Program Information
Dosimetric Accuracy of the CrystalBall™: New Reusable Radiochromic Polymer Gel Dosimeter for Patient QA in Proton Therapy
S Avery1*, J Kraus1 , L Lin1 , A Kassaee1 , M Maryanski2 (1) University of Pennsylvania, Philadelphia, PA (2) MGS Research, Inc., Madison, CT
Presentations
MO-FG-CAMPUS-T-12 (Monday, July 13, 2015) 5:30 PM - 6:00 PM Room: Exhibit Hall
Purpose: To evaluate the accuracy of monoexponential normalization in a new class of commercial, reusable, human-soft-tissue-equivalent, radiochromic polymer gel dosimeters for patient-specific QA in proton therapy.
Methods: Eight formulations of the dosimeter (sealed in glass spheres of 166 mm OD), were exposed to a 150 MeV proton beam (5 cm x 5 cm square field, range 15 cm, modulation10 cm), with max dose ranging from 2.5 Gy to 20 Gy, depending on formulation. Exposed dosimeters were promptly placed in the commercial OCTOPUS™ laser CT scanner which was programmed to scan the central slice every 5 minutes for 20 hours (15 seconds per slice scan). This procedure was repeated several times. Reconstructed data were analyzed using the log-lin scale to determine the time range over which a monoexponential relaxation model could be applied. Next, a simple test plan was devised and delivered to each dosimeter. The OCTOPUS™ was programmed to rescan the central slice at the end of each volume scan, for signal relaxation reference. Monoexponential normalization was applied to sinograms before FBP reconstruction. Dose calibration was based on a volume-lookup table built within the central spherical volume of 12 cm diameter. 3D gamma and sigma passing rates were measured at 3%/3mm criteria down to 50% isodose.
Results: Approximately monoexponential signal relaxation time ranges from 25 minutes to 3.5 hours, depending on formulation, followed by a slower-relaxation component. Noise in reconstructed OD/cm images is less than 0.5%. Dose calibration accuracy is better than 99%. Measured proton PDDs demonstrate absence of Bragg-peak quenching. Estimated number of useful cycles is at least 20, with a theoretical limit above 100. 3D gamma and sigma passing rates exceed 95%.
Conclusion: Monoexponential normalization was found to yield adequate dosimetric accuracy in the new class of commercial radiochromic polymer gel dosimeters for patient QA in proton therapy.
Contact Email: