Volume 34, Issue 6, June 2007
Index of content:
- Therapy Scientific Session: Room M100E
- New Dosimeters, Treatment Techniques and Clinical Applications
TH‐C‐M100E‐01: Numerical Feasibility Study of a Novel Absorbed Dose to Water Calorimeter‐Based Standard for 192Ir HDR Brachytherapy34(2007); http://dx.doi.org/10.1118/1.2761669View Description Hide Description
Purpose: To study the feasibility of developing a new standard for absorbed dose to water based on watercalorimetry for high dose rate iridium‐192 brachytherapysources.Method and Materials: The heat conduction pattern generated in water by the Nucleotron microSelectron‐HDR brachytherapysource was simulated using Comsol Multiphysics™ software.Source self‐heating due to self‐attenuation of photons was calculated with GEANT4. A smooth, well‐behaved three‐dimensional function was fit to the entire dose distribution data using TableCurve3D™. The heat‐loss correction Kc was calculated as the ratio of the temperature in the calorimeter under ideal conditions to realistic conditions. Results: The feasibility of a watercalorimeter based absorbed dose standard is determined by a balance between the requirements to obtain sufficient signal to perform a reproducible measurement, the effects of heat loss on the measured signal, and the positioning uncertainties. Due to self‐absorption, the source equilibrium temperature was found to be above the ambient temperature by a constant amount that depends only on setup conditions and source activity. For the source inside its nylon‐12 catheter inserted into water, the steady state excess temperature per unit source activity was found to be 0.5671 K/Ci. The source temperature reached 96% of its steady state temperature after 60 s. Conduction correction factors Kc were calculated for several exposure times and at various measurement points away from the source inside the calorimeter. A total exposure time between 140 s and 240 s at a distance that receives a minimum of 1 Gy of dose was found to allow reduction of Kc to below 0.1% of unity. Conclusions:Watercalorimetry for HDR brachytherapy is feasible and total uncertainties of significantly better than 5% on the dose can be achieved with current watercalorimetry techniques and instruments.
TH‐C‐M100E‐02: Optically Stimulated Luminescence of Aluminum Oxide Detectors for Radiation Therapy Quality Assurance34(2007); http://dx.doi.org/10.1118/1.2761670View Description Hide Description
Purpose: The purpose of this experiment was to: 1) Determine if a commercially available detector system used for monitoring personnel exposure could be adapted for use as a radiation therapydosimetry system; and 2) Evaluate the system's performance as an in‐vivo dosimeter and its ability to measure absolute surface dose, isocenter dose, and normal tissues dose in a phantom as part of patient‐specific IMRTquality assurance.Method and Materials: The dosimeters were evaluated for: 1) Signal decay; 2) Field size dependence; 3) Energy dependence; and 4) Angular dependence using the Landauer, InLight MicroStar system. In‐Vivo dosimetry measurements were taken for 22 patients treated on a Varian 21EX. The Landauer system was also tested for its ability to measure absolute dose from helical tomotherapy treatments. Results: The variation between dosimeters was evaluated and found to be ±1.6%. The dosimeters appeared to over‐respond in the first 10 minutes, however, after 10 minutes the chips were within 1 percent of the steady‐state reading. Unlike other detectors, the Al2O3dosimeters showed no field size, energy, or angular dependence. The agreement between the dosimeters and the calculated doses for the in‐vivo dosimetry patients was 2.2±6.1 cGy or 3.7±2.5%. The dosimeters were also tested for their ability to measure absolute dose inside an IMRT phantom. The agreement between the dosimeters and the calculated doses was 0.1±5.3 cGy or 0.7±6.7%. Conclusion: dosimeters can be a convenient, inexpensive alternative to TLDs, MOSFETS, and Diodes. The agreement between calculated and measured doses for in‐vivo dosimetry and IMRT QA is comparable to TLDs, MOSFETS, and Diodes. The dosimeters can be quickly read and analyzed after 10 minutes (to allow time for signal decay). The dosimeters do not appear to have an energy, field size, or angular dependence. In addition, the detectors can be erased and re‐used.
