Volume 35, Issue 6, June 2008
Index of content:
- Therapy Scientific Session: Auditorium A
- IMRT — Motion and Adaptation
WE‐E‐AUD A‐01: Controlling the Impact of Intensity Modulation When Treating Moving Targets with Dynamic IMRT35(2008); http://dx.doi.org/10.1118/1.2962767View Description Hide Description
Purpose: The treatment of moving targets using dynamic IMRT can result in dosimetric effects that have a predictable component (dose blurring) and an unpredictable component (interplay between organ motion and leaf motion). We investigate the impact of the intensity modulation on the unpredictable component. Method and Materials: Software was written to simulate MLCs moving across a moving target using dose distributions exported from Eclipse. The dose distribution delivered to a static target was blurred with the target motion to give the expected dose distribution. Dose errors were calculated by comparing the dose delivered to a moving target with the expected dose distribution. The impact of the intensity modulation (1:1 to 6:1) on the maximum dose error was investigated for a wide range of MLC sequences and target motion. MLC sequences were identified which kept dose variations within 10% of the expected dose. Results were confirmed experimentally by measuring the dose delivered to an ion chamber array in a moving phantom. Results: The maximum dose error increases with increasing intensity modulation. Dose errors also increase with a decrease in the following parameters: target period, target amplitude, MLC separation and MLC speed. Simple rules were developed to determine whether a planned fluence / MLC sequence will give a dose error larger than 10%. If the MLC speed is restricted to 0.1cm/sec and target period ⩽6sec, then the maximum daily dose error is always less than 10% for all levels of intensity modulation. For some combinations of intensity modulation and MLC sequences, faster MLC speeds are acceptable depending on details of the MLC parameters and target motion. Conclusion: Intensity modulation as high as 6:1 can be used when treating moving targets with dynamic IMRT, but the MLC speed must be carefully restricted if unpredictable and large dose errors are to be avoided.
35(2008); http://dx.doi.org/10.1118/1.2962769View Description Hide Description
Purpose:IMRT delivery follows planned leaf sequences, which are optimized before treatment delivery. Real‐time variations, such as respirations, are hardly to be modeled in planning procedure. We developed a real‐time Motion‐Adapted‐Optimization (MAO)‐guided delivery technique in TomoTherapySMtreatments. This technique models the radiation delivery with the real‐time motion as a negative feedback system. It updates the motion‐encoded cumulative dose and optimizes the leaf sequence in real‐time, right before the delivery of each projection. Method and Materials: TomoTherapySMtreatment delivery consists of thousands of projections with projection time around 200–500 ms. The leaf latency plus transition of TomoTherapy® binary MLC takes less than 50 ms. Real‐time MAO is to optimize leaf sequence of the coming projection right before its execution. It consists of several real‐time procedures including “motion detection and prediction”, “motion‐encoded dose accumulation” and “leaf sequence optimization” for the coming projection. To update leaf sequence in real‐time, all above procedures must be executed within 150 ms. We developed ultra‐fast algorithms and codes to approach such critical goal.
We implemented and tested this technique with the TomoTherapy® research system. The integrated system includes a real time camera system and a programmable motor‐driven phantom. We tested different TomoTherapySM plans with various simulated and real respiration traces. We used film dosimetry to verify and validate the final results. Results: MAO‐guided delivery runs smoothly in the integrated TomoTherapy® system. The whole MAO procedure takes less 100 ms per projection. Both simulated motion and real respiration of ∼2cm amplitude, the real‐time MAO‐guided delivery doses matched with the planning dose within 3% and 3mm criteria, for a typical TomoTherapySMtreatment configuration. No hot and cold spots are noticeable. Conclusion: We present a novel technique for real‐time MAO‐guided delivery within current TomoTherapy® hardware. Simulations and experiments conceptually proved this technique. Further validation and clinical implementation are underway.
