Abstract
In this thesis the main aspects of three photon positron annihilation processes and their potential use in medical imaging have been investigated as a positron emission tomography technique. The main objectives are focused on: three-photon positron annihilation measurement and imaging, analytical modelling and Monte Carlo simulation and the evaluation of the detection system requirements. A novel method as proof-of-principle of the three photon positron annihilation imaging concept based on a triple coincidence imaging technique using high energy resolution semiconductor detectors has been introduced. It has been shown that a simple system of three high-energy resolution detectors is able to produce images of three photon positron annihilations events. The full energy photopeak detected of the true 3y events can be easily identified in the spectrum. Although the sensitivity is small due to a very small solid angle (~0.05 str) subtended by the detectors and rather poor detection efficiency, it is a first step towards a scanner capable of a new imaging modality. This method has been investigated using Monte Carlo simulation results and experimental data acquired. Further a new three-photon yield measurement method based on three-photon positron annihilation imaging technique with correction for scattered and random events is proposed. The feasibility of this approach has been verified using experiments and compared to existing methods. Results show that this method is more accurate with better scatter correction due to electronic collimation than others but it has also some limitations. In order to obtain quantitative information from the detection system it is necessary to establish mathematical or analytical models, which describe the system. This was achieved for the triple coincidence condition. The count rate of single and triple detected events was investigated. Results have shown differences due to scatter and random events estimation. The effect of semiconductor detectors properties on three-photon image quality and scanner design was also investigated. It was in addition shown that computer simulations can be effectively used to predict the image quality and background noise for a particular scanner design. Important characteristics which affect scanner performance were evaluated. The effect of detector and scanner size on spatial resolution of three photon images was discussed. It was found that by reducing scanner size spatial resolution was improved for three-photon positron annihilation imaging as for conventional two photon-positron annihilation. Variation of scanner size (scanner diameter) affects the point spread function of the three photon positron annihilation image profile and introduces a combination of errors due to photon energy and detection position. To introduce the three-photon technique in a dedicated PET system, high- energy resolution detectors are needed to improve the quality of the image and reduce the noise due to scattered events arising from Compton scattering which do not correspond to 3y events. Semiconductor detectors, particularly CZT which have good energy resolution, significantly better stopping power and can be used at room temperature are proposed as the detectors of choice for the new detection system. Therefore, images of two photon-positron annihilation can be mapped with those of 3- photon events and new valuable information can be extracted. This information will be valuable to treatments involving external beam radiotherapy and may also be of use in brachytherapy.