnacle Treatment Planning System.

To model and validate, in a Pinnacle treatment planning system, a Brainlab micromultileaf collimator mounted on a Primus Siemens accelerator. The objective is to take advantage of the collapsed cone convolution algorithm and the ability of this system modelling rounded leaf- end MLC's. The micro multileaf collimator was modelled using fixed accelerator jaws with a value of 9.2×9.2 cm2 . Profiles and depth dose curves for a wide range of square fields at SSD of 100 cm and depths of 1.5,5,10 and 20 cm were measured using a Scanditronix stereotactic SFD diode. Output factors were measured using a stereotactic unshielded diode for field sizes from 0.6×0.6 cm2 to 3×3 cm2 . For wider fields a Scanditronix ic15 ionization chamber was used. EDR2 films were used to measure adjacent fields in the transverse and longitudinal direction. The film measurements were compared to Pinnacle calculations to model and validate the leaf tip radius, leaf offset calibration values and tongue and groove width. Pinnacle calculations and measurements agree within 2% or 2mm except for the tails of largest fields where differences are <3.5%. Comparison of film measurements and Pinnacle calculations give the optimal value for leaf tip radius of 15 cm and for tongue and groove width of 0.04 cm. Pinnacle models a Brainlab micromultileaf collimator mounted on a Siemens Primus accelerator with acceptable results for clinical treatments.

SU-E-T-200: IMRT Patient Specific QA for On-Line Adaptive Radiotherapy.

On-line adaptive IMRT requires a highly efficient radiotherapy team and an intensive amount of work, as well as significant amount of Quality Assurance (QA). With regard to patient specific QA, it is clinically unrealistic to perform fluence measurements via a QA phantom while patient is setup awaiting treatment. We therefore developed an alternative way to check point dose and the IMRT fluence right before beam delivery. In this study, 28 IMRT plans were generated with Prowess Panther 5.1 for three prostate cancer patients on 28 CT datasets with CTV and critical organs contoured by treating physicians. The corresponding QA plans were generated by Prowess, and then transfered to Pinnacle 9.2 for fluence recalculation on a flat surface phantom, In addition 28 QA plans were delivered on a Siemens Artiste accelerator and fluence were measured with Matrixx IMRT Device. We analyzed both point dose and fluence for these 28 samples. Of all 28 IMRT plans, the point dose difference between Prowess and Pinnacle are well within 1%. The point dose difference between measurements and Prowess calculation are all within 3%. The passing rates using gamma criteria (3%3mm) for the fluence comparison between Prowess and Pinnacle calculation are at least 98.5% while the passing rates of the gamma analysis between fluence measurement and Prowess calculation are all better than 98.5%. The passing rate of gamma difference between Prowess vs Pinnacle and Prowess vs QA measurement is less than 1.5% for all 28 samples. Therefore, second TPS (e.g. Pinnacle) can be used to verify planned fluencies and can serve as a valuable patient specific QA when conventional IMRT QA measurement is not feasible. A new IMRT plan is desired if significant anatomy change from the date of original CT scan is observed before the radiation delivery. It is not clinically feasible to check the fluence on the machine before radiation delivery while the patient is in the treatment suite awaiting radiation delivery. Our method showed here is an alternative way to verify the planned fluence with a second TPS and can serve as a valuable IMRT patient specific QA in online adaptive radiotherapy.

SU-E-T-513: Clinical Implementation of a GPU Accelerated Pencil Beam Dose Calculation Algorithm.

This work reports a clinical implementation of a GPU accelerated pencil beam dose calculation algorithm (GPU-PB). Model parameters were determined using in-house scripts written in MATLAB. Dose distributions in a water phantom were calculated using a Pinnacle TPS for various open field sizes. Lateral profiles at 2-mm incremental depths were used to calculate PB kernel parameters. Weighted sum of squares were used for least-squares parameter fitting utilizing a Levenberg-marquardt algorithm. Weightings were adjusted based on goodness of fitting and in accordance with field size to suppress deviations due to horn effects. The scale factor was fitted iteratively. The calculated doses for two patient cases were analyzed with a 3D gamma method (3%/3mm). Excellent agreement between Pinnacle calculation and GPU-PB calculation was achieved regarding PDD, profiles and output factor. For a head-neck 10-field step-and-shoot IMRT case, gamma passing rate was over 99% and maximum absolute dose value is 75.6 Gy vs. 73.9 Gy, differing by 2.3%. Similar results were obtained for a SBRT lung case. Gamma passing rate is a function of PB kernel cut-off distance and beamlet resolution. It appears that if the highest accuracy is desired, a resolution of 10 mm or better in the direction parallel to MLC travel and a cutoff distance of 10 cm or better should be used. The calculation time increases with both cut-off distance and beamlet resolution. For example, GPU calculation time is 1.36 seconds for 5 cm cut-off distance, and increases to 4.84 s for 15 cm cut-off distance. A GPU accelerated PB dose calculation algorithm has been implemented using clinical measurement data. Excellent agreement with Pinnacle TPS has been achieved. Beamlet size and PB cut-off distance should be chosen according to desired dose calculation accuracy and speed.

SU‐E‐T‐194: Accounting for Support Arm Backscatter in EPID‐Based IMRT QA

Purpose: To devise a method for IMRT QA using a Pinnacle model for the EPID response which accounts for non‐uniform backscatter from the support arm. Method: The method is based on a Pinnacle treatment planning system model. It relies on determining a correction matrix (CM) that multiplies the (flood‐field calibrated) EPID image to restore pixel response to backscatter and subsequently constructing a backscatter region within the Pinnacle model that reproduces the effect of backscatter from the EPID support arm. The CM is constructed from a series of profiles calculated as the ratio of modelled to acquired slit‐beam fields. Each slit beam is offset so that the series covers the entire EPID area. Slit beams are used to minimize the dependence of response to backscatter. The CM represents the flood‐field response of the EPID mounted on the support arm as if it had been calibrated in the absence of backscatter. An iterative fitting procedure is used to adjust the backscatter region in the Pinnacle model so that the modelled flood‐field response matches the CM. Results: As expected, arbitrary‐field EPID images multiplied by the CM agree well with Pinnacle model results incorporating the backscatter region. The method was tested for 10, 20, and 30 cm square fields and shown to reduce discrepancies between the Pinnacle calculation and EPID images from 3.3, 8.3 and 9.4 %, to 0.8, 2.2 and 1.1 % respectively for the three fields. Conclusions: A novel method for IMRT QA was developed whereby acquired EPID images are multiplied by a pre‐computed correction matrix and then compared to the response from a Pinnacle model which incorporates a non‐uniform backscatter region. The method is easy to use in clinical practice and does not require measurements involving the removal of the EPID from the support arm.

Investigation of dose verification of esophageal carcinoma intensity modulated radiotherapy

Objective To compare the results of three dose verification solutions of esophageal carcinoma IMRT plans. Methods Seven esophageal carcinoma cases were planned with Pinnacle 8.0 h.The MATRIXX and Delta4 were chosen as the two-dimensional dosimetry and three-dimensional dosimetry.IMRT plans and Delta4 phantom plans were also recalculated by Monte Carlo. Gamma values were evaluated for MATRIXX and Delta4 with 3 mm/3% gamma criteria. For the comparison of Pinnacle, Delta4 and Monte gamma maps, the dose distribution in central plane, dose profiles and dose-volume histograms were used to evaluate the agreement. Results The gamma maps comparison show that with 3 mm/3% gamma criteria an over 98% pass ratio was obtained by MATRIXX measurement. A 94. 4% gamma pass ratio whicl.contains 4 fields gamma pass ratio lower than 90%, was obtained by Delta4 measurement. A 97.6% and 99. 8% gamma pass ratio was obtained between the Delta4 measurement and Monte Carlo simulation with 2 mm/2% and 3 mm/3% gamma criteria. The dose distribution in central plane and dose profiles from Pinnacle calculation were almost in agreement with both the Monte Carlo simulation and Delta4 measurement. The DVH plot have slightly differences between Pinnacle and Delta4 measurement as well as Pinnacle and Monte Carlo simulation, but have excellent agreement between Delta4 measurement and Monte Carlo simulation. Conclusions It was shown that all the three methods can be used very efficiently to verify esophageal carcinoma IMRT delivery, Delta4 and Monte Carlo simulation no data missed. The primary advantage of Delta4 is the fact it can measure true 3D dosimetry while Monte Carlo can simulate in patients CT images but not in phantom. Key words: Esophageal neoplasms/intensity modulated radiotherapy; Two-dimensional dosimetry verification; Three-dimensional dosimetry verification; Monte Carlo method

Monitor Unit Checking in Heterogeneous Stereotactic Body Radiotherapy Treatment Planning

Treatment of lung cancer using very-high-dose fractionation in small fields requires well-tested dose modeling, a method for density-averaging compound targets constructed from different parts of the breathing cycle, and monitor unit verification of the heterogeneity-corrected treatment plans. The quality and safety of each procedure are dependent on these factors. We have evaluated the dosimetry of our first 26 stereotactic body radiotherapy (SBRT) patients, including 260 treatment fields, planned with the Pinnacle treatment planning system. All targets were combined from full expiration and inspiration computed tomography scans and planned on the normal respiration scan with 6-MV photons. Combined GTVs (cGTVs) have been density-averaged in different ways for comparison of the effect on total monitor units. In addition, we have compared planned monitor units against hand calculations using 2 classic 1D correction methods: (1) effective attenuation and (2) ratio of Tissue-Maximum Ratios (TMRs) to determine the range of efficacy of simple verification methods over difficult-to-perform measurements. Different methods of density averaging for combined targets have been found to have minimal impact on total dose as evidenced by the range of total monitor units generated for each method. Nondensity-corrected treatment plans for the same fields were found to require about 8% more monitor units on average. Hand calculations, using the effective attenuation method were found to agree with Pinnacle calculations for nonproblematic fields to within ±10% for >95% of the fields tested. The ratio of TMRs method was found to be unacceptable. Reasonable choices for density-averaging of cGTVs using full inspiration/expiration scans should not strongly affect the planning dose. Verification of planned monitor units, as a check for problematic fields, can be done for 6-MV fields with simple 1D effective attenuation-corrected hand calculations.

SU-FF-T-193: An Investigation of the Accuracy of Esophageal IMRT Dose Distribution Using Three-Dimensional Dosimetry Techniques and Monte Carlo Simulation

Purpose: Complex dose delivery techniques like intensity-modulated radiation therapy (IMRT) require dose verification in three dimensions. This work investigates the accuracy of dose distribution of esophageal IMRT plans using three-dimensional measurement and Monte Carlo simulation. Method and Materials: Ten esophageal cases delivered by Varian 23EX were planned with Pinnacle 8.0. The plans were recalculated in the three-dimensional measurement phantom Delta4 with the structures of every patients and accurate plans parameters. The plans were also recalculated by Monte Carlo using leaf sequences and MUs for individual plans. All plans of Delta4 phantom were delivered and measured using Delta4. The dose distribution of iso slice, dose profiles, gamma maps and dose-volume histograms were used to evaluate the agreement. Results: The dose distribution of iso slice and dose profiles from Pinnacle calculation were in excellent agreement with both the Monte Carlo simulation and Delta4 measurement at all points except the edge of lung. Gamma maps comparison show that all three distributions mutually agreed to within a 3% (dose difference) and 3mm (distance-to-agreement) criteria. A 94.4% gamma pass ratio was obtained between the Delta4 measurement and Pinnacle distributions with 3mm/3% gamma criteria. A 97.6% gamma pass ratio was obtained between the Delta4 measurement and Monte Carlo simulation with 2mm/2% gamma criteria and 99.8% gamma pass ratio with 3mm/3% gamma criteria. The DVH plot have slightly differences between Pinnacle and Delta4 measurement as well as Pinnacle and Monte Carlo simulation, but have excellent agreement between Delta4 measurement and Monte Carlo simulation. Conclusions: It was shown that Delta4 and Monte Carlo simulation can be used very efficiently to verify esophageal IMRT delivery and no data is missed. The primary advantage of Delta4 is the fact it can measure true 3D dosimetry while Monte Carlo can simulate in patients CT images but not in phantom.