WE‐E‐AUD A‐03: Evaluation of Dosimetric Margins in Prostate IMRT Treatment Plans Generated with Pinnacle DMPO35(2008); http://dx.doi.org/10.1118/1.2962770View Description Hide Description
Purpose: To quantify ‘dosimetric margins’ existing in prostate IMRT plans generated with Pinnacle Direct Machine Parameter Optimization (DMPO), due to imperfect conformance of the planned dose distribution to delineated targets. To demonstrate that when these margins are accounted for, the van Herk margin framework provides accurate target coverage estimates. Method and Materials: DMPO was performed on 27 prostate plans with 5mm PTV margins. Setup errors were simulated via fluence convolution and sampling from a systematic error distribution, and resulting dose volume histograms (DVHs) were used to generate van Herk style Dose Population Histograms (DPHs). The dosimetric margin distribution (DMD) is the 3D margin distribution between the clinical target volume (CTV) and the treated volume (TV). Consistent with ICRU guidelines, the TV was the volume enclosed by the planning target volume (PTV) minimum dose isodose surface. Due to imperfect conformance of the planned dose distribution, the DMD can extend beyond the PTV. This creates an additional ‘buffer zone’ around targets, which can absorb setup and other geometric errors. The DMD was measured by exporting CTV and TVs from Pinnacle as meshes, and measuring collision distances. Results: DPH analysis showed the DMPO plans could tolerate random plus systematic setup errors having standard deviations (SDs) up to 3mm. Naive application of the van Herk margin formula suggests that 5mm PTV margins should absorb errors with SDs up to only 1.6mm. Coverage calculations based on measured DMDs were in agreement with the DPH analysis: the presence of dosimetric margins allows the prostate plans to absorb larger errors than one would naively expect. Conclusion: These DMPO results agree with those previously obtained using an in‐house optimizer. Both optimizers create similarly large dosimetric margins. Accounting for these margins enables accurate coverage estimates, without the need for simulations. (Supported by NIH R01CA98524 and NIH P01CA116602).
WE‐E‐AUD A‐04: Motion Artifacts in Fast KVCT and Slow MVCT and Their Effect On Indirectly Derived Target Delineation in Adaptive Radiation Therapy35(2008); http://dx.doi.org/10.1118/1.2962771View Description Hide Description
Purpose:Livertumor cannot be directly delineated from MVCT, or kVCT without contrast. The external boundary of the organ in which the tumor resides is used for automatic segmentation of the PTV in adaptive radiation therapy (ART), and we study the effect of imaging speed on the accuracy of automatic segmentation and its dosimetric impact using a motion phantom. Method and Materials: A motion platform was programmed to move with a sinusoidal motion by amplitude of 1.9cm peak‐to‐peak and period of 3 seconds to simulate respiratory motion. A 15cm cubic water equivalent phantom was carried by the motion platform. 5 kVCT scans using a Philips Brilliance CT and 5 Tomotherapy MVCT scans were obtained. The external contour of the cube was manually and automatically delineated. A cylindrical PTV was created on the kVCT and a Tomotherapy was created. The contour of the PTV was generated based on the deformation of the cube in the MVCT and the effect on the DVH determined. Results: The volume of the contoured cube in the kVCT scans showed significant variation (±10%). However, the external contours derived from the MVCT scans were reproducible (+/− 1%), agreeing with previous reports. The external contour based on kVCT scans was larger than the MVCT in 3 cases and smaller in 2 cases. In the cases where the kVCT scan volume was larger than the MVCT scans, ART resulted in an underdosing of 10% of the PTV by 7% of the prescription dose.Conclusion: Without anatomical deformation or respiratory variation, the morphology of a solid moving phantom show significant differences between fast kVCT and slow MVCT, resulting in differences in automatic target delineation and subsequent dosimetric errors. The uncertainty needs to be taken into consideration for ART application.
35(2008); http://dx.doi.org/10.1118/1.2962772View Description Hide Description
Purpose: To demonstrate, both in phantom and patient, the feasibility of approximating the subject's density distribution via an average 4DCT image (AVG‐CT), calculate cumulative dose delivered during respiration with this technique, and evaluate the results with a full 4D dose summation. Method and Materials: A series of 4DCT numerical phantoms (9 phases, lungtumor excursions 2, 3, and 4 cm in S‐I direction) and their AVG‐CT images were generated. For full 4D dose summation, static dose was calculated on each phase, each dose matrix was sampled with known displacement, and dose was accumulated over all phases. Using the same clinical plan, the AVG‐CT cumulative dose was calculated by combining the static AVG‐CT dose with known tumor displacement, and assuming the dose distribution was the same for all phases. Four lungcancer cases were also evaluated for stereotactic body radiotherapy and conformal treatments. Here, deformable image registration was used to generate the patient‐specific motion model from 4DCT. Dose accumulation was analogous to phantom, however, each phase's dose matrix was sampled using the displacement vector field.Dose discrepancy (D) between full 4D summation and AVG‐CT approach was calculated and compared. Results: For all phantoms, AVG‐CT approximation yielded slightly higher cumulative doses compared to full 4D summation, with dose discrepancy increasing with increased tumor excursion. In vivo, using the AVG‐CT coupled with deformable registration yielded a modest increase in cumulative dose relative to full 4D dose summation and not at clinically applicable levels (D < 2%). Even for a patient with substantial tumor motion near the diaphragm, dose discrepancy was within 4%. Conclusion: Simplifying 4D dose accumulation via the AVG‐CT, while fully accounting for tumor deformation due to respiratory motion, has been validated, thereby introducing the potential to streamline the use of 4D dose calculations in clinical practice.
