Quantifying the limitations of small animal positron emission tomography

The application of position sensitive semiconductor detectors in medical imaging is a field of global research interest. The Monte-Carlo simulation toolkit GEANT4 [http://geant4.web.cern.ch/geant4/] was employed to improve the understanding of detailed γ -ray interactions within the small animal Positron Emission Tomography (PET), high-purity germanium (HPGe) imaging system, SmartPET [A.J. Boston, et al., Oral contribution, ANL, Chicago, USA, 2006]. This system has shown promising results in the field of PET [R.J. Cooper, et al., Nucl. Instr. and Meth. A (2009), accepted for publication] and Compton camera imaging [J.E. Gillam, et al., Nucl. Instr. and Meth. A 579 (2007) 76]. Images for a selection of single and multiple point, line and phantom sources were successfully reconstructed using both a filtered-back-projection (FBP) [A.R. Mather, Ph.D. Thesis, University of Liverpool, 2007] and an iterative reconstruction algorithm [A.R. Mather, Ph.D. Thesis, University of Liverpool, 2007]. Simulated data were exploited as an alternative route to a reconstructed image allowing full quantification of the image distortions introduced in each phase of the data processing. Quantifying the contribution of uncertainty in all system components from detector to reconstruction algorithm allows the areas in need of most attention on the SmartPET project and semiconductor PET to be addressed.