Verification of stereotactic radiotherapy

Investigations have been made into the use of a computer based simulation technique (Monte Carlo (MC)) to ionising radiation transport in order to verify the doses delivered during linear accelerator based stereotactic radiotherapy and radiosurgery. Due to the complex nature of the micro multi-leaf collimators (μMLC) used in this these treatments, a bespoke model of the μMLC was developed and combined with standard component modules to represent the remainder of the linear accelerator. Following validation of the above models, investigations were made into the dosimetry of small fields, defined by the μMLC and measured with a variety of detectors. Comparisons of relative output, profiles and depth doses were made against MC simulations, and a series of correction factors determined, to account for detector geometry and the non water equivalence of materials used in semiconductor detectors. An assessment was then made to determine the smallest fields that can be measured with each detector with confidence. Systems were then developed to independently simulate stereotactic treatments and compare doses simulated with those calculated by the treatment planning system (TPS); excellent agreement between TPS calculations and MC simulations was observed. The application of MC methods to determine the most appropriate treatment tactics and calculation algorithms for stereotactic body radiotherapy in the lung was then investigated with recommendations made on the most appropriate calculation algorithms and beam arrangements for the technique. The doses calculated using the type-b or collapsed cone algorithm agreed most closely with the MC simulation. There was little difference observed between plans using more than four beams in the treatment delivery. Treatment techniques using only three beams or less achieved poorer coverage of the tumour with dose, producing lower doses at the periphery of the tumour near the interface with the surrounding lung tissue, compared to using a greater number of beams. Finally, methods of transit dosimetry using Electronic Portal Imaging Devices were investigated for use in cranial stereotactic radiotherapy. Three methods were investigated based on a full MC simulation of the radiation transport through the patient and on to the imager, prediction of the dose based on a TPS calculation and an approximation of the radiological path length of the central axis of the beams to derive an expected dose at the imager plane. The MC method produced the best agreement at the expense of a longer time to acquire the comparison doses compared to the TPS calculation method. The equivalent path length method showed good agreement (within 3.5%) between delivered and predicted doses but at a single point.EThOS - Electronic Theses Online ServiceGBUnited Kingdo