WE‐E‐AUD A‐06: Image and Dose Processing for Image Guided Adaptive Radiation Therapy and Outcome Research35(2008); http://dx.doi.org/10.1118/1.2962773View Description Hide Description
Purpose: To develop a general image processing and dose computation procedure that allows for the accumulation of accurate dose distributions for purposes of both outcome research and IG‐ART (image guided adaptive radiation therapy). We applied the procedure to cone‐beam CT head‐neck cancertreatment cases, as well as Tomotherapy MVCT (mega‐voltage CT) GYN and prostate cancertreatment cases. Method and Materials: The image processing procedures include preprocessing, rigid and deformable registration, mesh‐ based structure contour propagation, and image volume composition for dose re‐computation. The dose processing includes dose recomputation on the daily CT or the composed CT volume, deforming and register the daily dose to the planning dose space, dose accumulation, evaluating of the accumulated dose against the planned dose. Accuracy of image registration is very important for the entire procedure because it determines the accuracy of all later subsequent steps. In order to improve the accuracy, we preprocessed the CTimages before the registration step by using various methods, including edge‐preserving smoothing, Gaussian lowpass smoothing, contrast enhancement, window‐level intensity transformation, and bowel gas pocket painting for abdominal regions. We used MI (mutual information) based algorithm for rigid image registration, and applied the Horn‐Schunck optical flow algorithm for deformable image registration. We used a mesh‐based algorithm for structure contour deformation, interpolation and smoothing. Results: By applying these preprocessing procedures we were able to achieve improved image registration results. The computed image deformation fields were then used to propagate the structure contours from the kVCT to the current daily CTimage, which could be used for dose deformation or re‐computation. Conclusion: We have developed a procedure of image processing and dose computation to support the accumulation of ‘true’ dose distributions. These methods could be used in both radiotherapy outcome research and adaptive radiotherapy applications. Partially supported by NIH R01 grant CA85181.
WE‐E‐AUD A‐07: Should a Gaussian Probability Density Function Be Used to Approximate Respiration Induced Dosimetric Effects for Proton Radiotherapy?35(2008); http://dx.doi.org/10.1118/1.2962774View Description Hide Description
Purpose: To compare the dose distributions generated by convolving a static dose distribution using a patient specific respiratory probability density function (R‐PDF) with those generated using a generic Gaussian PDF (G‐PDF) for proton therapy of lungcancer.Method and Materials: The R‐PDFs were obtained by identifying the centroid motion of the targets from the 4D‐CT scans of a phantom (CIRS Model 008 Dynamic Thorax) and a representative lungcancer patient. The CMS XiO® Treatment Planning System commissioned with 208 MeV nominal proton beam data from a passive scattering beam line at a proton therapy center was used for the static dose calculation. The dose convolution results from four different G‐PDFs with standard deviations (SD) of 0.2, 0.3, 0.4, and 0.5 multiplying by the peak‐to‐peak motion amplitude (letter “A”, 1.60cm in the phantom and 1.75cm in the patient) were compared to the R‐PDF convolved dose distributions using a commercial dosimetry analysis package (OmniPro I'mRT). Results: Respiration‐induced dose error was 29% and 16% of the prescribed dose (PD) compared to the static doses in the phantom and patient, respectively. The G‐PDF with SD of 0.4A most closely approximates the R‐PDF whilst the maximum dose disagreements (MDDs) between the convolved doses using the two methods were 6% and 4% of the PD in phantom and patient, respectively. When G‐PDFs with SD of 0.2A and 0.5A were used to approximate the R‐PDF, the resulting MDDs were 19% and 12% in the phantom respectively, and 12% and 10% in the patient, respectively. When the G‐PDF with SD of 0.3A was used to approximate the R‐PDF, the resulting MDDs were 10% in the phantom and 8% in the patient. Conclusion: A Gaussian function should not be used to approximate a patient specific respiratory PDF since it can lead to clinically significant dose errors.