SU‐FF‐T‐363: An Investigation of the Accuracy of Head‐And‐Neck IMRT Dose Distribution Using Three‐Dimensional Dosimetry Techniques and Monte Carlo Simulation

Purpose: Complex dose delivery techniques like intensity‐modulated radiation therapy (IMRT) require dose verification in three dimensions. This work investigates the accuracy of dose distribution of head‐and‐neck IMRT plans using three‐dimensional measurement and Monte Carlo simulation. Method and Materials: Eleven head‐and‐neck cases delivered by Varian 23EX were planned with Pinnacle 8.0. The plans were recalculated in the three‐dimensional measurement phantom Delta4 with the structures of every patients and accurate plans parameters. The plans were also recalculated by Monte Carlo using leaf sequences and MUs for individual plans of every patients and Delta4 phantom. All plans of Delta4 phantom were delivered and measured using Delta4. The dose distribution of iso slice, dose profiles, gamma maps and dose‐volume histograms were used to evaluate the agreement. Results: The dose distribution of iso slice and dose profiles from Delta4 measurement were in excellent agreement with both the Monte Carlo simulation and the Pinnacle calculation at all points. Gamma maps comparison show that all three distributions mutually agreed to within a 3% (dose difference) and 3mm (distance‐to‐agreement) criteria. A 93.2% gamma pass ratio was obtained between the Delta4 measurement and Pinnacle distributions with 3mm/3% gamma criteria. A 96.8% gamma pass ratio was obtained between the Delta4 measurement and Monte Carlo simulation with 2mm/2% gamma criteria and 99.2% gamma pass ratio with 3mm/3% gamma criteria. The DVH plot have slightly differences between Pinnacle and Delta4 measurement as well as Pinnacle and Monte Carlo simulation, but have excellent agreement between Delta4 measurement and Monte Carlo simulation. Conclusions: It was shown that Delta4 and Monte Carlo simulation can be used very efficiently to verify head‐and‐neck IMRT delivery and no data is missed. The primary advantage of Delta4 is the fact it can measure true 3D dosimetry while Monte Carlo can simulate in patients CT images but not in phantom.

SU-GG-T-103: Minimum Segment Size for the Collapsed Cone Convolution Superposition Algorithm

Purpose: A comparison between film measurements,ion chambermeasurements and the collapsed cone convolution superposition (CCCS) algorithm for small field sizes prone to electronic disequilibrium is presented in this study. Method and materials: Using a Varian Clinac 2100 C/D with a Millennium 120 leaf multi‐leaf collimator(MLC),field sizes from 10×10 cm2 to 0.5×0.5 cm2 were created using both square fields and fields comprised of small rectangular segments using control points, for 6 MV and 18 MV photon beams. Using an ion chamber inside a solid water phantom, and Kodak EDR2 film these fields were measured. The monitor units were set to 200 for each field or segment. The results were compared against calculations using the CCCS algorithm in Pinnacle3. Results: Agreement, within 2%, between Pinnacle and measurements was observed for all open fields. For the fields using control points, the discrepancy between Pinnacle and measurements was in the order of 10–15% for most of the fields for both energies used. Ion chamber and film measurements agreed within 3% for the same fields throughout the range of the fields and energies used. Conclusion: Because the ion chamber, film and Pinnacle calculation agreed very well for the open square field sizes, it is unlikely that a setup error caused the unexpected results. At present, it can be concluded that the CCCS is not in good agreement (fails to predict accurately) the dose when the minimum segment has been reduced to an area with the smallest side smaller or equal to 1 cm. Further investigation will be carried out and more measurements will be taken to confirm the accuracy of the current data, and provide guidelines when small fields are needed for treatment.

WE-E-ValA-07: 3D Dose Reconstruction with Megavoltage Cone-Beam CT and EPID Exit Dosimetry