TH‐C‐M100E‐03: A New Generation of Electronic Portal Imaging Devices (EPID) Using Thin‐Film CdTe for Radiation Oncology Applications34(2007); http://dx.doi.org/10.1118/1.2761671View Description Hide Description
Purpose: The commercially available EPIDs today commonly manufactured from amorphous‐Silicone (a‐Si) materials display numerous problems in imagecontrast & resolution. These are related to poor radiationhardness and low Z which is detrimental to their detecting capabilities. The CdTe‐based thin film devices with superior radiationhardness and much higher Z (∼50) have not yet found many applications in medical physics. Here we report first promising results on polycrystallinethin film CdTe‐based prototype EPID.Method and Materials: The proposed design utilizes a layer of high atomic number and density functioning as a converter transforming high energy X‐rays into the Compton electrons impeding onto CdTe detector operating in pulse mode. Such a converter can replace the standard scintillators used with a‐Si devices. We conducted Monte Carlo simulations testing the proposed structure and verified them with measurements using a prototype thin filmCdTe cells. Our modeling was amplified with a semi‐empirical algorithm accounting for the processes of device degradation, simultaneously applied to the a‐Si devices for comparative study. Results: We found that a layer of Pb less than 3 mm thick in combination with low‐Z material such as polystyrene, used for filtering out low‐energy scatter, is suitable for the converter. Detector output voltage in the range of the tenths of Volt for the typical radiation therapy dose rates allowed for a possibility of using the device without biasing. We have carried out verifying experiments with polycrystallineCdTe based cells: good agreement with our MC simulations was obtained. Conclusion:CdTe based thin film detectors have a high potential to become the next generation EPID's. They feature: small thickness <10 μm; efficient collection; exceptional radiationhardness; inexpensive deposition technology; sensitivity to low intensity radiation; excellent room temperature characteristics without biasing; small integration time < 500 ns.
TH‐C‐M100E‐04: Cerenkov Light From Phantom Cassettes in Absolute Dose Measurements Using Radiographic Film34(2007); http://dx.doi.org/10.1118/1.2761672View Description Hide Description
Purpose: To determine the impact of Cerenkov light on radiographic film response and to recommend a methodology for correcting absolute dose measurements when using bare film, but calibrated using prepackaged film. Method and Materials: The Gammex RMI Film Dosimetry Cassette Model 436‐AST (Solid Water) and two in‐house cassettes (white opaque, high‐impact polystyrene) were studied using Kodak XV or EDR2 film. Films were exposed perpendicular to 16‐MeV electron (15×15 cm2) or 6 MV x‐ray (20×20 cm2) beams. Films were oriented such that quadrant ♯1 had bare film; quadrant ♯2 had film covered by the prepackaged white paper; quadrant ♯3 had film in its prepackaged container; and quadrant ♯4 had film covered by the prepackaged carbon jacket. To account for beam asymmetry, dose response for each quadrant was normalized to that in the corresponding quadrant of a film irradiated in the carbon jacket, which blocked phantom‐produced Cerenkov light. A prepackaged film, irradiated using a multi‐exposure technique, provided the dose‐response calibration.Results: The “carbon jacket only” dose values averaged 96.1% of the “prepackaged” dose values, indicating that the prepackaged white paper produced Cerenkov light that increased film response by 4.0%. No significant difference due to radiation modality or film type was evident. The “white paper only” dose values ranged from 103.6–107.5% of the “prepackaged” dose values, indicating that Cerenkov light from the phantom material contributed to an increased film response. For white opaque, high‐impact polystyrene the “bare film” dose values ranged from 102.2–109.6% of the “prepackaged” dose values, depending on phantom and modality. For Solid Water the “bare film” dose value was 117.3% of the “prepackaged” dose values. Conclusion: When making absolute dose measurements using bare film and calibrating using prepackaged film, a correction for excess film response arising from Cerenkov light is required, and the reported quadrant method is recommended.
34(2007); http://dx.doi.org/10.1118/1.2761673View Description Hide Description
Purpose: A customized gating system was developed for a large bore CT scanner to produce gated images for use with an active breathing control (ABC) system. Operator triggering following a breath hold is replaced by an electronic gating signal from a bellows. A breathing phantom was constructed and imaged to insure accurate gating. Methods and Materials: A bellows strap wraps around the patient to monitor breathing during CT scans. A universal serial port (USB) cable provides power to the bellows sensor via the circuit box. Bellows signals route to the computer from the circuit box via another USB cable. Operators can set a threshold point in the breathing cycle that corresponds to an ABC breath‐hold to trigger the CT scanner. Softwaregenerated signals gate the CT through the circuit box. Unique to this system is an option to scan continuously while the patient is holding a breath, in contrast to commercial gated‐CT systems which take one image per breathing cycle. A breathing phantom was created using a computer programmed actuator, step‐motor, and off‐set ball. A sine wave modeled on human breathing was fed to the actuator. The moving range was 2.10 cm anterior‐posterior and 1.05 cm in lateral direction. The diameter and motion period was 6.4 cm and 5.54 sec. Phantom images taken were: stationary non‐gated, moving gated, and three un‐gated while moving. All image sets were analyzed with the Pinnacle planning system. Volumes generated from contours where used as a metric to determine imaging accuracy. Results: Target volumes of stationary and gated imaging were 146.8 and 145.5cm3, respectively (1% agreement). However, the discrepancies between stationary and all three un‐gated scans ranged from 4.84 to 5.45% with significant shape distortion.
