Index of content:
Volume 33, Issue 6, June 2006
- Therapy Scientific Session: Room 224 C
- Brachytherapy III
33(2006); http://dx.doi.org/10.1118/1.2241835View Description Hide Description
Purpose: A new model of palladium‐103 (Pd‐103) seeds with plastic encapsulation (OptiSeed103 Model 1032P, International Brachytherapy, Norcross, Georgia) is commercially available. The abstract reports the investigation of using the plastic Pd‐103 seeds in prostate brachytherapy.Method and Materials: The RTOG criteria were adopted in patient selection. A real‐time transrectal‐ultrasound (TRUS) guided transperineal implant technique was employed. A Foley catheter was used to localize the urethra. The seed placement followed a modified uniform loading protocol. At the completion of an implant, TRUSimages were recorded. Post CTdosimetry was accomplished in the patient's 30‐day follow‐up visit and a seed migration survey was also performed. Results: Three prostate cancer patients underwent the Pd‐103 seed implantation. 386 loose seeds with total activity of 656.1 U were implanted. The patients tolerated the procedure. No abnormal symptoms have been reported in the first half‐year follow‐up, except the detection of 3 migrated seeds to the lungs in a patient. The seed appearances in the pelvic radiographs and CTimages were similar to that of the metallic seeds. A significant seed enhancement was observed in the post‐implant TRUSimages. Phantom studies showed similar results. A post plan for each patient was generated on the post‐implant TRUSimages where 97% of the seeds were identified with confidence. The patterns of the seed distribution in the TRUS and CTimages were similar although the GTVs were different due to implant edema. Conclusion: Regarding the dosimetry and implant process, there are no differences between the plastic and metallic seeds. A superior property of the plastic seeds is its appearance in TRUSimages. The seed distribution can be immediately displayed in the TRUSimages at the completion of an implant. Cold spots can be identified in real time and additional seeds may be added so as to improve the implant quality.
33(2006); http://dx.doi.org/10.1118/1.2241836View Description Hide Description
Purpose: To evaluate and optimize the performance of a prototype robotic manipulator designed for prostate seed implantations in a gel phantom. Methods and Materials: The prototype design is a custom‐built six degree‐of‐ freedom robot engineered for highly accurate prostate seed implantations. The robot allows for a fully automatic insertion or a manually assisted implantation using the device as a guide. The accuracy was determined using a gel phantom with a holed‐template placed at 10 cm depth. Insertion speed and rotation were varied, and the distance from the target hole was measured. The effect on the gel due to needle rotation was observed for increased damage. The experiment included both manual and automatic insertions using 30 cm, 17‐gauge, beveled‐tip prostate brachytherapy needles. Results: The optimal techniques were automatic insertions with a rotation speed of 10 rev/s or 1 rev/s. This provided a “hit” of the target 67% of the insertions with maximum displacement in the “non‐hit” insertions of 2.1 mm. The least accurate technique utilized a non‐rotated needle, which deflects towards its beveled‐tip as inserted. The total displacement from the target with this technique reached distances of greater than 8 mm for a 10 cm depth insertion with an average of 6.9 ± 0.8 mm. Conclusion: Preliminary results yielded an optimization of insertion parameters which increased the accuracy from near 1 cm to sub‐millimeter displacements. The robot was determined to have a high accuracy in tip placement within the gel. Rotation of the needle can dramatically increase the accuracy of the tip's final position. Rotating the needle at 1 rev/s yielded an accurate implant with minimal increase in gel damage. The robot used in conjunction with an optimized needle insertion technique benefits the patient with increased accuracy, leading to a more successful outcome and reduced complications.
33(2006); http://dx.doi.org/10.1118/1.2241837View Description Hide Description
Purpose: To present preliminary results for a robotically‐assisted prostate brachytherapy treatment system. Method and Materials: A 4 degree‐of‐freedom (DOF) robotic device was developed to replace the ultrasound template in a commercially available prostate brachytherapy treatment system (Interplant, Computerized Medical Systems, St. Louis, MO). The robot mounts to the existing template mounting points on the ultrasound stepper, and is capable of positioning the needles at arbitrary positions and orientations. The robot is spatially co‐registered to the Interplant treatment planningsoftware through a calibration procedure. We performed a seed‐implantation experiment on a prostate training phantom in which the needles were positioned by the robot. The needles were preloaded with one seed, then inserted manually by the operator under transrectal ultrasound(TRUS) guidance. In this experiment, ten seeds were implanted, and their implanted positions were reconstructed using data from a post‐implant CT of the phantom and the Interplant post‐implant analysissoftware (iPAS). Results: Using the results from the iPAS software, we measured both the relative error of each seed (with respect to the other seeds), and the absolute error of each seed with respect to the treatment plan. The relative root‐mean‐ square (RMS) transverse error was 0.8 mm (worst case 2.1 mm, 70% under 0.7mm), and the relative RMS sagittal error was 2.5 mm (worst case 4mm, 60% under 2.5 mm). The absolute transverse RMS error was 2.4 mm (worst case 4.3 mm, 50% under 2.4 mm), and the absolute sagittal RMS error was 2.5 mm (worst case 4.5 mm, 80% under 2.5 mm). However, the absolute transverse errors were characterized by an offset in each direction, most likely resulting from errors in the measurement of the robot position relative to the phantom. Conclusion: Our system for robotically‐assisted prostate brachytherapy shows potential for improved needle placement, repeatability, and accuracy.
