Evaluation and improvement of angular response for a commercial 2D detector array for patient-specific QA.

Abstract

MatriXX ionization chamber array has been widely used for the composite dose verification of IMRT/VMAT plans. However, in addition to its dose response dependence on gantry angle, there seems to be an offset between the beam axis and measured dose profile by MatriXX for oblique beam incidence at various gantry angles, leading to unnecessary quality assurance (QA) fails. In this study, we investigated the offset at various setup conditions and how to eliminate or decrease it to improve the accuracy of MatriXX for IMRT/VMAT plan verification with original gantry angles. We measured profiles for a narrow beam with MatriXX located at various depths in increments of 0.5mm from the top to bottom of the sensitive volume of the array detectors and gantry angles from 0° to 360°. The optimal depth for QA measurement was determined at the depth where the measured profile had minimum offset. The measured beam profile offset varies with incident gantry angle, increasing from vertical direction to lateral direction, and could be over 3cm at vendor-recommended depth for near lateral direction beams. The offset also varies with depth, and the minimum offset (almost 0 for most oblique beams) was found to be at a depth of ∼2.5mm below the vendor suggested depth, which was chosen as the optimal depth for all QA measurements. Using the optimal depth we determined, QA results (3%/2mm Gamma analysis) were largely improved with an average of 99.4% gamma passing rate (no fails for 95% criteria) for 10 IMRT and VMAT plans with original gantry angles compared to 94.1% using the vendor recommended depth. The improved accuracy and passing rate for QA measurement performed at the optimal depth with original gantry angles would lead to reduction in unnecessary repeated QA or plan changes due to QA system errors.