Conclusions: We successfully created a custom gating system for a large‐bore CT simulator.
TH‐C‐M100E‐06: Determining Optimal Respiratory Gating Parameters for Passively Scattered Synchrotron Based Proton Irradiation34(2007); http://dx.doi.org/10.1118/1.2761674View Description Hide Description
Purpose: Respiratory gated irradiation offers potential for margin reduction and dose escalation for treating moving tumors in the thorax or abdomen. Unfortunately, for synchrotron‐based proton irradiation, it may not be efficient. We have determined the optimal respiratory gating parameters for passively scattered proton irradiation on a synchrotron through a simulation study. Method and Materials: An in‐house software program was developed to investigate the interaction of the respiratory gating intervals with different synchrotron magnet excitation cycle patterns. Test data was obtained by using the recorded respiratory trace of 94 patients who underwent 4DCT. A typical magnet excitation cycle, Tcyc consists of proton acceleration, flat top and deceleration periods. Proton beam delivery occurs only during the flat top portion of each such excitation cycle. Respiratory gating was simulated at expiration for a 30% duty cycle around peak exhalation. The time required to deliver 100 MUs was estimated for the following scenarios: (a) Ungated irradiation with Tcyc set to the minimum value (2.7sec) and (b) Gated irradiation with Tcyc set to (i) the minimum value, (ii) approximately equal each patient's average respiratory cycle, and (iii) a variable value according to each individual respiratory cycle. Overall treatment time and efficiency of treatment delivery were studied in each case. Results: Average times required to deliver 100 MUs were 1.1 minutes for ungated irradiation; and 3.7 (1.7 – 6.0), 3.2 (1.6 – 7.1), 2.3 (1.4 – 3.1) minutes respectively for gated irradiations at various scenarios mentioned above. For gated irradiation, variable Tcyc mode of operation yielded least overall treatment time and greatest efficiency of proton beam delivery. Conclusion: Respiratory gated passively scattered proton delivery using a synchrotron‐based system is feasible without significantly increasing treatment time. Based on above results, variable Tcyc mode of operation offered least overall treatment time and greatest efficiency for respiratory gated irradiation.
34(2007); http://dx.doi.org/10.1118/1.2761675View Description Hide Description
Purpose: To report experience with a novel total body irradiation (TBI) technique. 3D planning techniques are used to deliver a uniform dose to a patient using a conventional linear accelerator in a standard bunker. Manually segmented intensity modulated fields are employed to provide dose compensation for contour variation, tissue heterogeneity, inverse square law effects and junction dose stability. Methods and Materials: The technique uses a conventional Elekta Synergy linear accelerator together with a custom designed floor couch. The couch, positioned 102.5 cm below the machine isocentre, provides treatment distances near 180 cm SSD. The couch is oriented in the gantry rotation plane, with couch motion along the cranial‐caudal axis enabling a match of beam divergence through patient translations and gantry rotations. Treatment is delivered by a set of 2 to 3 divergence matched abutting fields, with field modulation feathering junctions through 4 cm on the patient. Treatment plans are created using conventional beam models in Pinnacle 7.6C and whole body CT scan data. Independent plans for supine and prone orientations are constructed to deliver a uniform dose at mid‐separation throughout the patient and create a composite uniform dose. Segmentation is used to adjust the dose at mid‐plane, correcting for effects of patient thickness, inverse square law, and lung density. Results: A total of 11 patients have been treated with this new technique. Dosimetry measurements in phantom at extended distance and in‐vivo measurements have demonstrated an accurate dose delivery. Composite AP‐PA dose assessments based on contributions to uniquely identified anatomical points have shown that a dose within 10% of the prescribed dose is achieved throughout the treatment volume. Conclusions: A new TBI technique has been implemented which employs modern imaging and delivery methods to achieve a uniform patient dose. The technique utilizes standard equipment, and does not require specialized bunker design.
TH‐C‐M100E‐08: Evaluation of Multiple‐Isocenter IMRT Planning Technique for Field Matching with Limited Collimator Field Size34(2007); http://dx.doi.org/10.1118/1.2761676View Description Hide Description
Purpose: The Elekta Beam Modulator with fully integrated miniMLC has the precision suited for the treatment of smaller tumors. However, the maximum collimator length in the inf/sup direction is 16 cm when most of the head and neck target sizes have a larger dimension. This study used a multi‐isocenter IMRT planning technique with overlapping fields generated using inverse planning. Measurements have been taken to evaluate plans in terms of volume dose and delivery accuracy. The total monitor units for this technique were compared with a standard IMRT plan. Methods and Materials: Ten patients were planned with multiple beam arrangements; eight beam directions for the upper tumor volume without the supraclav and three to four beam arrangements for the lower tumor volume and supraclav. Treatment planning was performed on a CMS/XiO workstation. Treatment fields were delivered on a flat acrylic phantom containing EDR2 film. Isodose distributions were exported from XiO and compared to the measured data using Omni‐Pro software.Results and Conclusions: Multi‐isocenter techniques provide good coverage to the tumor volume while sparing organs at risk. Film distributions show that XiO does not model the effective tongue and groove that is evident in the film measurement. The magnitude of the tongue‐and‐groove dose decrease is on the order of up to 10%. An analysis of the gamma values shows that even with this discrepancy the agreements between plan and measurement were acceptable. The plans are not significantly degraded with small table inaccuracies on the order of 2mm. The total monitor units for multi‐centered plan increased 20% compared to single isocenter delivery.
IMRT planning techniques using two or more isocenters for this collimator show clinically acceptable results when splitting the geometry.
TH‐C‐M100E‐09: Assessment of Skin Dose for Breast Chest Wall Radiotherapy as a Function of Bolus Material34(2007); http://dx.doi.org/10.1118/1.2761677View Description Hide Description
Purpose:Skindose assessment to the chest wall is important to ensure sufficient dose to the near‐surface target volume without undue skin reaction. Bolus is often used for a portion of the treatment course, but removed if clinically necessary because of skin toxicity. This study quantifies changes to the surface dose as a function of bolus material for conventional and IMRT techniques using a thermoluminescent dosimeter(TLD) extrapolation technique. Methods and Materials: Three types of bolus materials (2mm solid, 2mm fine mesh, and 3.2mm large mesh aquaplasts) were compared with Superflab (Med‐Tec). Surface dosemeasurements were performed using the Attix (parallel‐plate) chamber in a flat solid water phantom with the bolus materials for 10×10 cm2 and 10×20 cm2 jaw fields at 00, 450 and 700 incident angles. The Attix chamber measurements were used to validate the TLD extrapolation technique (0.89, 0.38, 0.15 mm thicknesses). TLDs were used to measure the surface dose on an anthropomorphic phantom for conventional and IMRT tangential fields. Results: Surface dose increased with increasing angles and field sizes. Oblique incidence has a larger influence on the surface when no bolus is present (from 20% to 48% for 10×10 cm2). The skindose of solid 2mm aquaplast was larger than that of fine mesh aquaplast (22% for 10×10cm2−00 incidence, and 11% for 10×10cm2−700 incidence). Compared to conventional tangential fields, the skindose for IMRT decreased ∼5%. For the conventional tangential fields, skindoses of fine mesh, solid, and large mesh aquaplasts were 21%, 11% and 9% less than that of superflab, respectively. For IMRT fields, skindoses were 22%, 12% and 10% less than that of superflab, respectively. Conclusion: For chest wall radiotherapy, the bolus type can be selected to compromise near‐skin target dose vs skin tolerance dose for optimal clinical outcome.
TH‐C‐M100E‐10: A Correction Method for the MOSFET Energy Dependence Response to Therapeutic Proton Beams34(2007); http://dx.doi.org/10.1118/1.2761678View Description Hide Description
The energy‐dependence response of the metal oxide semiconductor field‐effect transistor(MOSFET) dosimeter has been investigated with regard to therapeuticproton beams. The MOSFET configurations used include the commercial standard MOSFET (TN‐502RD), the microMOSFET (TN‐502RDM) dosimeters, and a prototype MOSFET with no encapsulation. Proton beams of non‐modulated pristine and modulated Spread‐Out Bragg Peak (SOBP) of 5 cm and 10 cm widths were used with beam ranges of 8.9cm, 15.9cm and 25.9cm in water. Each MOSFET dosimeter was calibrated at the center of the modulated 10 cm SOBP proton beam with conditions of 1 cGy per MU. The MOSFET energy‐dependence response was quantitatively evaluated by the ratio of measured doses between the MOSFET and an ionization chamber in a same condition. The three dosimeters showed a similar response for the pristine proton beams at various beam ranges. This indicates that the variation in dosimeter response is dominated by the change of the linear energy transfer (LET) for the used proton beam and not by the MOSFET encapsulation thickness. The observed MOSFET trends for various pristine proton beams have been modeled by an analytical function , where Rres is the residue range (distance to the distal 90% level dose), and A, B, C are constants. A similar modeling has been performed for modulated SOBP proton beams at different widths and for various beam ranges. The k (Rres ) function has been used as a correction factor for patient dose measured by MOSFETs for corresponding residue ranges; this resulted in dose measurements with an uncertainty of less than 3.0% for various proton beams.