- Brachytherapy II
TH‐C‐AUD A‐02: The Air‐Kerma Strength Standard for 192Ir HDR Sources at the University of Wisconsin ADCL35(2008); http://dx.doi.org/10.1118/1.2962837View Description Hide Description
Purpose: To present a compilation of a seven‐year study of measurements using several HDR afterloaders and to provide a comprehensive analysis of the various published methodologies for interpolating between NIST standards to determine the air‐kerma calibration coefficient for . Ultimately an update of the current interim standard will be considered. Method and Materials: An acrylic apparatus for performing the seven‐distance measurement technique equipped with laser alignment was used to acquire all datasets. Measurements were performed with an Exradin A3 spherical ion chamber that had been calibrated at NIST for beam qualities of M250 as well as . A total of four different afterloaders were measured during multiple trials and a comparison of the results was made to assess any trends in measurements due to source geometry. Recently published interpolation methods were compared to the method used in the original establishment of the interim standard in 1991 with proper accounting of the revisions in the NIST air‐kerma standards in 2003. Three different methods for solving the non‐linear system of equations were compared to assess stability and minimize uncertainty. Results: Depending on the interpolation method, deviations of −1.12% to −0.37% from the long‐standing air‐kerma calibration factor were observed. In comparing the measurements from the last seven years, (2000–2007), to the well chamber transfer standards, (1991), certain trends between various source models were identified, but the overall effect was found to be in the range of — 0.95% to 0.18%. Conclusion: Based on the data recorded it is reasonable to assume, given the uncertainty in the method, that a single calibration factor would indeed be appropriate for all source models. The possibility of a formal update to the standard will be considered.
TH‐C‐AUD A‐03: A New Approach for Afterloading Brachytherapy Inverse Planned Dose Optimization Based On the Accurate Monte Carlo Method35(2008); http://dx.doi.org/10.1118/1.2962838View Description Hide Description
Purpose: In brachytherapy, considering the perturbations from the heterogeneities in the planning system will give a better dose conformity for specific sites. The goal of this work is to demonstrate the feasibility of replacing the TG43 analytical approach by a Monte Carlo (MC)dose calculation engine in the optimization process. Method and Materials: The novel method is based on pre‐computed 3D dose kernels. The CT clinical images and the dwell positions (DWP) are loaded from the DICOM‐RT files to create a voxel based simulation of the treatment. MCdose calculation is used to create the dose kernel specific for each possible DWP. Density and tissue compositions are fully taken into account in MC. The Inverse Planning Simulated Annealing (IPSA) algorithm is used for the optimization process. IPSA reads and analyzes the MCdose kernels before the beginning of the optimization process; it replaces the TG43 parameterization. A breast interstitial HDR plan is used to demonstrate the approach. Results: Computation of precise 3D‐kernels is the most time consuming portion and is proportional to the number of DWP. However, the optimization process itself takes the same amount of time as a standard (TG43) optimization. The breast, TG43/MC plan shows an underdosage in the CTV relative to the TG43/TG43 plan by 4.3 % on D90 and 3.2 % on D50. For the surgical bed, the difference is 4.2 % and 3.5 % for D90 and D50 respectively. This was corrected in the MC/MC plan, with a minimal dose increase of the skin.Conclusion:MCoptimization improved the dose conformity. The method presented is straightforward and can be applied to any site and any afterloading process (HDR or PDR) using various type of sources, from Ir‐192 to micro‐XRay devices, as long as a precise MC model is made.
35(2008); http://dx.doi.org/10.1118/1.2962839View Description Hide Description
Purpose: To perform calibration and evaluation of an image‐guidedbrachytherapyrobotic system. Method and Materials:Brachytherapyrobotcalibration consists of three steps: mechanical roboticcalibration,imagecalibration and overall system calibration. Mechanical calibration determines 1)system resolution, as the smallest incremental movement that the robot can physically perform, 2)repeatability as a measure of the ability of the robot to move back to the same position and orientation and 3)accuracy, as the robot's ability to precisely move to a desired position in 3D space. Imagingcalibration for our system is the procedure where distance from ultrasoundimages is transformed to the metric distance in robot absolute coordinates. Overall system calibration is to determine the exact position of image stacks, calculated in absolute robotic system coordinates. These steps were performed using high‐resolution camera, specially‐designed phantom box for imagingcalibration and CMM device for mechanical calibration. The system software allows mutual calibration between mechanical and imagingrobotic modules. Mechanical calibration consists of DHV table definition for robotic system, matrix transformation, definition of composite matrices, direct kinematics solution, inverse kinematics solution, definition of robot initial position, calculation of position error, and error correction method. The robot errors gathered by position measurement are minimized by numerical optimization. Results: The calibrated precision of translation movements for the stylet and cannula are in the range of 0.03–0.08mm (depending on load); lateral and vertical movements for the gantry are 0.03mm; probe translation and rotation are 0.05mm and 0.03deg, respectively. The fiducial error for imaging is less than 0.1mm in x and y image coordinates. Conclusion: Overall roboticbrachytherapycalibration plays a crucial role in accurate delivery. The calibrated precisions of the image‐guidedbrachytherapyrobotic system are considered satisfactory for the given clinical application.
