Home / Our Team / Publications


Commissioning Experience Of A Superficial kV System for Guinea Pig Cutaneous Irradiation

Saikanth Mahendra1, Adam Schoen1, Karla D. Thrall2
1Northwest Medical Physics Center, Lynnwood, WA; 2SNBL USA, Ltd., Everett, WA

Research Poster presented at 63rd Radiation Research Society (RRS) Annual Meeting, Cancun, Mexico, (Oct 2017)

Cutaneous radiation injury (CRI) studies in guinea pigs are considered a good small animal model to assess the human dermal response to low-penetrating acute irradiation exposures for prospective medical countermeasures (MCMs). The characteristics of superficial Xstrahl 100 kV irradiator were utilized for development of dose-response models and key dosimetric parameters are presented. In order to mimic the depth of penetration and dose fall off under the skin similar to a pure beta source, we selected a 60kV beam potential, a beam current of 6mA and a custom 0.15mm inherent filter setting. The irradiation area was achieved by a 4x4cm square applicator placed in contact with the skin. All measurements were performed in accordance with the AAPM Task Group 61 (TG-61) protocol. A soft X-ray parallel plate chamber was utilized in air and in a Soft X-ray slab phantom to determine different beam characteristics. The Half Value Layer (HVL) of the beam was measured in air and compared using three different purities of Aluminum (Al) filters. Al purity of 99.0%, 99.99% and 99.999% yielded 0.163, 0.171, 0.161 mm of Al as HVL’s respectively. The Percent Depth Dose (PDD) at 50% was found to be 1.884mm and less than 30% at 4mm in water equivalent plastic slabs. A steep dose fall off was desired to limit the dose to deeper tissues to decrease any normal tissue complications. The absolute beam output was determined to be 6.532 Gy/min at surface, 4.290 Gy/min at 1mm depth and 3.165 Gy/min at 2mm depth. Beam consistency and constancy check measurements were performed with minimal coefficient of variation (0.1%). GafChromic EBT-3 film was used to determine dose homogeneity across the irradiated area and was found to be well within the required +/- 5%.

The Radiation Dose Response Relationship for Whole Thorax Lung Irradiation in the Female Rhesus Macaque

Karla D. Thrall1 , Saikanth Mahendra2, Thomas J MacVittie31SNBL USA, Ltd., Everett, WA; 2Northwest Medical Physics Center, Lynnwood, WA; 3UDept. Of Radiation Oncology, University of Maryland, School of Medicine, Baltimore, MD

Research Poster presented at 63rd Radiation Research Society (RRS) Annual Meeting, Cancun, Mexico, (Oct 2017)

Concerns over nuclear and radiological threats have prompted the need to improve methods to protect the general population from the health hazards associated with exposure to ionizing radiation. The development of effective medical countermeasures requires efficacy studies be conducted in an appropriate animal model (“the Animal Rule” 21 CFR 314.600 for drugs). However, the reliance on non-human efficacy data places an enormous importance on appropriately developed and validated animal models. A whole thorax lung irradiation (WTLI) model has been developed for the male rhesus macaque (Garofalo et al. 2014). The objective of this study was to develop a lethality dose response profile for WTLI specific to this institution, an endeavor necessary to validate the model prior to conducting efficacy studies under the criteria of the Animal Rule. The study design was similar to that described previously, with the exception that female, rather than male, rhesus macaques were utilized. In brief, a computed tomography (CT) scan was conducted prior to irradiation and used for treatment planning. Animals were exposed to a single 6 MV photon exposure focused to the lung as determined by the CT scan and treatment planning at a dose of 9.5, 10, 10.5, 11 or 11.5 Gy. Supportive care, including administration of dexamethasone, was based on trigger-to-treat criteria and clearly defined euthanasia criteria were used to determine moribund condition over 180 days post irradiation. Percent mortality per radiation dose was 12.5% at 9.5 Gy exposure, 25% at 10 Gy exposure, 62.5% at 10.5 Gy exposure, 87.5% at 11 Gy exposure, and 100% at 11.5 Gy exposure. The resulting probit plot for the WTLI model estimated a 180 day LD50 of 10.28 Gy, which compared well with the previously published estimate of 10.27 Gy for the male rhesus. Additionally, WTLI allows longitudinal definition of combined organ injury to both lung and heart. These data suggested the absence of a gender influence on the radiation dose response for WTLI in the rhesus and provided an inter-laboratory validation of the previously reported model.