TH‐B‐224C‐04: Feasibility of Calibrating Elongated Brachytherapy Sources Using a Well Type Ionization Chamber33(2006); http://dx.doi.org/10.1118/1.2241838View Description Hide Description
Purpose: TG‐43 recommended parameters of a brachytherapy source require calibration of the source using the WAFAC system by the NIST. However, the presently available NIST standard system is limited for calibration of sources with active lengths ⩽1 cm. A new procedure has been introduced for calibration of elongated brachytherapy sources (i.e. active lengths > 1 cm) as an interim solution to the calibration standards by the NIST. This procedure is based on commercially available well type ionization chambers.Materials and Methods: The variation of the source calibrator response as a function of the source position along the longitudinal axis of the chamber was measured to determine the relative correction factor (RCF). Then the NIST calibrated source was used to calibrate the response of the source calibrator. A train of the 1 cm source segments were used to create elongated sources with different active‐lengths ranging from 1 cm to 7 cm. The measured air kerma strength were compared with the total source strengths calculated as the sum of the individual 1 cm source segments utilized to compose the source. Results: The results of these investigation have indicated that the response of the Capintec CRC was nearly constant (within 0.5 %) for distances ranging from 4.9 cm to 14.9 cm from the bottom of the chamber. With this information and a calibration factor of 4.881 Reading/U, air kerma strengths of RadioCoil™ Pd‐103 sources with active length of 1 to 7 cm were measured. The results of these measurements were found to be within ± 0.4% of the values calculated by addition of the 1cm source segments used in creating the sources. Conclusion: A well type chamber with the calibration for the 1 cm long RadioCoil™ 103Pd source segment can be utilized for calibrating of an elongated source with different active lengths sources.
- Clinical Measurements I
33(2006); http://dx.doi.org/10.1118/1.2241536View Description Hide Description
Purpose: Liquid ionization chambers (LICs) have characteristics that can remedy some of the drawbacks of air‐filled ionization chamberdosimetry: large sensitive volumes (i.e. low spatial resolution), fluence perturbations, and energy dependence over the clinical range of beam qualities. However, there are significant problems in liquid chambers. High ionization density and low ion mobility lead to high ion recombination rates. In this work, extensive experimental work has been performed to investigate properties of a new liquid chamber. This includes chamber stability over time, chamber reproducibility, and establishing recombination corrections. Methods and Materials: The new chamber is called the GLIC‐03 (Guarded Liquid Ionization Chamber). The diameter of the collecting electrode is 1.5 mm and the plate separation is 0.4 mm, giving a sensitive volume of 0.7 mm3. The dielectric liquid used is isooctane. We used the 18 MV beam of a Varian Clinac 21EX linear accelerator. The lowest pulse rate setting, 100 MU/min, was used in order to avoid incomplete collection of charge from one pulse before the arrival of the next. Measurements were taken in solid water at 15 cm depth, with various field sizes and SSDs. Boag's theory for general collection efficiency for parallel‐plate gas ionization chambers, applied to isooctane in pulsed radiation, was used for recombination corrections. Results: The GLIC‐03 response varied by less than 1% over 10 hours, and was reproducible within 1.5% of the mean over different liquid fills. The collection efficiency decreased with increasing dose per pulse due to general recombination of ions from a larger number of ionizing particle tracks. Recombination corrections were within 1% for low dose rates and high electric field strengths. Conclusion: The establishment of these characteristics in the present work allows us to perform accurate relative measurements in high gradient non‐equilibrium fields as well as energy dependence investigations.