Acknowledgement: Supported by NCI‐R01‐CA091763.
TH‐C‐AUD A‐05: Experimental Validation of An Iterative Forward Projection Matching Algorithm for Seeds Center Localization Using Conebeam‐CT X‐Ray Projections35(2008); http://dx.doi.org/10.1118/1.2962840View Description Hide Description
Purpose: To experimentally validate a new algorithm for reconstructing the 3D positions of implanted brachytherapy seeds from 2D projection images.Method and materials: The iterative forward projection matching (IFPM) algorithm consists of finding the 3D seed geometry that minimizes the sum‐of‐squared‐difference of the pixel‐by‐pixel intensities between computed projection images and measured auto‐segmented images of implanted seeds. IFPM starts with an approximation to the initial seeds configuration, e.g., the pre‐implant seed arrangement. It then iteratively refines the 3D seed coordinates until the computed projections match with the measured projections. Three pairs of computed and measured projection images, with known imaging geometry, are used. Two brachytherapy phantoms were fabricated with 12 and 72 seeds in known configurations. Three projections of each phantom were acquired using an Acuity digital simulator along with a full 660 projection Conebeam CT(CBCT).Image pre‐processing steps were performed to create the binary seed centroids images for use by IFPM algorithm. To quantify IFPM accuracy, the actual seed positions were extracted from the CBCTimages by the Brachy Vision‐planning system. Results: For the 12 seed phantom data, the mean reconstruction error was found to be 0.83±0.34mm where as for 72 seed phantom it was 0.97±0.37mm. The each test trials converged in 4–10 iterations with computation time of 2.8–62 min on a 2 GHz processor. Discussion: The IFPM algorithm avoids establishing seed projection correspondence required by standard back‐projection methods. In phantom studies we have demonstrated 1 mm accuracy in reconstructing the 3D positions of brachytherapy seeds from 2D projection images. This supports the potential of this algorithm for accurate and robust seed reconstruction in patients.
This project was supported by grant from Varian Medical Systems.
TH‐C‐AUD A‐06: Evaluation of 3 Tesla MR Image Distortion and Artifacts in a Titanium Applicator Presence: Toward 3T MRI Guided HDR Brachytherapy for Cervical Cancer35(2008); http://dx.doi.org/10.1118/1.2962841View Description Hide Description
Purpose: Characterize 3 Tesla (T) magnetic resonance image(MRI) distortion and artifacts induced from a titanium applicator presence. Method and Materials: Based on the ASTM International method, a titanium tandem and ovoids (Varian) was placed in a reference phantom, and embedded in a solution (30L distilled water, 1.5g/L CuSO4). A reference phantom was designed to be free from distortion, to suspend an applicator, and to provide a reference for distortion. MRimages were scanned for transverse, sagittal, and coronal views; and also generated both with and without the applicator in place. Image artifact and artifact width were quantified for all three datasets to determine maximum width. For the purpose of this study we used three tandems to simulate an applicator. Two different gels (both water‐soluble) were tested around tandems: lubricating jelly for ultrasoundimage and white petrolatum gel.Results: Image artifacts were evaluated for pixels changing their intensity by ⩾ 30% and found at mainly three regions; the tip (its artifact width ⩽ 4mm) of tandem and the shoulder region (⩽ 5mm) of tandem and the triangular area (its image artifacts area ⩾ 0.8cm2) surrounded by the three tandems. A shoulder region is located inferior‐outside of uterus and a triangular region also represents the gauze‐packed space in the vagina. Hence, their impact on tumor delineation is minimal. At the tip of tandem, the artifacts width (4mm) potentially leads to limiting microscopic tumor delineation but is within the tolerance (5mm) of MRimages registration (AAPM TG53). The distortion was determined to be no more than 1.2mm. The gels described above were found to be helpful in determining the boundary but not in improving artifacts. Conclusion: Artifacts and distortion from a titanium applicator presence were found within the tolerance. 3T MRimage is feasible to be implemented into brachytherapy planning process.