- Tx Planning and Delivery — Clinical Planning
34(2007); http://dx.doi.org/10.1118/1.2761728View Description Hide Description
Purpose: Dose equalization along the long axis of the patient for total body irradiation requires the use of a compensator. At our institution the compensator consists of multiple layers of lead strips and is based on measurements along the patient's mid‐sagittal plane. In this study, we examine the feasibility and limitations of replacing the lead compensator with a one‐dimensional electronic compensator using fluence management tools available in our treatment planning system. Method and Materials: The patient compensator was based on body thickness, SSD and off‐axis distance of 12 mid‐sagittal patient specific points. A phantom was modeled using these measurements from a previously treated patient. A fluence map of the transmission values for the compensator used at treatment was created. Dose calculated on the phantom was compared with patient surface dose measurements. Physical limits and software limitations of this method were evaluated. Results: The maximum patient height accommodated by our treatment room and setup is 225 cm. Using a fluence map with the largest transmission factor gradient is not an issue, nor is the use of any reasonable field width that would be seen in a clinical setting. Calculated dose to the phantom using the electronic compensator was found to be on 3.4% lower (range: +2.6% to −7.8%) than diode dose measurements taken on the patient skin surface time of treatment.Conclusion: It is feasible for most patients that an electronic compensator be used. Discrepancies between calculated and measured dose can reasonably be accounted for. Further study is planned to measure the dose using an anthropomorphic phantom and also to automate the construction of patient‐specific virtual phantoms and patient‐specific optimal fluence.
34(2007); http://dx.doi.org/10.1118/1.2761729View Description Hide Description
Purpose: To examine characteristics of electron beams collimated by an electron multileaf collimator (eMLC) for modulated electron radiotherapy (MERT). Method and Materials: An eMLC, which has 25 pairs of Tungsten leaves with 2 cm of thickness and 0.6 cm width, was made and can be attached to a medical LINAC. In this work, measurements using films and ion chamber were performed to investigate electron beam characteristics with eMLC. The source surface distance is 70 cm, at which the distance between the bottom of eMLC and the phantom surface is 5.5 cm. The beam characteristics, such as leakage and transmission, spatial resolution, beam abutment, beam penumbra and percentage depth dose (PDD) were evaluated. Except that the PDDs were measured in water, all other measurements were conducted in a solid water phantom. Results: Leakage and transmission increases from 0.24% to 3% as the electron energy increases from 6 MeV to 20 MeV. The shape of one leaf can be fairly distinguished while that of 2 or 3 leaves can be clearly seen in the dose profile on the phantom surface. The difference between an abutting field sand a single field is obvious on the phantom surface but tends to diminish at the treatment depth. The beam penumbra is of a comparable quality with that shaped by an electron cutout, and in general, it decreases with the increase of energy and increases with the increase of the depth. The therapeutic range and the bremsstrahlung magnitude obtained from the measured PDDs with eMLC are similar to the electron beams shaped by an electron applicator and cutout. Conclusions: Characteristics of the electron beams collimated by eMLC with a thin air gap (about 5.5 cm) is investigated. Acknowledging these characteristics is essential for optimizing treatment plans of the MERT.