33(2006); http://dx.doi.org/10.1118/1.2241537View Description Hide Description
Purpose: At present, clinical dosimeters are limited to point or planar measurement, and hence do not provide the comprehensive 3D information ideal for verification of advanced delivery techniques. In this work we present a clinically viable 3D dosimetry system comprising a PRESAGE™ dosimeter with read out by an optical‐CT scanner.Method and Materials: A novel solid dosimeter called PRESAGE™ has been developed which is composed of polyurethane polymer and radiochromic leuco dyes. PRESAGE™ exhibits a stable color change and hence optical density (OD) change when exposed to ionizing radiation. A PRESAGE™ cylinder of 16cm diameter × 11cm height was taken through the treatment planning process and a 5‐field 6MV conformal radiation treatment was delivered by a Varian® linear accelerator. The radiation induced OD change was imaged in 3D by an optical‐CT scanner and this measured distribution was then compared with the corresponding dose distribution calculated by the treatment planning system, as well as to independent measurement by GAFCHROMIC® film. Intercomparisons between the three dose distributions were made by superimposing isodose lines and calculating gamma maps (with criteria 4% dose difference and 4mm distance to agreement). Results: Given stable temperature and protection from exposure to incandescent light, the dose response of PRESAGE™ was observed to be robust to all aspects of the lab. The 3D dose distribution measured in PRESAGE™ showed good agreement with the calculated treatment plan (Eclipse) as well as the independent film measurement at all percent doses >30% (i.e. in regions further than 1cm from the wall). Gamma comparison shows that the PRESAGE™ measurement agrees with both the calculation in treatment plan and the film measurement within 4% dose difference and 4mm distance to agreement. Conclusion: This work presents the PRESAGE™/optical‐CT combination as a practical 3D dosimetry system which can provide comprehensive quality assurance of advanced treatment techniques.
33(2006); http://dx.doi.org/10.1118/1.2241538View Description Hide Description
Purpose: To characterize the behavior of MOSFETs under radiation of various clinically relevant energies used in radiotherapy and radiology by evaluating its sensitivity or threshold voltage shift (CF) with regard to total integrated dose.Method and Materials: Seven p‐type, duals bias MOSFETs from Thomson & Nielsen were investigated. They were exposed to four radiationssources: (1) 60Co unit (〈E〉γ: 1.25 MeV), (2) 192Ir HDR unit (〈E〉γ;: 0.38 MeV), 30 kV beam (〈E〉γ: 14.8 keV) and (4) 150 kV beam (〈E〉γ: 70.1 keV). The MOSFET's sensitivity (CFw) was evaluated at various moments in time and was calculated as the ratio of the measurement Mw (mV) over the estimated dose value Dw (cGy) both in water. Results: The sensitivity of MOSFET is express by their calibration factor (CFw), and allows the user to associate the reading displayed by the device (mV) to a dose value (cGy). The CFw value diminishes with increasing threshold voltage, especially for low energy radiation. It is stable for 60Co irradiations, while it decreases of 6%, 5% and 15% for beam energies of 192Ir, 150 kV and 30 kV respectively. This behavior is explained by an alteration of the effective field applied on the MOSFET (bias), caused by the accumulation of holes at the SiO2 interface. It is strongly dependent on the radiation nature (LET) and particularly affects low x‐ray energies. Conclusions: Those results are of major interest since, following the company recommendations to calibrate the device every 7 000 mV, it could lead to a significantly underestimated dose. A calibration of the device before every use and performing more than one measurement (thus using a mean dose value) should compensate the observed behavior.
33(2006); http://dx.doi.org/10.1118/1.2241539View Description Hide Description
Purpose: To determine if a CR system that had been used for radiation therapy digital imaging since 2001 could also be used for IMRTquality assurance, and if the system could be used in helical tomotherapy quality assurance.Methods and Materials: The CR system used consisted of a desktop CR reader that utilizes storage phosphor plates, a 650 nm laser diode scanning beam source, and a high luminance light box for plate erasure. The CR plates are made of phosphor, coated with a photostimulable storage phosphor (BaFBR:Eu2+). Three types of dose‐to‐responsecalibrations were performed 1.) Static square fields; 2.) An IMRT step‐wedge; and 3.) A rotational helical tomotherapy delivery with concentric rings of known dose. All readings were taken with 6 MV beams. Like TLDs, some of the trapped charge carriers in the storage phosphor gradually decay with time. Because of the decay effect, it was important to determine the best time to wait between exposure and scanning. Five helical tomotherapy patients were selected as test cases for the CRdosimetry. Measurements were made in a cylindrical phantom with the CR plate and again with radiographic film. Calibration techniques #1, #2, and #3 were applied to the CRimages to determine which was the most appropriate. Dose differences and gamma comparisons were made between the calculated and measured doses.Results: A time and field size dependence was observed. After comparing readings from different time intervals, ranging from one to twenty minutes, it was decided that four minutes was an optimal time to wait between exposure and scanning. Also, gamma for the CRimages was significantly worse than the film images taken for the same patient. Conclusions: The field‐size dependences, inconsistencies between calibration techniques, and plate decay make the CR system used in this study non‐usable for IMRTdosimetry.