Radiation physics and dosimetry of 6MeV linear accelerator beam for swine cutaneous irradiation

Saikanth Mahendra, Karla D. Thrall
Northwest Medical Physics Center, Lynnwood, WA; SNBL USA, Ltd., Everett, WA

Research Poster presented at 62nd Radiation Research Society (RRS) Annual Meeting, Big Island, Hawaii, (Oct 2016)

The characteristics of a linear accelerator generated 6MeV radiation beam are well suited for development of radiation-induced skin injury models. Based on literature, swine has been widely accepted as the closest skin model compared to humans. To study the evolution of cutaneous radiation injuries, the lowest electron energy – 6MeV, available in our Varian 21EX linear accelerator was preferred, taking the below parameters into consideration – high surface dose deposition within the first few millimeters, uniformity of dose distribution across the irradiation site and practical range (Rp) of the beam. The dorsal skin surface of three different porcine breeds – Yorkshire, White Sinclair and Göttingen were exposed to a single fraction of 25, 35, 45, and 60Gy to two 4-cm circular fields using a nominal dose rate of 4Gy/min. Ultrasound measurements of swine skin thickness enabled us to employ a 0.8cm superflab bolus material to ensure approximately 95% dose homogeneity from the swine skin surface to a depth of 0.7cm. The depth of maximum dose (dmax) was found to occur around 0.5cm below the swine skin surface. A steep dose fall off beyond the dose deposition depth of 0.8cm was desired to limit the dose to deeper tissues to decrease any normal tissue complications. Relative doses of less than 50% at 1.5cm, 15% at 2cm and less than 5% at 2.2cm below the swine skin surface were observed with virtual water phantom measurements which mimic tissue. The percentage depth-dose dosimetry was characterized using a small-volume parallel-plate ionization chamber which serves as an ideal choice in measuring doses near to the skin surface. The output factors and uniformity profiles were measured using Electron diodes for desired applicators. A 20×20 applicator cone was used along with 3.2mm Lead (Pb) surface cutouts to minimize the radiation penumbral effects on the skin ensuring a more defined irradiation boundary. The variability of beam uniformity – both flatness and symmetry across the 4cm circular irradiated field was observed to be less than 1%. Optically Stimulated Luminescence (OSL) dosimeters were placed on the dorsal surface to evaluate dose delivery and below the lead cutouts to evaluate shielding efficacy and average measured differences were less than 2% and 0.25% respectively.

A Cross-Breed Comparison of Cutaneous Radiation Injury in Swine

Karla D. Thrall, Ronald Manning, Saikanth Mahendra,George De Los Santos, Koichiro Fukuzaki, Ryoichi Nagata
SNBL USA, Ltd., Everett, WA, Northwest Medical Physics Center, Lynnwood, WA; SNBL USA, Ltd., Kagoshima, Japan

Research Poster presented at 61st Radiation Research Society (RRS) Annual Meeting, Weston, FL, (Sep 2015)

Development of medical countermeasures to treat radiation-induced injuries requires an appropriate animal model. For skin burn injuries, the pig is widely accepted as the best model, and the literature contains numerous examples of the use of pigs in the study of thermal and radiation burns. However, no single breed of pig has emerged as a preferred model. In the present study, the evolution of radiation-induced skin injury was evaluated in three porcine breeds. Age matched female Yorkshire, white Sinclair, and Göttingen minipigs were exposed to a single fraction of 25, 35, 45, and 60 Gy with 6-MeV electrons to a 4×4-cm field on the dorsal skin surface of the animal (4 fields per animal, 1 dose per field, and 4 doses per animal). Ultrasound was used to estimate differences in skin thickness at each irradiation field and skin was overlaid with tissue equivalent bolus material to ensure that greater than 90% of the prescribed dose was limited to a depth of 2 cm. Furthermore, bolus material was varied appropriately to ensure uniform depth of dose delivery across all irradiation locations for all animals, regardless of breed. Following irradiation exposure, animals were observed and skin was numerically scored for erythema and desquamation separately over a sufficient period of time to allow full progression of injury. Skin scores were used to compare qualitative observations of gross changes, including erythema, dry desquamation and moist desquamation. A comparison of skin scores indicate clear differences between breeds in both degree of injury, as evidenced by higher skin scores, as well as in the time course of onset and progression. Based on skin scores per delivered dose, the Göttingen minipig appears to be more radioresistant than either the Yorkshire or Sinclair pigs. In summary, the study reported here provided a unique side-by-side comparison of radiation-induced skin injury over a range of doses across three commonly utilized research breeds.

