Evaluation of a planar diode matrix for SRS patient-specific QA in comparison with GAFchromic films.

PURPOSE: We validate the routine use of a two-dimensional (2D) diode matrix for patient specific pre-treatment verification for Cyberknife (CK) stereotactic radiosurgery and to compare it with film dosimetry. MATERIALS AND METHOD: A total of 46 patients were selected according to the most frequent diseases treated at our institution with the CK system, that is, brain metastases, meningiomas, spine metastases, and prostate tumors. All cases were evaluated with GAFChromic EBT-3 films and SRS MapCHECK for Fixed cone, IRIS, and MLC collimators of the CK. RESULTS: The highest mean passing rate was observed for the SRS MapCHECK system compared to films. In order to assess if the two techniques provide statistically different results, a Wilcoxon Signed-Rank non-parametric test was performed (p < 0.05) and we found gamma values significantly lower for EBT-3 films with respect to the SRS MapCHECK. We noticed a moderately significant association between the two techniques using Spearman's rank correlation coefficient (rs > 0.4). We also performed the Bland-Altman statistical method: less than 5% of the differences resulted outside the range (mean ± 1.96 × SD), so the two methods can be considered interchangeable within the combined inaccuracy. CONCLUSIONS: The use of SRS MapCHECK for CK patient specific quality assurance (QA) is feasible for a variety of clinical districts and could be reliably used as a replacement for radiochromic films.

Zero field PDD and TMR data for unflattened beams in conventional linacs: A tool for independent dose calculations.

PURPOSE: To investigate the applicability of the formalism described in BJR supplement n.25 for Flattening Filter Free (FFF) beams in determining the zero-field tissue maximum ratio (TMR) for an independent calculation method of Percentage Depth Doses (PDDs) and relative dose factors (RDFs) at different experimental setups. METHODS: Experimental PDDs for field size from 40×40cm2 to 2×2cm2 with Source Surface Distance (SSD) 100cm were acquired. The normalized peak scatter factor for each square field was obtained by fitting experimental RDFs in water and collimator factors (CFs) in air. Maximum log-likelihood methods were used to extract fit parameters in competing models and the Bayesian Information Criterion was used to select the best one. In different experimental setups additional RDFs and TPR1020s for field sizes other than reference field were measured and Monte Carlo simulations of PDDs at SSD 80cm were carried out to validate the results. PDD agreements were evaluated by gamma analysis. RESULTS: The BJR formalism allowed to predict the PDDs obtained with MC within 2%/2mm at SSD 80cm from 100% down to 50% of the maximum dose. The agreement between experimental TPR1020s and RDFs values at SSD=90cm and BJR calculations were within 1% for field sizes greater than 5×5cm2 while it was within 3% for fields down to 2×2cm2. CONCLUSIONS: BJR formalism can be used for FFF beams to predict PDD and RDF at different SSDs and can be used for independent MU calculations.