TU‐C‐224C‐05: Energy Response of LiF:Mg,Ti Thermoluminescent Dosimeters to Moderately Filtered X‐Ray Spectra in the Range of 20 to 250 KV Relative to 60Co33(2006); http://dx.doi.org/10.1118/1.2241540View Description Hide Description
Purpose: To use experimental methods to determine the response of LiF:Mg,Ti thermoluminescent dosimeters(TLDs) irradiated using moderately filtered (M‐series) x‐ray spectra in the energy range of 20 to 250 kV relative to the response to 60Co photons. Also, to determine if LiF:Mg,Ti TLDs are intrinsically linear detectors (i.e. the response is proportional to energy imparted). Method and Materials:TLDs were irradiated to a known air kerma using the NIST traceable M‐series x‐ray beams, which were located at an Accredited DosimetryCalibration Laboratory (ADCL), in the range of 20 to 250 kV. Using each x‐ray beam, several sets of TLDs were irradiated to different air kerma levels to take into account any dose non‐linearity. TLD response was then compared to that from several sets of TLDs irradiated at corresponding air kerma levels using 60Co. The Monte Carlo code MCNP5 was used to correct for scatter from the holder and to determine the predicted/expected TLD response to the experimentally used x‐ray beams. Results: The measured TLD energy response compared to the response to 60Co shows a rapid decrease toward very low photon energies. This response dropped to approximately 0.90 at the lowest effective energy of 11.5 keV. The highest response was found to be 1.37 at an effective energy of 28.5 keV. The results showed poor agreement between measured energy response and calculations using the mass‐energy absorption coefficients of pure LiF. A significant increase in measured response compared to calculated response was seen at effective energies higher then 25 keV. Conclusion: These results demonstrate that the measured energy response differs by up to 14% from Monte Carlo calculations and is highly dependent on the energy of the source. The results also suggest that LiF:Mg,Ti TLDs are not intrinsically linear with energy imparted.
TU‐C‐224C‐06: Determination of TG‐43U1 Recommended Parameters for Elongated RadioCoil™ Brachytherapy Sources 1.0 to 6.0 Cm in Length Using Experimental and Monte Carlo Simulation Techniques33(2006); http://dx.doi.org/10.1118/1.2241541View Description Hide Description
Purpose: In this project TG‐43U1 recommended dosimetriccharacteristics of newly designed elongated RadioCoil™ brachytherapy sources(1 to 6)cm in length have been determined following experimental and Monte Carlo simulation techniques. Monte Carlo simulated dose profiles have also been calculated for sources 1 to 6cm in length. Materials and Methods: TG‐43U1 recommended dosimetriccharacteristics (Λ, g(r), F(r,θ), φan) of RadioCoil™ sources have been determined using experimental(TLD) and Monte Carlo(MCNP5) simulation techniques. MCNP5 simulations were performed in spherical Solid Water™ and water phantoms 20cm in diameter for 108 histories. F(r,θ) of RadioCoil™ were determined for angles 0° to 90° for r⩾L/2(where L=active length), and angles 5°⩽θ⩽90° for r⩾L/2, with 5° increment. TLD measurements were performed in solid Water™ using same geometric arrangement as that of Monte Carlo simulations. Measured and calculated dose profiles along the longitudinal axis of these sources were utilized to validate the dose calculation with commercially available treatment planning systems. Results: Results of these investigations indicate good agreement between MCNP5 simulated and TLD determined values for RadioCoil™ sources. Upon the good agreement between these two methodologies, confirming the accuracy of our simulation geometry, dosimetric parameters of these sources have been determined in liquid water for their clinical applications. It has been found that in order to achieve a good agreement between the treatment planningdose distribution, the F(r,θ) of these sources have to be determined at radial distances ranging from 0.5 to 5.0cm with 0.5cm increment and L/2 ±0.2cm. Conclusion:Dosimetriccharacteristics of newly designed RadioCoil™ have been determined following TG‐43U1 recommendations using experimental and Monte Carlo simulation technique. In these determinations it has been found that the dose profile can be closely reproduced if the 2D anisotropy function is determined at 0.5 cm increments for radial distance ranging from 0.5 cm to 5cm, and L/2 ± 0.2cm.