Low Energy Therapeutic X-Ray Calibration Methods

M. M. Zaini,  A. R. Schoen, L. M. Arvan, E. I. Marshall, G. E. Robertson, C. G. Beck, A. L. Eagle, L. E. Sweeney
Northwest Medical Physics Center, Lynnwood, WA

World Congress on Medical Physics and Biomedical Engineering, 90, Toronto, Canada (2015).

NMPC also chaired this scientific session.

The application of low energy X-rays has recently increased in the dermatology workspace, with the generating tube potential of these radiation beams spanning from 10 to 70kV. Absolute calibration of these low energy radiation units from three different manufacturers has been conducted. Various low-energy calibration protocols were considered for calibrating the low energy beams of the X-ray tubes from three different manufacturers. These protocols included the American Association of Physicists in Medicine (AAPM) Task Group 61 Report (for 40-300 kV beams), the UK Code of Practice from the Institution of Physics and Engineering in Medicine and Biology (IPEMB) for very low-energy X-rays (8-50 kV), International Atomic Energy Agency (IAEA) Technical Reports Series No. 398 (for X-ray up to 100 kV), and Report 10 of the Netherlands Commission on Radiation Dosimetry for low energy X-ray (50-100 kV). Review and comparison of the methodology of these protocols, pertaining to the energy range of Grenz-rays to superficial X-rays, are conducted.

Superficial Therapy: More Than Just Skin Deep

E. Marshall, M. Zaini, A. Schoen, L. Arvan, G.Robertson
Northwest Medical Physics Center, Lynnwood, WA

NWAAPM Spring 2015 meeting, Young Investigators Presentations. Portland, Oregon (2015).

Book Chapter: “Brachytherapy Physics” in “Comprehensive Biomedical Physics”

B. Thomadsen,1 R. Miller,2
1 University of Wisconsin-Madison, Madison, WI; 2 Northwest Medical Physics Center, Lynnwood, WA

Anders Brahme, editor-in-chief, Vol 9, Amsterdam: Elsevier, 315-381 (2014). This book is in press.

Brachytherapy is the treatment of disease using small radioactive sources placed in or near the tissues to be treated.    This chapter explores some of the most common sources used for brachytherapy as well as their therapeutic applications.  The delivered dose rate is one means utilized to categorize brachytherapy treatments into low, medium or high dose-rate delivery.  Further characterization is based on how long the sources are implanted, either as a temporary or a permanent implant.  Several clinical examples are reviewed.  Calculation methods, optimization and quality management are also discussed.

Calypso(R) and Laser-Based Localization Systems Comparison for Left-Sided Breast Cancer Patients Using Deep Inspiration Breath Hold

S. Robertson,1 D. Kaurin,1,2 J. Kim,2,3 L. Fang,2,3 L. Sweeney,1,2 K. Halloway,2 A. Tran2,3
1 Northwest Medical Physics Center, Lynnwood, WA; 2 Seattle Cancer Care Alliance, Seattle, WA; 3 University of Washington Medical Center, Seattle, WA

Medical Physics, 41, 162 (2014).

Purpose: Our institution uses a manual laser-based system for primary localization and verification during radiation treatment of left-sided breast cancer patients using deep inspiration breath hold (DIBH). This primary system was compared with sternum-placed Calypso(R) beacons (Varian Medical Systems, CA). Only intact breast patients are considered for this analysis.

Methods: During computed tomography (CT) simulation, patients have BB and Calypso(R) surface beacons positioned sternally and marked for free-breathing and DIBH CTs. During dosimetryplanning, BB longitudinal displacement between free breathing and DIBH CT determines laser mark (BH mark) location. Calypso(R) beacon locations from the DIBH CT are entered at the Tracking Station. During Linac simulation and treatment, patients inhale until the cross-hair and/or lasers coincide with the BH Mark, which can be seen using our high quality cameras(Pelco, CA). Daily Calypso(R) displacement values (difference from the DIBH-CT-based plan) are recorded.The displacement mean and standard deviation was calculated for each patient (77 patients, 1845 sessions). An aggregate mean and standard deviation was calculated weighted by the number of patient fractions.Some patients were shifted based on MV ports. A second data set was calculated with Calypso(R) values corrected by these shifts.