Purpose: The ability to reconstruct the delivered patient dose can help ensure that the integrated doses to important structures faithfully adhere to prescription. The objective of this work is to develop and test a 3D dose reconstruction procedure based on exit‐dose measurements at treatment time and MV cone‐beam CT.Methods and Materials: The proposed dose reconstruction method uses a Megavoltage cone‐beam computed tomography (MVCBCT) image acquired on the treatment table prior to treatment, 2D portal images taken with an amorphous‐silicon electronic portal imaging device(EPID) during treatment, and an independent validated dose calculation engine. The energy fluence obtained from the EPID is back‐projected through the 3D MVCBCT image. A dose calculation engine based on a collapsed‐cone convolution algorithm subsequently calculates the dose in each voxel. To test the model, a MVCBCT of a cylindrical solid‐water QA phantom was acquired and the MVCBCT numbers mapped to appropriate attenuation coefficients. The phantom was then treated with a 5cmx5cm beam and a portal image acquired. During the treatment, a CC13 ion chamber and MOSFETdetectors were used to measure the dose at 21 points to compare with reconstructed dose. A Pinnacle dose calculation using a conventional CT was also performed for comparison. Results: The mean difference between reconstructed and measured doses was −0.2% (standard deviation = 2.8%). The reconstructed dose in the inner regions of the cylinder differed less than 2% from the measured, although discrepancies of about 10% occurred at one point in the buildup region and at two other peripheral points. In comparison, the mean difference between Pinnacle calculations and measurements was −2.9% (standard deviation =1.6%). Conclusion: Preliminary calculations of reconstructed dose demonstrated good agreement with experiments. Further refinement of the model and its application to clinical conditions are under investigation. Conflict of Interest: Research supported by Siemens.

SU-FF-T-261: Image Guided High Definition Dosimetry of IMRT Plans Using the MobileMOSFET System

Purpose: To investigate the feasibility of using multiple MOSFETsensors, available with the mobileMOSFET® system as high definition dosimeter for IMRT plan verification. Methods and Material: Wireless dosimetry system (mobileMOSFET from Thomson Nielsen) was first tested for reproducibility, linearity, sensitivity, long‐term performance, and angular dependency. Twenty head and neck plans, each consisting of 8–10 treatment fields were verified. Plans were generated in Pinnacle 7.6C, and exported to a cylindrical solid water phantom. The in‐phantom dose distribution was calculated on the phantom CTimage set, and two regions of interest were selected: GTV area and avoiding area. The doses at these points were measured using MOSFETs and compared to ionization chamber measurements and Pinnacle calculations. Site specific fixed configurations of 5 MOSFET positions has been developed, and ten patient plans were verified using this module. CBCTimage guidance was used to accurately position the MOSFET phantom. Results: The response of MOSFET was found to be linear, reproducible (within ±2%), independent of angular positions (within ±2%) and stable with time. For 20 head and neck patients, average variations of (0.68±2.11)% at high dose point and (0.06±1.94)% at low dose point were observed between measured dose using MOSFET and ionization chamber. Average variations of (0.73±1.85)% and (0.96±2.00)% were observed between measured dose using MOSFET and plan dose at high and low dose points respectively. A (0.47±2.45)% variation was observed using the special insert for head and neck and prostate plans in four points out of five. Conclusions: These investigations indicate that the use of mobileMOSFET device with image guidance would be suitable for efficient and high‐resolution dose verification of complex IMRT plans.

SU‐FF‐T‐135: Complex IMRT Plan Verification Using a Commercial MU Calculation Package

Purpose: The general availability of IMRT treatments is constrained by the need to perform dosimetry measurements as part of the patient plan verification. While secondary IMRT MU calculators have been successfully used to validate some plans for a restricted number of sites, these calculators have suffered from unacceptably large uncertainties when applied to sites confined to the head and neck region. A strategy to reduce these calculational uncertainties has been developed which permits a greater use of IMRT MU calculators and reduces the need for dosimetric measurements, enabling a larger patient population to receive the benefits offered by IMRT treatments. Method and Materials: Segmental IMRT head and neck treatments were developed using the Pinnacle 6.2b inverse planning module. The IMRT treatment plans where then calculated on a CT image set of PMH IMRT phantom, and validation points corresponding to key target and avoidance regions where identified. The dosimetric data corresponding to these points was exported to RadCalc, a commercial IMRT MU calculator, and the calculations were compared to Pinnacle calculations and in‐phantom measurements. Results: A total of 10 clinical patient cases, each containing 7 to 9 gantry angles, were assessed. Agreement was assessed on basis of total dose delivered to the point of calculation. The measured and RadCalc calculated doses were found to agree within 3% at the high dose point and 5% at the low dose point in all cases, while the Pinnacle calculated dose was found to agree with the measured dose within 2.5% at the high dose point, with deviations as large as 13.5% observed at the low dose point. Conclusion: In‐phantom IMRT verification calculations of the total dose yields similar results as in‐phantom measurements. Consequently, a secondary MU calculator can be used to verify IMRT treatment plans and reduce the frequency of validation measurements.