33(2006); http://dx.doi.org/10.1118/1.2241542View Description Hide Description
Purpose: To propose dosimetric guidelines specifically designed for the Cyberknife radiosurgery system. Non‐availability of 10×10cm2 field and use of small circular collimators (5mm to 60mm) pose serious problems, that have been faced in this study by means of 8 different detectors and Monte Carlo simulation. This work is oriented to measurement of total scatter factors (Sc,p) and to reference dosimetry, though indications will also be given in view of a comprehensive guideline. Method and Materials: PTW PinPoint 31014, Exradin A16 and T14P microchambers, TN 502RDM micromosfet, PTW 30008 diode and TM60003 diamond, MD55 and EBT radiochromic films were used to measure Sc,p. Monte Carlo simulations (BEAMnrc) were used to produce phase space descriptions at the exit plane of each collimator, to calculate: 1) theoretical Sc,p values in water, and 2) correction factors to be applied to Sc,p as measured by 5 detectors (PinPoint, A16, T14P, diode, diamond), obtained by simulating shape and chemical composition of each detector. BEAMnrc was also used to calculate stopping power ratios and chamber correction factors for the Cyberknife linac, to decide whether values of kQ from the IAEA398 protocol could be applied without using a 10×10cm2 field. Results: Sc,p of the 5mm collimator as measured by simulated detectors averaged 0.653 − 9%+14%. Variation for larger collimators was smaller. After Monte Carlo correction, Sc,p of the 5mm collimator became 0.686 −2%+1%. Pure Monte Carlo calculation gave Sc,p=0.715 +/−1%. Calculation of correction factors showed that kQ values for the investigated chambers could be chosen when using IAEA398, introducing +/−0.2% uncertainty. Conclusion: Pure Monte Carlo calculation gave higher values of Sc,p compared to Monte Carlo‐corrected measurement. The latter is to be preferred because correction factors are less sensitive to beam parameters than pure calculation of Sc,p. For determination of Sc,p use of microchambers and Monte Carlo correction is recommended.
33(2006); http://dx.doi.org/10.1118/1.2241543View Description Hide Description
Purpose: To design a small radiation facility for the partial‐ and fully ‐ body irradiation of zebrafish embryos, cell cultures or any other small specimen used for radiobiology studies. Method and Materials: Zebrafish embryos larger than 1mm are the main animal to be irradiated in this micro‐irradiator. Radiation is provided by a 50 kV photon beam from a miniature x‐ray source, Xoft Inc., CA. Radiation field is delimited by a pinhole collimator. Diameter of the pinhole ranges from 0.5mm. A movable table and a video camera connected to a computer are used to position the specimen under the beam. Radiochromic film has been irradiated to test positional accuracy and dose distribution. Results: Coordinates of the position of the zebrafish with respect to the collimator are calculated from the image provided by the video camera and sent to the computer‐controlled movable table to position the specimen under the beam. The micro‐irradiator is totally portable and it can fit on the desktop. Positioning can be acquired with uncertainties of 50 μm in X and Y axis. Collimator design and portable shielding reduce both primary beam and scattering at safe radiation levels. Dose distribution is good enough for zebrafish irradiation and penumbra is in the order of 150μm ± 50μm. Conclusion: The designed micro‐irradiator has attractive characteristics to facilitate zebrafish irradiation. Its portability and shielding simplicity make it adequate for any radiobiology laboratory. Its uncertainty in positioning is significantly smaller than zebrafish embryo size and radiation penumbra is acceptable for specimen larger than 1mm. This novel micro‐irradiator design is appropriate for irradiation of partial‐and fully‐ body zebrafish, cell cultures and any other small specimen used in radiobiology.