Results: Mean displacement values indicate agreement within 1±3mm, with improvement for shifted data (Table).
Conclusion: Both unshifted and shifted data sets show the Calypso(R) system coincides with the laser system within 1±3mm, demonstrating either localization/verification system will Resultin similar clinical outcomes. Displacement value uncertainty unilaterally reduces when shifts are taken into account.

Output Constancy: Reducing Measurement Variations in a Large Practice Group

K. Hedrick, T. Fitzgerald, and R. Miller
Northwest Medical Physics Center, Lynnwood, WA

Medical Physics, 41, 282 (2014).

A standardized linear accelerator output constancy check for both photons and electrons with acrylic phantoms was instituted within the practice group in 2010. The goal of this process is to eliminate small errors caused by differences in technique and equipment within a practice group that encompasses multiple sites (>20) and many physicists (>30). The resulting Kacrylic values and output data are sent to a project manager for review and auditing purposes. The results are published to the group monthly. The improvements seen in the consistency of output measurements as well as the reporting system enable us to quickly identify potential errors by investigating outlying Kacrylic and output values.


Radiation Physics of a New Electronic Brachytherapy System for Treating Skin Lesions

M. M. Zaini,  A. R. Schoen
Northwest Medical Physics Center, Lynnwood, WA

International Journal of Radiotion Oncology•Biology•Physics, 90 (1), S928 (2014).

Purpose/Objectives: To commission and analyze a new low energy high dose rate electronic brachytherapy system.

Materials/Methods: Half-value layer (HVL), absolute dose output, beam profiles, and insert factors of an electronic brachytherapy system (EBT) were measured and analyzed. A thimble-shaped parallel-plate ionization chamber was utilized for HVL, absolute output, and insert factor determinations. HVL measurements of the 69.5 keV beam of EBT were conducted using thin aluminum foils in narrow beam geometry. In this setup, the aluminum filters were positioned at a distance of 30 cm from tip of the EBT device and 30 cm from the ion chamber. Absolute calibration of this X-ray beam was carried out based on the AAPM Task Group 61 (TG61) report. Beam profiles and depth ionization curves were measured and analyzed using radiochromic films, a professional scanner, and a scientific software package. The relative output factors of the five standard inserts of this device (10, 15, 20, 25, and 30 mm in diameter) were determined employing the ion chamber in air. Finally, a secondary dose calculation system was developed in a spreadsheet.

Results: HVL of the X-ray beam of EBT was measured to be 1.61 mmAl. For TG61-based calibration of this low energy kV beam employing the ion chamber at the SSD of 6 cm, it was determined that for this system Pion, Ppol, backscatter factor, ratio of mass energy absorption coefficient, inverse square correction from the surface to the effective point of measurement, and end-effect are 1.009, 0.995, 1.121, 1.017, 1.061, and 0.997, respectively. Measuring the PDD of this beam was proved difficult. Per TG61 recommendations, PDD values from BJR supplement #25 were employed for translating the measured output at the surface to the depth of 3 mm in tissue. The PDD values from BJR were in agreement with those supplied by the manufacturer to within <0.5%. The absolute dose at the depth of 3 mm was therefore established to be 3.1 Gy, for when the EBT system was programmed to deliver 3 Gy. This agreement level was deemed acceptable for the low-energy and small SSD of this device. Beam profiles showed a relatively flat beam (<5%) with small (1 mm) penumbra. Insert factors varied from 0.95 to 1.01 for the five standard inserts of this EBT device in relations to the 30 mm insert output value.

Conclusions: Radiation physics parameters of an EBT X-ray unit were measured to match those claimed by the manufacturer to an accuracy of better than 5%, except HVL. Our measured HVL was 1.61 mm Al, whereas that indicated by the manufacturer is 1.83 mm Al. This discrepancy in HVL has a 0.3% effect on the NK value used for this chamber and less than 2% impact on the PDD employed for determining the absorbed radiation dose at depth. Reproducibility of the radiation beam of this device was on the order of less than 0.3%.


A Liquid Ion Chamber For Clinical Stereotactic Output Measurements

J. Mathews,1,2 P. Stevens,1,2 M. Zaini1,2
1 Northwest Medical Physics Center, Lynnwood, WA; 2 Skagit Valley Regional Health Center, Mount Vernon, WA

International Journal of Radiotion Oncology•Biology•Physics, 90 (1), S736 (2014).

Purpose/Objectives: To measure output factors produced by the liquid-filled ion chamber at different field sizes and use the data to evaluate the liquid-filled ion chamber’s performance in a clinical setting compared to its gas-filled counterparts.