34(2007); http://dx.doi.org/10.1118/1.2761730View Description Hide Description
Purpose: This work investigates the use of Gafchromic EBT film for dosimetry in homogeneous and heterogeneous phantoms for static and IMRT fields. This information is necessary for thorough algorithm verification for static and IMRT fields. Method and Materials: EBT film was used to measure dose in three phantom configurations. The first configuration consisted of solid water. The second configuration consisted of 6 cm thickness of lung‐equivalent slabs sandwiched between solid water slabs. Film measurements were compared to EDR film and calculations with the Dose Planning Method (DPM) Monte Carlo code. The third configuration, non‐slab, consisted of solid water with an 8×10×2 cm3 lung‐equivalent region in the solid water. EBT film was placed perpendicula and parallel to the beam within and between materials. Measurements were made for static fields and sample IMRT fields. Results: When converted to dose, EBT measurements showed good agreement with EDR film. For an example IMRT field, agreement was between 86% and 83% for EBT to EDR and EBT film to DPM calculations comparisons, respectively, using a χ evaluation of 2%/2 mm. In the non‐slab inhomogeneous phantom, the EBT film clearly showed the effect of disequilibrium at the interfaces. Conclusions: Because EBT film is not light sensitive, EBT film is a practical choice for phantom measurements that require precise placement. In addition, EBT film can be used for measurements parallel to the beam because it is less dependent on the photon energy spectrum. Tentative results indicate that EBT can be used for dosimetry for static and IMRT fields at planes perpendicular to the beam for homogeneous and heterogeneous slab geometries. Results show that the use of EBT film at interfaces is promising in providing dosimetric data to verify dose calculations in previously difficult to measure geometries.
TH‐D‐M100E‐04: Evaluation of MVCT Images Containing Lead Alloy Masks for Electron Beam Treatment Planning34(2007); http://dx.doi.org/10.1118/1.2761731View Description Hide Description
Purpose: To evaluate the accuracy of electron beam dose calculations in MVCT images containing lead alloy masks. Method and Materials: A phantom consisting of two 30×30×5‐cm3 slabs of CIRS plastic water® was imaged using kVCT (GE Lightspeed‐RT) and MVCT (TomoTherapy Hi⋅Art). The MVCT scans were taken with nine square masks of Cerrobend® (density = 9.4gcm−3) on top of the phantom. The masks contained square openings of 3×3cm2, 6×6cm2 and 10×10cm2 and had thicknesses of 6mm, 8mm and 10mm. The same collimation was simulated in the kVCT images by creating regions‐of‐interest (ROI) duplicating the sizes, shapes, and density of the masks. Using the Philips Pinnacle3treatment planning system, twelve treatment plans were created using electron energies of 6, 9, 12, and 16 MeV for each opening size. For each plan, the mask thickness appropriate for the electron energy was used and the dose distributions calculated using the kVCT and MVCT images were compared. In uniform dose regions (doses above 90% of maximum) dose differences were calculated; in high‐dose gradient regions (doses below 90% of maximum) distances‐to‐agreement (DTA) were determined. Results: In the uniform dose region, the maximum difference between the doses in the MVCT images and the doses in the kVCT image was greater than or equal to ±5% for all opening and energy combinations. In the high‐dose gradient region, almost half of the maximum DTA values exceeded 2mm. Analysis of the MVCT images showed that DTA differences were largely due to distortions in the phantom CT numbers caused by the masks. Conclusion: Although Cerrobend® produces dramatically less distortion in MVCT images compared to kVCT images,image distortion is still too great for accurate electron beam dose calculations.
Supported in part by a research agreement with TomoTherapy, Inc.
34(2007); http://dx.doi.org/10.1118/1.2761732View Description Hide Description
Purpose: This study aims to obtain a realistic estimate of proton range calculation uncertainties resulting from the calibration curve of CT‐number to protonstopping power conversion, which determines the accuracy of proton dose distribution prediction. Range in tissue is determined by integration of protonstopping power along its path, and uncertainties on the CTcalibration curve translate directly into ambiguities on the range. Uncertainties of calibration curves arise mainly from dependence of CT number on the size of imaged object. Method and materials: Materials used include an electron density phantom (CIRS) scanned on Brilliance CT (Philips) and irradiated on Proteus235 proton therapy system (IBA). The modular phantom can simulate a head, a medium and a large body and houses inserts made off 13 tissue substitutes. Hounsfield numbers were measured for different phantom configurations. Mean value and standard deviation established for each insert over all configurations were used to generate three calibration curves. Insert material stopping powers were measured and calculated. All curves were applied to patient data and resulted to three proton ranges from the skin to PTV distal edge. Results:CT number uncertainties were found to increase with physical density, from 3 CT numbers standard deviation for lung (0.5g/cm3) to 140 for dense bone (1.8g/cm3). Range uncertainties for different patients, treatment sites and field orientations vary from 1.5 to 2.4% or 1.6 to 4.5mm. Modulation uncertainties, for every case were less than 1mm and depended on bone in the PTV. Conclusion: Range uncertainties due to CT number variability as a result of beam hardening artifacts are significant for tumor local control as well as sparing of neighboring critical structures. Future work includes patient size specific curves that will reduce the range uncertainty.
