Abstract
Purpose: Assess the applicability of the MIM Maestro (MIM Software, Inc.) dose accumulation algorithm for adaptive planning using data from MOSFET dosimeters implanted in the prostate. Methods: Under a previously approved IRB protocol, prostate cancer patients were implanted with MOSFET dosimeters (Dose Verification System (DVS) by Sicel Technologies) and treated with daily CBCT. Daily readings were compared to dose predictions generated by an automated dose accumulation workflow (MIM Software) for five patients (179 total fractions): (1) contours from the planning CT (PCT) were rigidly transferred to the day 1 CBCT (CBCT1), (2) contours on CBCT1 were manually reviewed by the treating physician, (3) contours for each subsequent fraction (CBCTf) were deformably transferred from CBCT1, (4) planning dose was rigidly applied to CBCTf, (5) dose was accumulated. Results: The MIM workflow was efficient and applicable in a clinical setting. When comparing the MIM readings to the daily DVS readings, 53% fell within 3% of the measured dose, 80% fell within 5% and only 0.8% were greater than 10% from the measured dose. The largest deviations from the expected dose found in the MIM readings resulted from difficulties in registering of the CBCTs due to artifacts in the images. Conclusion: The MIM workflow was able to accurately predict the dose to the DVS despite inferior image quality with CBCT. Rigid dose registration was used to avoid unrealistic dose deformations, but this approach does not take into account changes in external contour. To further assess the efficacy of the workflow for adaptive planning and dose accumulation, the original treatment plan will be applied to the daily CBCT with the deformed contours. Additional dosimetric endpoints will be analyzed on deformable structures such as the PTV and OARs. Sara Pirozzi is an employee at MIM Software, Inc.