Materials/Methods: The setup of this experiment closely followed the method used to obtain the clinical data. The liquid-filled ion chamber’s cavity was positioned at isocenter (SSD = 90cm and d = 10cm) in a water tank on an Elekta Axesse medical linear accelerator. An HV power supply was set to –500V and connected to the liquid-filled ion chambervia an electrometer.The liquid-filled ion chamber had a nominal bias voltage of +/-800V (with +/-400V being the absolute minimum), but, due to equipment and budget constraints, only a -500 V bias could be applied.  The ensuing collection inefficiency was negligible for the relative measurements of this study.

The liquid-filled ion chamber was pre-irradiated. Various rectangular fields were made by the MLC.   This experiment was repeated at different timesin order to ensure accuracy and repeatability.The raw liquid ion chamber data were corrected by the published techniques for these types of measurements as well as the Daisy Chain method.

Results: Reproducibility of the measured results with this liquid ion chamber was better than 1%.   There was a statistically significant difference between the clinical stereotactic output values and those of this liquid-filled ion chamber data.    The clinical data were collected using four different chambers and diodes in multiple sessions by various physicists.   These clinical output factors were also congruent with the published results for this linac type.    The differences between the small stereotactic output factors for three photon beams of 6, 10, and 15 MV was determined using the liquid ion chamber versus those of the clinical values ranged between 0 and 7% for rectangular field sizes of 2 X 2 cm down to 0.8 X 0.8 cm.    The largest discrepancies were observed for the smallest field sizes.

Conclusions: Although the liquid-filled ion chamber provided accurate output factors for large fields when compared to clinical data, it failed to do so for small fields (1.2×1.2cm and smaller). The inaccuracies of the stereotactic output factors measured with the liquid-filled ion chamberdid not seem to warrant its advantages of high signal and small collection volume.


Total Body Irradiation (TBI) Optically Stimulated Luminescence In Vivo Dosimetry

C. Holloway,1 S. Mahendra1, D. Kaurin,1, Larry E. Sweeney1
1 Northwest Medical Physics Center, Lynnwood, WA

Medical Physics, 40, 223 (2013).

Purpose: To adequately monitor the doses for Total Body Irradiation (TBI) patients using opposed fields, we carry out both active and passive in vivo dosimetry, with the passive dosimeter being Optically Stimulated Luminescence (OSL) detectors. The OSLs are embedded in hemispherical brass buildup caps with the OSL/flat surface placed against patient skin. The in vivo dosimetry devices are not repositioned when the patient is rotated, so both entrance and exit dosimetry are carried out, which arguably gives a better midline dose estimate, and improves marrow graft acceptance by minimizing treatment time. The OSLs were calibrated using a 100cm SAD entrance method with brass buildup caps. The buildup caps have a maximum density thickness of 6.6g/cm2; dmax for our 18MV beam is 3.4g/cm2. This method results in an overestimation of exit dose due to the density thickness differences, requiring a correction factor for exit measured doses.

Methods: To determine the exit dosimetry correction factor, OSLs were placed on the entrance and exit sides of solid water phantoms under TBI conditions having thicknesses of 10, 30, and 49.2cm. Combinations of OSLs included entrance‐only, exit‐only and combined entrance and exit measured doses. The entrance and exit dose to the dosimeter position was calculated.

Results: The exit dose correction factor calculated from the ratio of calculated to measured values was larger than anticipated (∼17%). The correction factor was applied to the exit dose calculations and compared to the combined measurement for verification. Results were within 1.2% (Table 1).The method was tested on ten patients, giving agreement within 1.1±4% (Table 2). Acceptable TBI tolerance is ±10%. Conclusion: While this is acceptable, we will consider using non‐brass buildup caps to obtain smaller correction factors.


VMAT testing for an Elekta accelerator

Darryl G.L. Kaurin,1,2 Larry E. Sweeney,1,2 Edward I. Marshall,1,2 Saikanth Mahendra1,2
1 Northwest Medical Physics Center, Lynnwood, WA; 2 Seattle Cancer Care Alliance – Radiation Oncology, Seattle, WA

Journal of Applied Clinical Medical Physics, 13(2) (2012).