34(2007); http://dx.doi.org/10.1118/1.2761733View Description Hide Description
Purpose: A new index for scoring treatment plans that unifies the four dosimetry indices of coverage, conformity, homogeneity and dose gradient, into one simple equation, is introduced. We present results of actual clinical cases evaluated with new scoring index, validating its effectiveness. Methods and Materials: We formulated a unified dosimetry index (UDI) that computes for any given treatment plan its deviations in terms of dose coverage, conformity, homogeneity, and dose gradient vis‐á‐vis an ideal plan. We define an ideal plan as one with perfect dose coverage, conformity, homogeneity, and a step‐wise fall‐off to zero dose outside the planning target. To demonstrate effectiveness of the scoring system, a retrospective evaluation of 21 stereotactic radiosurgery cases was conducted. The cases presented were planned on BrainSCAN (by BrainLAB Inc) utilizing 5–8 single‐isocenter non‐coplanar fixed beams collimated with micro multi‐leaf collimator (MMLC). This index is designed to be utilized also for scoring of radiotherapy treatment plans obtained using various other forward and inverse planning techniques including IMRT, and multiple‐isocenters noncoplanar arcs. Results: For most of the cases evaluated, conformity and dose gradient were the two dominant components. The contribution from dose homogeneity was significant only in a few cases. All plans received very close to full dose coverage, and as a result the contribution from dose coverage component to the overall score was negligible. However, for inverse planning the contributions from dose coverage and homogeneity components to the overall score are expected to be quite more significant and variable. Conclusion: The unified dosimetry index is demonstrated as an effective tool for scoring treatment plans. Utilizing the mean and standard deviation of UDI scores from a pool of treatment plans of similar modality and planning technique, we suggest a general guide for ranking treatment plans as “excellent”, “good”, “average”, or “poor”.
TH‐D‐M100E‐07: Determining the Geometric and Dosimetric Accuracy of MRI Based IMRT Treatment Plan for Patients with Prostate, Brain, and Head and Neck Cancers34(2007); http://dx.doi.org/10.1118/1.2761734View Description Hide Description
Purpose: To determine the geometric and dosimetric accuracy of MRI‐based IMRT treatment plan for patients with prostate, brain, and head and neck cancers.Methods and Material:CT simulation images were obtained for prostate, brain and head and neck cancer patients in treatment position with required immobilization. T1 and T2 weighted MRIimages were also obtained for these patients. Contours for organs‐at‐risk (OARs) and planning target volumes (PTVs) from CTimages were copied to the MR images. Contours created for OAR and PTV in MR images were verified by a radiation oncologist. The intensity map from CT‐based plans and the contoured MR images were used to create MR‐based treatment plans with both homogenous and inhomogeneous tissue density (average HU from CT was assigned to MR structure). Treatment plans (Varian Eclipse TPS) were created by assigning water equivalent homogenous tissue density to CT‐based plan compared to CT‐based plan with inhomogeneity correction. We evaluated the geometric accuracy of MR‐based plans by determining the percent difference of structure volumes between CT and MR‐based contours. Furthermore, we evaluated the dosimetric accuracy of MR based treatment plans by comparing DVH, isodose lines, absolute dose and MUs. Results: The mean absolute dose difference between approved CT plan and one with homogenous tissue density in CT was < 1% for the anatomical sites. Volume differences between MR based and CT based contours were as great 25% ± 5 and dose differences varied 20% ± 1 5 depending on the initial structure volume. Less than 2% MU variation was observed between CT and MR‐based treatment plans using contours from CT and with uniform or non‐uniform electron density. Conclusion: Homogenous tissue density could be used for MR‐based treatment planning for prostate and head cases. Further, better dose coverage equivalent to CT‐based plan could be achieved by optimizing structures contoured on MR images.
34(2007); http://dx.doi.org/10.1118/1.2761735View Description Hide Description
Purpose: A technique for beam angle optimization based on minimization of field aperture eccentricity is introduced. A proof of concept of this new technique is presented. Methods and Materials: By selecting beam angles with minimal aperture eccentricities one can improve the likelihood of achieving a well optimized dosimetry plan. A test of our hypothesis requires a method for quantifying beam aperture eccentricity. We introduce a new concept based on the major and minor axes of the beam's eye view (BEV) projections of the field apertures. A coefficient of variance defined as the ratio of standard deviation and mean of the major and minor axes of apertures for any set of fields utilized for dose optimization is defined as a way to quantify field aperture eccentricity. To prove our concept we rank 21 stereotactic radiosurgery treatment plans utilizing a unified dosimetry index (UDI) presented in a separate paper. We then correlate the dosimetry scores to the coefficient of variance that quantifies aperture eccentricity for the selected set of beam angles utilized for dose optimization of the respective treatment plans. Results: The UDI scores of 21 radiosurgery treatment plans are plotted as a function of corresponding coefficient of variance (of aperture eccentricities of field sets utilized for dose optimization in each case). The computed linear correlation coefficient value of 0.377, is high enough to indicate that the UDI scores are somewhat correlated to CV. There is not a complete positive correlation (R2 value of 1.0) between the UDI scores and the beam aperture eccentricity. Conclusion: There is not a complete positive correlation (R2 value of 1.0) between the UDI scores and the beam aperture eccentricity. However, because other factors such as the number of beams utilized and planning technique also impact the quality of the dosimetry plan, a perfect correlation is not expected.
