Optimisation of modelling parameters for small MLC defined fields

Abstract

Abstract: Advances treatments such as IMRT and VMAT which mean small leaf gap fields can be found in all sorts of plans. A "small field" is generally defined as a field with smaller lateral dimensions than the lateral range of the dose-contributing electrons. The aim of this thesis is find the optimising modelling parameter for sample statistic small fields using different type of detectors .then examine the modelling parameters in Eclipse treatment planning system and match their output with measurement result .finally , using our modelling parameters with complex VMAT plans to confirm the improvement that will obtain from new model using different type of QA test . Materials and Methods This thesis demand on four-step, firstly Measurement of the dosimetric leaf gap by using The photon energies, 6x MV, 10x MV and 10x FFF MV were made using the multileaf collimators (MLCs) Varian LINAC. the chamber Exradin A19, Semiflex 3D, a 3D pinpoint, Diode E, and Micro diamond were used to find any effect on the result of our experiments. secondly, measure profiles for MLC fields by field sizes studied were 4 mm,12 mm,18 mm. Measurements were performed using beam scan phantom and normalise them to the output of a 10x10cm field at Dmax (i.e. reference conditions). thirdly, optimising the configuration beam by using result were obtain from the first step and compare with current clinical calculation parameters. finally, recalculate clinical plans and compare with measurement. Result : The feasibility study showed that changing the MLC dosimetric leaf gap and spot size play important role to optimise MLC modelling. The combination between 1.5 mm dosimetric leaf gap and 0.6 spot size give the optimal fit for modelling parameter.in addition, the result of t-pair test sampleOn recalculating previous clinical plans show that the T test value was 9.794. By standard criteria this distinction is regarded to be highly statistically significant due to The two-tail P value is less than 0.0001 . therefore, new model optimal than the old model