TH‐C‐AUD A‐07: Evaluation of the Correction Factor Due to the Lack of Full Scatter Conditions in Cs‐137 and Ir‐192 Brachytherapy Dosimetric Studies35(2008); http://dx.doi.org/10.1118/1.2962842View Description Hide Description
Purpose: Use of a finite phantom to derive dose rate distributions around brachytherapysources implies a lack of backscattering material near the phantom periphery. Conventional planning algorithms and newly‐developed 3D correction algorithms are based on physics data under full scatter conditions. Presently, most published Monte Carlodosimetric studies have been obtained using either a spherical phantom (15cm in radius) or a cylinder phantom (40×40cm2). The study objective was to derive a simple relationship to correlate the radial dose function, g(r), obtained for each one of these phantoms to that obtained for an unbounded phantom. Method and Materials: Assuming bare point sources of and , kerma was calculated using Monte Carlo GEANT4 code for 1) a spherical phantom of 40 cm in radius, R, which is assumed an unbounded phantom for r ⩽ 20cm, and 2) spherical phantoms of R=15cm and R=21cm. The later size mimics the scatter conditions of a 40×40cm2 cylindrical phantom for both radionuclides. From the ratio of the dose rate distributions for unbounded/bounded phantoms we derived the relationship between g(r) for both phantoms. Results: Phantom size correction results to g(r) were obtained and fit to 3rd order polynomials (R2 > 0.999) valid for r ⩽ 10cm, which is the clinical range of interest. To validate the method, published dose‐rate distributions for two and sources in bounded/unbounded phantoms were compared with the fits of this study. Agreement was typically within 0.2% over all distances studied. Conclusion: In order to compare the dose rate distributions published for different phantom sizes, a simple expression based on fits of the dose distribution ratios for bounded/unbounded phantoms was developed for and . Using these relations, it was possible to correlate g(r) between bounded and unbounded phantoms for improved accuracy and consistency of clinical dosimetry.
35(2008); http://dx.doi.org/10.1118/1.2962843View Description Hide Description
Purpose: For high‐energy photon‐emitting brachytherapysources such as , , , and , the main contribution of the systematic uncertainty in the dose distributions near the sources is understanding of electronic equilibrium and the contribution of β‐rays due to radioactive disintegration. Thus, it is important to study these effects in detail to accurately depict dose distributions near these brachytherapysources. This work studies the relative importance of β‐ray contributions to total dose (β + γ + x‐ray), and feasibility of using the approximation “collision kerma equals dose in electronic equilibrium conditions.” Method and Materials:Characteristics of kerma and dose distributions were studied for spherical , , and sources with composition, encapsulation, and dimensions similar to those existing in the literature. Dose contribution of β‐rays and γ+x‐rays were individually examined using the GEANT4 Monte Carlo radiation transport code. Results: The comparison of kerma and dose rate distributions indicate ∼ 20% electronic disequilibrium within 1 mm of sources, with have the most pronounced effect. When examining the dose contribution of β‐rays, again had the most pronounced effect out to 5 mm beyond the capsule, with β‐rays contributions for and 1.5 and 0.5 mm beyond the capsule, respectively. Conclusion: The dosimetriceffect of β‐rays for high‐energy photon‐emitting radionuclides and the influence of the electronic disequilibrium near of the sources were studied using Monte Carlo methods. For and brachytherapysources, electronic disequilibrium has an important role near the source. For the main perturbing dosimetriceffect near the source is from β‐ray contributions and not electronic disequilibrium.
35(2008); http://dx.doi.org/10.1118/1.2962844View Description Hide Description
Purpose: To model the geometry of eye plaque therapy using MCNP5 and study the dose distribution at specific points in the eye. Effects of tissue heterogeneity and coverage for small tumors (<10mm) which have poor clinical outcomes are concentrated. Method and Materials: Geometry of the eye plaque for eye plaque therapy is modeled using MCNP5. The dose at the tumor base and apex has been calculated and compared against published data. For this work, a simple geometry of the eye plaque has been studied. The eye diameter, tumor base diameter, plaque diameter, plaque thickness and depth are chosen to be in compliance with the COMS standards. This model uses 14 radioactive NASI MED 3631 A/M mode I‐125 seeds. The source energy distribution for this problem defines only the photon energies and not the beta energies as they are absorbed by the titanium capsule in the I‐125 seed. Results: Dose at the tumor apex was calculated using *F8 and F6 tallies in MCNP5. Results show that F6 tally gives more accurate results than *F8 for this problem. So, F6 tally was used further in this research. Dose at tumor base and depth dose profile were studied for this problem using F6 tally. 54540 source particles were simulated using MCNP5 and the resulting particle trace was obtained. Conclusion: The geometry of the eye plaque has been modeled using MCNP5 and the dose distribution at specific points of interest has been studied. The MCNP5 results were found to be agreeable with the published data for several treatment techniques. The effects of tissue/tumor heterogeneity have been studied and the reason for poor tumor control / toxicity outcomes for small tumors is found using Monte carlo dose calculation due to lack of coverage which could not be realized with 2D/3D planning.