33(2006); http://dx.doi.org/10.1118/1.2241544View Description Hide Description
Purpose: To study the dosimetric characteristics of a gated fiber‐optic‐coupled detector for measuring absorbed dose from a linear accelerator. The favorable properties of these dosimeters are their superior spatial resolution, real‐time readout and potential as in‐vivo dosimeters. In an application that takes advantage of its spatial resolution, small field tissue‐phantom‐ratio (TPRs) measurements from 0.6×0.6 to 2.0×2.0 cm2 are obtained and compared to those measured with a diamond and diode detector.Method and Materials: The detector is a short length of Cu1+ fused doped quartz fiber coupled to a fiber‐optic cable. It has an atomic number of about 10.8 with a density of 2.2 g/cm3 and is 1 mm long with a diameter of 0.4 mm. The background signal generated by Cerenkov radiation and native fluorescence within the optical fiber during irradiation is separated from the detector luminescence via gating the signal with the radiation pulses from the accelerator. A linear accelerator provides mega‐voltage photon and electron beams to investigate its energy response, dose rate dependence, dose linearity and reproducibility. Results: There is no measurable difference in the detector response between 6 and 18 MV photons. However, for electrons the dose response increases gradually by 7% from 6 to 20 MeV. Its dose rate response relative to a Farmer chamber exhibits a behavior similar to a diamonddetector decreasing by about 4.5% from 0.8 to 10.7 Gy/min for both 6 and 18 MV photons. The measured response is linear from 0.2 to 10 Gy and its reproducibility is better than 2%. The small field TPR measurements are in good agreement with the diamond and diode detector.Conclusion: The dosimetric properties of this detector compare favorably with other radiation detectors, and its small size and optical interface make it potentially very useful for small field and in‐vivo radiation dosimetry.
33(2006); http://dx.doi.org/10.1118/1.2241545View Description Hide Description
Purpose: To elaborate a transfer‐function approach to characterizing the response of watercalorimeters to abbreviated exposures in radiation beams. Method and Materials: Typical watercalorimeters in use today derive from the original design of Domen, inferring absorbed dose from temperature measurements at a single point in a partially irradiated, extended volume. The success of the approach in deducing locally absorbed dose depends on heat transfer being “sufficiently” slow temporally and broad spatially compared to time‐ and space‐scales of the measurement. In order to address these issues of “sufficiency” more quantitatively, we have undertaken an approach to the problem that assesses the impulse‐response of the calorimeter to spatially and temporally localized radiation. The approach involves simple analytical models of heat conduction and finite‐element methods that yield predictions of single‐point temperature waveforms obtained from thermistors in a calorimeter. Output from the models is compared with experiment over wide variations in shutter period (30s–10,800s) and duty cycle (5%–50%). Results: Our findings show that heat conduction due to typical dose inhomogeneities within watercalorimeters induces systematic variations in estimated dose that traditional data‐analysis techniques ignore. Moreover, these variations do not diminish at smaller duty cycle, suggesting that they are intrinsic to the exposure time of the experiment, and not to the use of periodic exposure techniques. Essential features of the measurements are reproduced by both finite‐element simulations and a simple analytical model. An RC‐circuit analogue derived from the latter suggests that conduction phenomena with a spatially masked beam proceed by a multiplicity of times scales, ranging from seconds to hours, which contribute significantly to systematic variations in estimated dose. Conclusions: The variations of estimated absorbed dose described here suggest that a small systematic error (<1%) may affect existing standards measurements based on watercalorimetry.
- Clinical Measurements II
33(2006); http://dx.doi.org/10.1118/1.2241780View Description Hide Description
Introduction: Scintillation dosimeters consisting of a single scintillating fiber coupled to an optical fiber and read with a CCDcamera have shown accurate dosemeasurement in high‐energy external beam radiotherapy. This work presents the next step, which is to investigate the development of a multi‐channel dosimeter array in conjunction with a CCDcamera for photodetection. Methods and Materials: The light collection of the CCDcamera (Apogee Alta U2000c) was first studied to evaluate the maximum number of detectors that can be read simultaneously. We then looked at possible sources of dose perturbation with a single‐fiber detector surrounded by other optical fibers. We constructed a prototype array with 10 detectors. Finally, depth‐dose and cross‐plane profiles were compared with measurements taken with small ion chambers (Exradin models A14 and A16). Results: We found up to 3000 individual fibers could be detected simultaneously with the full CCD chip. No dose perturbations were seen when a plastic optical fiber was used to transmit the light from the scintillating fiber to the CCD. Depth dose curves measured in water with a single scintillation dosimeter with up to 75 plastic optical fiber in the beam showed no discrepancy to within 0.3% when compared to the same curve taken with an ion chamber and without the plastic fibers. The ten‐fiber prototype allowed precise evaluation of profile and depth dose curves in a single irradiation. Conclusion: This work has shown that use of a multi‐channel scintillation dosimeter is feasible. The prototype of 10 detectors produced excellent results and could be extended up to 3000 detectors in the near future.