Volumetric-modulated arc therapy (VMAT) has been shown to be able to deliver plans equivalent to intensity-modulated radiation therapy (IMRT) in a fraction of the treatment time. This improvement is important for patient immobilization/localization compliance due to comfort and treatment duration, as well as patient throughput. Previous authors have suggested commissioning methods for this modality. Here, we extend the methods reported for the Varian RapidArc system (which tested individual system components) to the Elekta linear accelerator, using custom files built using the Elekta iComCAT software. We also extend the method reported for VMAT commissioning of the Elekta accelerator by verifying maximum values of parameters (gantry speed, multileaf collimator (MLC) speed, and backup jaw speed), investigating: 1) beam profiles as a function of dose rate during an arc, 2) over/under dosing due to MLC reversals, and 3) over/under dosing at changing dose rate junctions. Equations for construction of the iComCAT files are given. Results indicate that the beam profile for lower dose rates varies less than 3% from that of the maximum dose rate, with no difference during an arc. The gantry, MLC, and backup jaw maximum speed are internally consistent. The monitor unit chamber is stable over the MUs and gantry movement conditions expected. MLC movement and position during VMAT delivery are within IMRT tolerances. Dose rate, gantry speed, and MLC speed are accurately controlled. Over/under dosing at junctions of MLC reversals or dose rate changes are within clinical acceptability.


A Retrospective Review of Target Motion for Prostate Patients using Calypso

M. Fredrickson, L. Arvan, G. Courlas, L. Sweeney, D. Kaurin
Northwest Medical Physics Center, Lynnwood, WA

We retrospectively studied target motion in prostate IMRT patients using the Calypso System (Varian Medical Systems, Palo Alto, CA) by tracking the number of setup adjustments for each patient in their first 5 and last 5 fractions. Additionally, the time of the first adjustment was recorded for each fraction that was interrupted.

Medical Physics, 39, 3780 (2012).


Temporal Analysis of Inter-Fraction In-Vivo 3D Dosimetry Using DVH Difference Curves Based on Daily Exit Fluence Measures and Cone-Beam CT Images

M. M. Zaini,1,2 G.A. Sandison2
1 Northwest Medical Physics Center, Lynnwood, WA; 2 University of Washington, Seattle, WA

International Journal of Radiotion Oncology•Biology•Physics, 84 (3), S747 (2012).

Purpose/Objectives: Cancer patients undergoing highly fractionated radiation therapy may experience changes in the relative position of their tumor and surrounding normal tissues during the course of their therapy.  These temporal changes impact the planned distribution of radiation, and it is vital that the attending radiation oncologist continually assess information on these changes to decide whether a modification to the patient’s treatment plan is warranted.  The purpose of our study is to assist the attending in this assessment by providing 4D (temporal and spatial) patient dose distribution changes based on exit fluence measures overlaying CBCT images of the patient’s anatomy presented at the time of each treatment fraction.

Materials/Methods: Exit fluence was collected with the EPID during each treatment fraction and back-projected onto the daily acquired CBCT image volumes.  Actual daily delivered dose was computed based on the daily CBCT by the Math Resolutions, DosimetryCheck® software.  Any patient shift corrections based on the registration of the daily and planning CT volumes prior to administering the radiation treatments were incorporated in the 3D dose calculations.  Processing of daily CBCT images involved the demarcation of the tumor.   Daily DVH of the 3D in-vivo tumor dose was compared against that of the originally planned IMRT or VMAT treatment.

Results: Daily tumor DVH deviated from the planned DVH on different days.  Subtraction of the daily DVH curves and the planned one was performed to generate a difference curve. The area under the planned DVH curve was then compared to this difference curve.  The ratio of the absolute difference integrals and the planned DVH integral varied from 0.5% to 4.5%.  The running sums of the absolute difference between the daily DVH curves and the planned DVH were also computed.  Fluctuations in the daily DVH curves compared to the planned one did not result in target dose coverage changes by more than 5% in four of the patients studied.  The running sum difference curves diminished in value over the course of treatment indicating that daily fluctuations in tumor DVH are more random than systematic in terms of dose coverage.

Conclusions: An attending may easily assess daily whether corrections to planned treatment are required based on DVH difference curves between planned and actual treatment dose distributions.  Although the visual inspection of the slice-by-slice daily tumor coverage showed some positional dependence of the dose coverage deficiencies, it illustrated that the daily variation of dose is predominantly disseminated throughout the tumor.  Absolute differences in daily DVH curves provides an accurate measure of daily variation in dose deposition but running sum differences may overestimate the biological impact of this anatomic-based treatment inaccuracy.