34(2007); http://dx.doi.org/10.1118/1.2761736View Description Hide Description
Purpose: In treatment planning systems, it is common to view anatomical structures or dose levels using segmented boundary curves. While such segmentations can be provided either manually or automatically on a number of transverse planes, interpolation has to take place on in‐between or non‐transverse planes. Naïve image‐basedinterpolation by blending voxel values converted from provided segmentations produces artifacts in locations of dramatic change of segmentations between planes or due to inconsistencies in manual segmentations. We propose a geometry‐based interpolation approach that significantly reduces these artifacts by building and slicing 3D meshes. Method and Materials: The provided segmentations are first converted into labeled voxels (e.g., structure names, dose levels). We then use standard iso‐contouring techniques to extract surfaces bounding voxels with a common label, followed by iterative Laplacian‐smoothing. The intersection of the smoothed mesh and an arbitrary view plane is then calculated to obtain the interpolated segmentation on that plane. To accelerate the mesh generation process for multiple structures or dose levels, an interval‐tree optimization is introduced. Results: Mesh extraction, smoothing and plane‐mesh intersection have been implemented in C++ for efficiency, which is interactively invoked within CERR (coded in MATLAB). On a 1.4 GHz processor, the initial surface extraction and smoothing takes between 2–5 seconds for a single structure or dose level. Once the surfaces are generated, building intersections with view planes (and generating interpolated contours) takes negligible time and allows interactive viewing. Our smooth contours have dramatically reduced jaggedness comparing to naïve image‐basedinterpolation at locations of sharp dose changes or between inconsistent physician‐drawn contours, as demonstrated using clinical treatment plans. Conclusion: The geometry‐based interpolation approach improves the visual quality as well as accuracy over naïve image‐basedinterpolation techniques for viewing segmentations or dose levels on in‐between and non‐transverse planes without a significant sacrifice of running time.
- Radiobiology: Treatment Planning and Evaluation
TH‐E‐M100E‐02: Superiority of Equivalent Uniform Dose (EUD)‐Based Optimization for Breast and Chest Wall IMRT34(2007); http://dx.doi.org/10.1118/1.2761772View Description Hide Description
Purpose: To investigate whether IMRT optimization based on generalize equivalent uniform dose1 (gEUD) objectives for target volumes and organs at risk (OAR) alike can lead to superior plans as opposed to multiple dose‐volume (DV) based objectives plans, for intact breast and postmastectomy chest wall (CW) cancer.Methods and Materials: Four IMRT plans with six or seven coplanar 6‐MV beams were prepared for a number of chest wall and breast CA patients (10 patients). The first three plans utilized our standard in‐house physician‐set of DV objectives (phys‐plan), gEUD‐based objectives for the OARs (gEUD‐plan), and multiple, “very stringent”, DV objectives for each OAR and PTV (DV‐plan), respectively. The fourth was only beam fluence optimized plan (FO‐plan), without segmentation and utilized the same objectives as in the DV‐plan. The latter plan was to be used as an “optimum” benchmark without the effects of the segmentation for deliverability. Various dosimetric quantities, such as mean dose (Dmean) for heart, contralateral breast, and contralateral lung; and V20 (volume of organ receiving 20Gy) for the ipsilateral lung were employed to evaluate our results. Results: For all patients in this study, we have seen that the gEUD‐based plans allow greater sparing of the OARs while maintaining excellent target coverage. The use of gEUD allows selective optimization of the dose for each OAR and results in a truly individualized treatment plan. Conclusions: gEUD requires a smaller number of parameters for optimization and allows exploration of a much wider space of solutions, thereby making it easier for the optimization system to balance competing requirements in search of a better solution. Thus, gEUD optimization can be used to search for or evaluate plans of different DVHs with the same gEUDs. This method can be efficiently used in routine clinical IMRTtreatment planning.