35(2008); http://dx.doi.org/10.1118/1.2962845View Description Hide Description
Purpose To study the Monte Carlo codes PTRAN_CT and MCNPX2.5 and compare the dosimetry results with the TG‐43 formalism for a permanent seed implant of a breast brachytherapy.Method and Materials: The geometry validation of a model 6711 iodine seed was studied calculating the radial dose function in water and the energy spectrum of the seed in air. The results for the calculated spectrum were compared with an experiment carried out with an Amptek XR‐100T spectrometer. The calculation times for MCNPX and PTRAN_CT were analyzed by calculating the figure of merit for water phantoms with different voxel numbers. The absolute dose was validated comparing the absolute dose in water with TG‐43 and the absolute isodoses obtained from EBT Gafchromic film. The results of a treatment plan for a breast brachytherapy were compared with TG‐43 calculations. Results: The discrepancy between the calculated and published radial dose function values is less than 3% for the two MC codes. The comparison of the energy spectrum with the experiment reveals a contribution of the detector diode for the energy inferior to 5 keV. The calculation time comparison between MCNPX and PTRAN_CT shows that PTRAN_CT is 10%–30% faster than MCNPX for a voxel number between 200,000 and 500,000. A good agreement is obtained for the absolute dose calculated by the two MC codes compared to the TG‐43 calculation in water and the absolute isodoses measured in the Gafchromic film. The breast cancer patient plan shows that the MC results differ 12.7% in comparison with the TG‐43 results. Conclusion:MCNPX and PTRAN_CT simulations agree with the absolute dose in water obtained with TG‐43 and experiments. Moreover, the patient dosimetry study reveals the interest to use a MC code where the tissue composition and the interseed attenuation are taken into account.
- Advanced Techniques and Risk Assessment
35(2008); http://dx.doi.org/10.1118/1.2962893View Description Hide Description
Purpose: To identify new ways of increasing the proton energy from a laser driven accelerator without increasing the laser power. To design an optimized interaction geometry regarding target positioning and laser energy delivery, that will result in more energetic protons.Method and Materials: Fully relativistic 2D3V particle‐in‐cell (PIC) simulations are used in this study. The initial conditions are chosen to correspond to a real experiment with a 40fs laser pulse (λ=800nm) and energy in the pulse in the range 4J–7J, focused to 2.8–3.4μm. The loading of the laser pulse in the simulation is adequately controlled in order to study the influence of angle of incidence and wave‐front curvature on the proton acceleration. The target is Cu with thickness of 400nm and width of 10μm. A thin proton layer attached to its back surface. Results: When the laser beam is focused on the target optimum incidence angle is found at ∼30° angle for a 21% gain in proton energy compared to normal incidence. When the laser pulse is split in two and both sub‐pulses are focused on the target at opposite angles (+30° and − 30°) the proton energy is increased and reaches maximum for equal splitting for a total energy gain of 42%. Positioning the target exactly one Rayleigh range behind the beam's waist is found to be beneficial as well. This position of the target corresponds to a wave‐front with the lowest positive radius of curvature. Conclusion: The combined optimization of angle of incidence, pulse splitting, and wave‐front curvature leads to energy gain between 65% and 140% compared to normal incidence for two realistic experimental situations. Increasing the proton energy by such a significant amount without increasing the energy in the laser pulse can prove to be the way to reach the therapeutic range of proton energies.