VMAT Couch Attenuation Model Testing for Prostate Plans

Darryl Kaurin, Ph.D.,1,2 Edward Marshall B.S.,1 M. M. Zaini, Ph.D.,1,2 Larry Sweeney, Ph.D.,1,2 Saikanth Mahendra, M.S.1,2
1 Northwest Medical Physics Center, Lynnwood, WA; 2 Seattle Cancer Care Alliance, Seattle, WA

International Journal of Radiotion Oncology•Biology•Physics, 81 (2), S845 (2011).

Purpose/Objective(s): Our institution has two accelerators that are beam matched, one having a kevlar couch and the other carbon-fiber. We were concerned about couch attenuation for volumetric modulated arc therapy (VMAT), as well as being able to switch patients to either unit without replanning. The couches were characterized dosimetrically to determine if the couches needed to be included in the treatment plan, and if patients could be treated on either couch without additional planning.

Materials/Methods: Attenuation measurements of both couches were made by positioning an ion chamber with buildup cap on top of the couch and measuring transmission of a 5cmx5cm beam at ten different posterior gantry angles (PA and then every 10 degees up to 10 degrees below the horizontal) for 6MV and 10MV energies, relative to an unattenuated beam. Both couches were CT’d and imported into the treatment planning system (TPS).

Attenuation measurements were simulated in the TPS. The couches were contoured as shells of varying thickness and density, with air in the middle of the couch, as the recommended TPS air threshold density for the entire couch attenuated the beam too much. We had good measurement agreement for a kevlar couch model shell of 3-mmthickness, and the carbon-fiber model shell of 4-mm-thickness, both with 0.65g/cc density. Both couches were stored in the TPS organ model library.

To validate the couch models, five prostate patients (two 6MV, three 10MV cases) with test VMAT plans were copied onto a cylindrical phantom with a central 0.13cc ion chamber and also a rectangular phantom with a 2-demonsional diode array in a sagittal orientation. Doses and fluences were calculated for both phantoms, both couches and no couch (30 plans); and measured (20 measurements).

Results: Ion chamber results comparing the phantom plan with and without the couch gave a 1% improvement in measured-to-calculated values. Planar fluences were evaluated using absolute dose with 3% in 3mm gamma criteria, with comparisons between the phantom plan with and without the couch showing no improvement in measured-to-calculated values (agreement range/average: 93.7-100/98.51% with couch, 97.2-100/99.1% no couch). To further investigate the planar doses, calculated planar doses between the kevlar and carbon-fiber couches were compared to each other; with differences only noticeable using a criteria of 1% in 1mm gamma (agreement range/average: 92.1- 100/97.1%).

Conclusions: Prostate patient treatment plans may benefit about 1% with inclusion of the treatment couch. It will be fine clinically for patients to be treated using the couch they were not planned with for several fractions. These results may not be applicable for other disease-sites that have more of a posterior-beam component.


Four-Dimensional Image Processing of Daily Cone-Beam CT Volumes

M. M. Zaini, Ph.D.,1,2 W. Bryan Jackson, M.S.,1,2 Kellen K. Thuo, M.S.1
1 Northwest Medical Physics Center, Lynnwood, WA; 2 Skagit Valley Regional Health Center, Mount Vernon, WA

Radiological Society of North America, 398 (2011).

Purpose: Daily cone-beam CT-based 3D dosimetry allows the observation of anatomical and dosimetric changes in the patients undergoing radiation therapy.   To derive meaningful anatomical and dosimetric information from the daily changes to the patient, one needs to observe the temporal changes in the three-dimensional cone-beam CT (CBCT) image volumes.   Temporally varying 3D CBCT image volumes constitute a four-dimensional dataset.   Four-dimensional spatiotemporal image processing techniques are employed for studying the three-dimensional anatomical and dosimetric videos of the radiation therapy patients.

Methods: CBCT image volumes of radiation therapy patients are obtained daily.   The exit fluence of 3D-conformal, IMRT, and VMAT radiation beams are acquired using an electronic portal imaging device.    Back-projected fluence images onto the CBCT image volumes employing inverse convolution produce 3D dosimetric maps of the daily radiation treatments.

Organs and volumes of interest (VOI) are contoured on the daily CBCT image volumes.   3D dose values inside the daily normal and disease contours are tracked in time.   Spatiotemporal filters and models are used to assist in extracting clinically relevant information.   Velocity vector topographs are employed for visualization of the temporal changes in the radiation treatment delivery and outcome.   Movies of the dose maps in the VOIs also aid in the clinical decision making process.

