The development of a clinical decision making framework for image guided radiotherapy

This thesis developed a decision-support Bayesian network (BN) and complementary visualisation tool for image-guided radiotherapy (IGRT). These tools form the framework for a future decision-support system that will rapidly analyse cone beam computed tomography (CBCT) images and provide advice to clinicians whether to proceed with treatment delivery or if a pre-treatment intervention is required. Clinical experts completed an IGRT practice survey, which along with published and clinical data were used to develop the BN. The visualisation tool graphically summarises daily tumour and organ-at-risk variations routinely observed on CBCT images, which can result in incorrect targeting of the radiation fields