Digital signal processing techniques for semiconductor Compton cameras

The work presented in this thesis has focused on the development of a low dose Compton camera for nuclear medicine. A Compton camera composed of two high-purity planar germanium orthogonal-strip detectors has been constructed. Fast digital data acquisition has been utilised for the application of pulse shape analysis techniques. A simple back projection imaging code has been developed and validated with a Geant4 radiation transport simulation of the Compton camera configuration. L A 137CS isotropic source and a 22Na anisotropic source have been experimentally reconstructed. Parametric pulse shape analysis was applied to both data sets and has been shown to increase the detector spatial resolution from a raw granularity of 5x5x20mm to a spatial resolution that can be represented by a Gaussian distribution with a standard deviation of 1.5mm < u < 2mm in all dimensions; this result was in-part derived from Geant4 simulations. Qualitatively poor images have been shown to result - based wholly on simulation - from Gaussian spatial-resolution distributions that have a standard deviation of greater than 4mm. A partial experimental basis-data-set has been developed and proved capable of providing 1.9mm FWHM average spatial resolution through the depth axis of a single detector crystal. A novel technique to identify gamma ray scattering within single detector c1osed-face-pixels - hitherto unrecognised - has also been introduced in this thesis. This technique, henceforth known as Digital Compton Suppression (DieS), is based on spectral analysis and has demonstrated the ability of identifying events in which the Compton scattering and photoelectric absorption sites are separated by 13mm in the direction ofthe electric field. Supplied by The British Library - 'The world's knowledge'EThOS - Electronic Theses Online ServiceGBUnited Kingdo