Results: Daily changes in the imparted radiation dose and the shape of the VOIs are clearly visible in the four-dimensional dataset of this work.  Spatial extent of the dosimetric shift in some anatomical regions was larger than the estimated values by up to 1.5 cm.   It was found that intraframe filters are not needed for smoothing the sparsely sampled time dimension.

Conclusion: Temporal changes in the treated areas and their dose values are anticipated.  Such expectations arise from the progress of the treatments, daily setup variations, and patient motion.   At some points in the course of the treatment of certain anatomical sites, the need for altering the treatment plan could arise.   Movies are a better visualization tool than both the velocity maps and the temporally modulated dose-volume histogram plots.

Clinical Relevance/Application: A clinician could employ the results of this work in assessing whether an alteration in the treatment plan during the course of a fractionated radiation therapy is warranted.


Anatomical Dose Tracking for Adaptive Radiation Therapy

M. M. Zaini, Ph.D.,1,2 W. B. Jackson, M.S.,1,2 K. K. Thuo, M.S.,1 G. A. Sandison, Ph.D.,3 H. P. Patel, M.S.,1,2 M. T. Luckstead, M.S.,1,2 M. A. Whiton, M.D.,2 L. E. Sweeney, Ph.D.1
1 Northwest Medical Physics Center, Lynnwood, WA;  2 Skagit Valley Regional Health Center, Mount Vernon, WA;  3 University of Washington, Seattle, WA

Cancer Imaging and Radiation Therapy Symposium, 151, 45 (2011).


Photoneutron Activation of an IMRT QA Device and the Radiation Safety Implications

A. Eagle,1 M. Mann1, J. Washington,2, L. Sweeney,2 D. Kaurin,2 S. Qui,3 W. Simon,4 F. Newman5

1 Lutheran Medical Center, Wheat Ridge, CO, 2 Northwest Medical Physics Center, Lynnwood, WA, 3 Univ. Of Colorado, Denver, Aurora, CO, 4 Sun Nuclear Corp, Melbourne, FL, 5 University of Colorado Health Science, Aurora, CO

Medical Physics, 38, 3545 (2011).

Purpose: To characterize the activation of the Sun Nuclear MapCheck2 diode array when exposed to high energy photons.

Methods: A Sun Nuclear MapCheck2 diode array was exposed to 6, 15 and 18 MV photons. After activation, radiation levels were measured and analyzed with a calibrated survey meter and with the device moved well away from the linac. Additionally, gamma camera images and energy spectra were obtained.

Results: As expected, activation increases with increasing exposure and photonenergy. Irradiating the device at 100 cm SSD for 1000 and 3000 MUs yielded survey meter readings at the surface of the device of 30 and 60 mR/hr respectively from 18 MV photons, and 2 and 5 mR/hr from 15 MV photons. 6 MV photons produced no detectable activation. The observed half‐life was approximately 10 minutes. Gamma cameraimages showed the activated material to be approximately uniformly distributed. The energy spectra showed a strong peak centered on 511 keV, consistent with annihilation photons from positron emission. These data are consistent with the Cu63 (y, n) Cu62 reaction. The decreased activation from 15 MV photons is also consistent with Cu63 activation which has a strongly peaked cross section centered around 16 to 18 MeV. Also observed were strong energy peaks centered near 76 keV and 177 keV, which require further analysis.

Conclusions: Although normal operation of this device does not usually require such high doses, the Medical Physicist should be aware of the possibly higher than expected activation of any such QA device and the resultant exposure. Without proper precautions this could have ALARA consequences. For example, Cu63 accounts for nearly 70% of naturally occurring copper and is used extensively in virtually all similar measuring devices. Though the Cu63 (y, n) Cu62 dominates, the other peaks will be discussed as well.


Modeling Dose Modulation in VMAT

M. R. Zaini, Ph.D.,1,2 T. A. Blackwell, M.S.1,2
1 Northwest Medical Physics Center, Lynnwood, WA; 2 Skagit Valley Regional Health Center, Mount Vernon, WA

International Journal of Radiotion Oncology•Biology•Physics, 75 (3S), S731 (2009).

Volumetric Modulated Arc Therapy (VMAT) is a new treatment modality in radiation therapy marketed by Elekta. VMAT uses both temporal and spatial modulation in delivering large arcs of radiation to patients. The aim is to quantify dose modulation levels based on the VMAT treatment parameters.