Optimisation of radiation dose to patients in computed tomography

Methodologies for optimisng CT scanning protocols have been developed in this project. In the first instance, a novel Monte Carlo CT model has been constructed to calculate effective dose. Comparisons With experimental data in CT dose phantoms indicate that the accuracy of the model is better than 10%. The model has been used to demonstrate that, if used with care, commercial dosimetry packages can calculate effective dose to standard sized subjects undergoing multi-slice CT scans of the torso without incurring errors larger than 20 %. For electron-beam CT, the model has been used to calculate effective doses for screening and diagnostic scans; it has also demonstrated that commercial packages are still valid except for scans of the thorax, where a correction factor of 0.75 should be applied. In the second instance, an optimisation methodology, based on phantom experiments to evaluate image quality and the novel Monte Carlo CT model to calculate patient doses, has been devised. It has been applied to scans of the paediatric abdomen and pelvis. If the noise levels observed in adult scanning are maintained across all patient sizes, there is a potential for dose reduction in paediatric scanning by a factor of 3-6 compared to local clinical practice. The tube voltage can be reduced to increase the visualisation of iodine-enhanced tissue and visceral fat. The optimisation methodology has also been applied to CT perfusion of the head and neck where a risk of skin radiation injury exists. A scan protocol has been recommended and implemented in a patient trial. It increases by 10 % the contrast-to-noise ratio of iodine-enhanced tissues achievable with the manufacturer's recommended protocol and delivers half the surface dose to the patient with no detrimental effect on normal perfusion parameters.EThOS - Electronic Theses Online ServiceGBUnited Kingdo