WE‐D‐224C‐02: High Dose Rate Mode Linear Accelerator Based Stereotactic Radiosurgery and Image Guided Radiation Therapy33(2006); http://dx.doi.org/10.1118/1.2241781View Description Hide Description
Introduction: In order to improve the standard linear accelerator's ability to address the demands of intracranial and extracranial stereotactic radiosurgery as well as for image‐guided and intensity modulated radiotherapy, a high dose rate output accelerator was developed. The objective of this study was to assess the device performance and compare it to standard accelerators. Materials and Methods: The 6 MV x‐ray beam of a Siemens Oncor linac was modified by removing the flattening filter, enabling dose rates to reach 1000 MU per minute. Ion chambers, diodes, and film dosimeters were used to assess monitor chamber behavior, energy, and dose profile characteristics. Treatment time required for radiosurgerytreatments in 40 patients was measured. Results: Even at this high dose rate, the linacdosimetry system remains robust; constancy, linearity, and beam energy remain within 1% for 3 to 1000 MU. Measurements at incrementally reduced dose rates (1000 to 300 in 100 MU increments) showed the output calibration to fall from 1.000 cGy/MU to 0.994 cGy/MU, with an average value of 0.997 cGy/MU. Over this range of dose rates the beam energy held consistent, with the ratio of percent depth dose values within 0.5%. Dose profiles for larger field sizes are not flat, but they are radially symmetric and as such able to be modeled by a treatment planning system. Radiosurgerytreatment times, computed here as the beginning of x‐ray delivery until the end of treatment, were reduced to an average of 2 minutes and 18 seconds per arc, or 11 minutes 27 seconds per isocenter. Conclusions: Even at this high dose rate, the linacdosimetry system remains robust. Because stereotactic IGRT can require significantly longer times for treatment delivery, the advantages of the high dose rate design should be pursued.
33(2006); http://dx.doi.org/10.1118/1.2241782View Description Hide Description
Purpose: To measure the surface and near‐surface doses for a variety of IMRTdelivery and planning techniques, and to determine the optimal PTV size and planning scenerio for skindose maximization or minimization. Methods and Materials: A primary PTV was defined on an anthropomorphic head phantom for a typical parotid‐sparing head &neck treatment. Treatment plans were created with 0, 1, 2, 3, 4, and 5‐mm separation between the skinsurface and PTV boundary. IMRT treatments were planned using the Pinnacle treatment planning system and delivered on a Varian 21EX with a 120‐leaf multileaf collimator. Inverse planning was performed using Direct Machine Parameter Optimization (DMPO) for step and shoot IMRT and gradient decent optimization for sliding window leaf sequencingIMRT. Helical tomotherapy cases were planned using the TomoTherapy HI‐ART treatment planning system and delivered using a Tomotherpy HI‐ART treatment delivery system. Surfacedoses were measured for each of the treatment deliveries using film placed in an anthropomorphic head phantom and thermoluminescent dosimeters(TLD) chips placed on the phantom's surface. A chip‐specific TLDsurface calibration factor was determined and applied to the raw TLD readings to account for measurement efficiency at the surface. Relative dosemeasurements were made using calibrated film placed in the phantom. Surface and near‐surface doses were measured from digitized film images for a 1.5‐cm range inside the phantom to in‐air outside the phantom. Results:Surfacedose values measured with film were consistently lower, while TLDmeasurements were higher than planned for the cases studied. Conclusions: The helical tomotherapy treatment plans were found to have better parotid sparing and PTV dose uniformity than both the step‐and‐shoot DMPO and sliding window plans. The tomotherapy planning system was observed to overestimate the surfacedose from 9 to 18 percent.
WE‐D‐224C‐04: Investigation of MLC Effects On Secondary Neutron Spectra for Varian, Siemens, and Elekta33(2006); http://dx.doi.org/10.1118/1.2241783View Description Hide Description
Purpose: To compare the secondary neutron spectra in accelerators with different multileaf collimator(MLC) configurations. The addition of MLC, have lead to design modifications in modern linear accelerators; Elekta has replaced upper collimating jaws with MLC, Siemens has replaced lower jaws with MLC, Varian added tertiary level MLC.Method and Materials: Measurements were made of the neutron fluence and energy spectra for several modern accelerators. The detector consisted of 197Au activation foils, which were placed on the surface of the holder and inserted into Bonner Spheres. An HPGe detector was used to measure counts under the 411keV photopeak for each foil. Data were unfolded with the MXD_FC33 code with a response matrix specifically calculated for this measurement system using MCNP5. In this investigation, neutron spectra, fluence per MU, and ambient dose equivalent are reported for 18MV x‐ray beams generated by Varian 21EX, Siemens Oncor, and Elekta Precise accelerators. The impact of the jaw and MLC configuration were further studied for the Varian 21EX by taking measurements following the complete removal of the MLC.Results:Jaws and MLC closed: similar spectra for the Siemens and Elekta but the Varian spectrum has a lower energy distribution. Varian X and Y jaws closed: With MLC in place, less neutrons are detected and spectrum shifts to lower energies (compared to MLC removed). MLC attenuates neutrons created higher in treatment head. Varian X and Y jaws retracted: With MLC in place more neutrons are detected and the spectra shifts to lower energies (compared to MLC removed). The MLC attenuate neutrons created higher in treatment head but MLC become the primary source of contamination neutrons when jaws retracted. Conclusion: Secondary neutron spectra are different in accelerators with different MLC configurations. This difference translates to a difference in ambient dose equivalent to patients receiving high‐energy radiation therapy.