TH‐D‐AUD A‐02: Measurement of Neutron Spectrum and Ambient Dose Equivalent Around a Mini‐Phantom at a Proton Therapy Facility35(2008); http://dx.doi.org/10.1118/1.2962894View Description Hide Description
Purpose: To determine neutron spectra and ambient dose equivalents (H*(10)) for out‐of‐field and in‐field‐out‐of‐range locations around a mini‐phantom irradiated by proton beams. Method and Materials: A dual‐activation foil‐based Bonner sphere (BS) and BS extension (BSE) system was used to determine the neutron spectral fluence in conditions typical for treatment of pediatric patients with proton beams. Proton beams with nominal energies of 120 MeV and 180 MeV were modulated to generate in water 5.0‐cm Spread‐Out‐Bragg‐Peak (SOBP)/5.5‐cm range and 15.0‐cm SOBP/15.5 cm range, respectively. Brass apertures were placed in the large snout at the end of the treatment nozzle to project a 5×5 cm2 field at isocenter. Lucite blocks with a cross‐section of 6×6 cm2 and thicknesses of 5.7 cm and 14.7 cm were used for the 120‐ and 180‐MeV beam irradiations, respectively. Neutron H*(10) was calculated using the fluence‐to‐ambient dose equivalent coefficients from ICRU report 57. Results: The H*(10) was determined at three locations around the mini‐phantom: 25 cm from the isocenter perpendicular to the beam axis (L1), 25 cm from the isocenter along the axis downstream (L3), and 35.4 cm from the isocenter along 45° downstream (L2). The H7*(10) for the 180‐MeV irradiation were 6.89, 4.07, and 4.60 mSv/Gy at L1, L2, and L3, respectively, and was 6.18 mSv/Gy at L1 when the mini‐phantom was removed. For the 120‐MeV irradiation, the H*(10) were 1.21, 0.774, and 0.919 mSv/Gy at L1, L2, and L3, respectively, and was 1.06 mSv/Gy at L1when the mini‐phantom was removed. Conclusion: The neutron spectrum near the isocenter has a two‐peak structure, with peaks near 1 MeV for both energies and a peak near 110 MeV for the 180‐MeV and near 80 MeV for the 120‐MeV protonirradiations.Neutrons below 100 keV contribute less than 2% of the ambient dose equivalent.
35(2008); http://dx.doi.org/10.1118/1.2962895View Description Hide Description
Purpose: The purpose of this work was to determine which components along the central axis of a passive beam delivery system for proton therapy contributes the most to the production of secondary neutrons.Method and Materials: In this work a passive beam delivery system was modeled based on the MD Anderson Cancer Center treatment nozzle. We performed Monte Carlo simulations with Los Alamos code MCNPX. In these simulations a 200 MeV proton beam is shaped by a rotational modulator wheel (RMW), a secondary scatterer and by a collimating system including a variable snout. Cylindrical volumes were placed along the beam central axis to determine the radial distribution of the neutrons produced. The volumes were made of concentric cylinders with radius ranging from 50 mm to 20 mm. The volumes were placed after the RMW, the secondary scatterer, before and after the snout. The neutron flux and energy spectra were determined for each volume radii and for three treatment volumes. Results: After the RMW the neutron flux was higher for all treatment volumes diminishing as the distance along the central axis increased. The flux increased slightly just before the final snout for the smaller field sizes indicating a backscatter contribution as the proton beam is finally collimated. At the end of the nozzle the flux was lower than after the RMW. The larger neutron flux with energies ranging from 130 MeV to almost 200 MeV was found at smaller radii. As the radial distance increased the flux of energetic neutrons diminished. Conclusion: We found that the RMW was the major source of neutrons in the treatment nozzle. The flux diminished as the distance increased indicating a 1/r2 dependency. The other shaping components contribute to the neutron production but it is difficult to differentiate between contributors after the RMW.
35(2008); http://dx.doi.org/10.1118/1.2962897View Description Hide Description
Purpose: To characterize neutron dose equivalent per proton Gray (H/D) as a function of proton energy, spread‐out Bragg peak (SOBP) and aperture size at a clinical proton treatment facility using both active and passive neutron detection instruments, and to evaluate the performance of the instruments for a possible candidate for routine patient neutron exposure monitoring. Method and Materials: The H/D were measured at 2 locations near isocenter, namely in proton field but 2 cm out of range and approximately 20 cm out of field, with neutron instruments of various types. Two types of rem‐meters were used, the REM‐500, a tissue equivalent proportional counter and SWENDI‐2, a wide energy (up to 5GeV) neutron probe. The passive detectors consist of Neutrak® 144 Fast, Intermediate and Thermal Neutron Dosimeters, an etched‐track detector, and two flavors of bubble detector, BDS®, a spectrometer, and BD‐PND®. The results from these instruments were compared for 200 MeV and 160 MeV beam configurations. REM‐500 and Neutrak® were used to study the dependency of H/D on proton energy, SOBP and aperture size. Results: Various instruments had comparable H/D measurements at the out field location but varies more at the in field location. The H/D value increased linearly with proton energy, with a factor of ∼6 from 100 MeV to 250 MeV for medium snout. The H/D value increased ∼50% with increasing SOBP from 2 cm to the maximum of clinical allowed value and had a small decrease with aperture size. Conclusion: The measured results show H/D increases strongly with proton energy, moderately with SOBP, and has minimal dependence on aperture size. This is consistent with the simulation results. The Neutrak® etched‐track detector is a good candidate for patient neutron exposure monitoring due to its small size, excellent dose range, good energy range and commercial support.