Angular dependency correction of 2D planar detector I’mRT MatriXX an offline dosimetry system used for IMRT pre-treatment verification

Abstract

I'mRT MatriXX is one of the common offline dosimetry tools used for the pre-treatment verification quality assurance (QA) of intensity modulated radiation therapy (IMRT) treatment plans in radiotherapy. It has 1020 pixel ionisation chambers (PIC) arranged in 32×32 grid in a plain. In this study the angular response of the I'mRT MatriXX was measured using treatment planning system (TPS) calculated dose and I'mRT MatriXX measured dose. The angular response correction factor (CF) of the I'mRT MatriXX was found by measuring the ratio of MatriXX measured dose to the treatment planning system (TPS) calculated dose in 5°step size of MatriXX angle from 0°to 355°u nder the similar set up condition for MatriXX with MULTICube Phantom and TPS plan. This CF can be applied directly in OmniPro software and used if the user have gantry angle sensor. The aim of this work is to apply this CF without gantry angle sensor to each of the 1020 PIC with the help of a simple program on MATLAB and verify with the TPS treatment plan field by field. For the 10 Ca cervix patients which were planned with standard 7 fields IMRT, there is improvement in the pass percentage of gamma index 3% dose difference and 3 mm distance to dose agreement. At 0°IMRT field, there is insignificant improvement but at other fields like 50°, 100°, 150°, 200°, 250°and 300°there are significant improvement in the gamma index pass percentage.

A novel angular dependency model for MatriXX response and its application to true composite dose verification for IMRT plans

“…The asymmetry about 90 • of the Δd(θ) agrees well with the asymmetric about 90 • of the CF profiles. The reproduce of the CF profile in Figure 2b using the DEMP model showed analogous shape to those of previous works, 7,8,13 verifying the efficacy of DEMP model.The fact that neither the depth offset curve of the EMP nor the CF curve is much influenced by the simulated set up error of 1 mm in Figure 2a,b demonstrated the robustness of the DEMP model to the set up uncertainties. Additionally, the derivatives of CF curve in Figure 2c revealed that the sensitivity of CF to the gantry angle is largest in the gantry angle range of (100 • , 120 • ).…”

“…Previous methods accounting for the angular dependency of MatriXX response mostly were based on some sort of CFs. 7 , 8 , 13 , 14 It should be noted that the DEMP model in this study is to calculate the angularly corrected composite planar dose, which is to be compared with the MatriXX response, using the TPS‐calculated dose distribution from each beam; while the conventional CF methods is to correct the MatriXX response from each beam and then synthesize them to match the TPS‐calculated dose on the measurement plane. The distinction between the two methods could be summarized as the conventional CF method corrects the MatriXX response, while the DEMP‐based approach corrects the TPS‐calculated planar dose.…”

“…Firstly, due to their phenomenological nature, CF methods normally require a large number of sampling points, for the rotational dimension of gantry angle and the off‐axis dimensions (2D) on the sensitive area of MatriXX, to make the correction sufficiently accurate. In this regard, the correction methods had evolved from using only CF for the center PXC for each angle, 7 to using CFs of a line of PXCs, 13 and to using CFs for all the PXCs on the whole 2D sensitive area 8 , 14 of MatriXX. The increase of the correction parameters demanded more time for MatriXX calibration and raised the possibility of overfitting.…”

“…Several MatriXX detectors have been tested, and rather large dose bias (up to 8%–11%) was observed for the non‐zero‐degree beams. 7 , 8 Hence, the true composite dose verification 9 was normally performed using dosimetric detectors without angular dependence of response, such as Gafchromic EBT3 films. 10 , 11 Nevertheless, the tedious calibration of films limits their uses for routine QA.…”

“…Shimohigashi et al 13 considered the off‐axis dependence of CF in one dimension, and calculated CFs using MatriXX measured dose and the dose calculated by the Monte Carlo (MC) algorithm. Boggula et al 14 and Robert et al 8 calculated CFs using MatriXX measured dose and treatment planning system (TPS)‐calculated dose to investigate the angular dependence of all the pixel chamber detectors. While the former used the CFs for VMAT QA, the later applied them for the composite dose verification of IMRT.…”