33(2006); http://dx.doi.org/10.1118/1.2241784View Description Hide Description
Purpose: To determine the magnitude of superficial doses produced from a static TomoTherapy beam directed at varying SSDs and angles of incidence. Method and Materials:Measurements of superficial doses have been made along the central axis of a static TomoTherapy beam at normal incidence for SSDs of 55, 70, and 85 cm for typical TomoTherapy jaw sizes (40×2.5 cm2 and 40×5 cm2). Measurements have also been made along the central axis of a TomoTherapy beam at oblique angles of 30°, 45°, 60°, 75°, and 85°. Data were collected with a Gammex Model 449 parallel‐plate chamber embedded in a solid water phantom and LiF TLDpowder. For comparison, measurements have been made on a 6X Varian 2100C accelerator with the same jaw width (5 cm at 85 SSD) and at the same SSDs. Percentage depth dose (PDD) profiles for depths ⩽2 cm have been obtained from the data. Results: TomoTherapy surface dose measurements vary weakly with SSD, ranging from 16%–18% for the 40×5 cm2 field, and from 12%–14% for the 40×2.5 cm2 field. The measured doses increase rapidly with depth, with PDD>90% obtained at depths < 0.6 cm. Surface dose ranges from 17%–26% on the Varian 2100C for the same SSDs and 5cm jaw width. TomoTherapy surface dose increases from 16% to 43% as the angle of incidence increases from 0° to 85° for the 40×5 cm2 field and increases from 12% to 40% for the 40×2.5 cm2 field. Conclusion: TomoTherapy surface doses do not vary significantly with SSD but increase dramatically with increasing angle of incidence. Generally, the magnitude is less than that measured from a conventional, flattening filter‐based linac. These data should assist in assessing the accuracy of the TomoTherapy planning system in the calculation of superficial doses.
Supported in part by a research agreement with TomoTherapy, Inc.
WE‐D‐224C‐06: Leakage Characteristics of Two Common Multi‐Leaf Collimators: Implications for Intensity Modulated Radiotherapy33(2006); http://dx.doi.org/10.1118/1.2241785View Description Hide Description
Purpose: To evaluate leakage characteristics of two Varian MLCs during IMRT delivery. Introduction:IMRT increases total body dose due to increased leakage proportional to the additional monitor units required. During MLCIMRT delivery, collimator jaws are positioned outside the edges of the entire intensity map, increasing the prevalence of leakage through and potentially around the MLC. The magnitude of the latter component is dependent upon the physical size of the MLC.Method and Materials: The Varian Mark‐I 52‐leaf and Millennium 120‐leaf MLCs have widths of 26 and 40cm, respectively, in the direction perpendicular to leaf motion. Identical, 1cm bixel‐size small (8×9cm2), medium (12×16cm2), and large (20×24cm2) intensity maps were delivered in SMLC mode with each MLC using 6MV photons.Ionization chamber measurements were performed outside these fields at the surface and 10cm depth in a patient‐simulating phantom at 15, 20, 30, 40, and 50cm off‐axis both along and perpendicular to the direction of leaf motion. Results: Mark‐I leakage in the leaf motion direction was very similar to Millenium‐120 leakage in either direction. However, Mark‐I leakage was considerably higher perpendicular to the leaf motion direction. This increase was apparent at all points, with a maximum of 100–150% at 20–30cm off axis for all fields. The average increase for all fields and positions was 90% at the surface and 65% at 10cm depth. Previous investigations have indicated significant risks of secondary malignancy induction associated with total body doses encountered in IMRT. Our results indicate that this risk is nearly doubled using a Mark‐I MLC when the patient axis lies along the narrow dimension of the MLC.Conclusion: The Mark‐I MLC can profoundly increase total body dose in comparison to a wider MLC. This increase may be largely alleviated by judicious choice of collimator angle. This risk should be considered during IMRT planning.