In this thesis we consider the task of dynamic imaging using a gcPET system. Our technique is based on a mathematical method (developed for SPECT), which processes all dynamic projection data simultaneously instead of reconstructing a series of static images individually. The algorithm was modified to account for the extra data that is obtained with gcPET (compared with SPECT). The method was tested using simulated projection data for both a SPECT and a gcPET geometry. These studies showed the ability of the code to reconstruct simulated data with a varying range of half-lives. For SEPCT data the characteristic parameters of half-life (T<sub>1/2</sub>) and initial activity (A<sub>0</sub>) were reconstructed with a percentage error of 35.1%, and 40.8% (at 50 iterations) for a 2 minutes half-life, respectively. The reconstruction of gcPET data showed improvement in half-life and activity compared to SPECT data by 27% and 31%, respectively (at 50 iterations). The method was also extended to enable reconstruction of images in which some regions increased in activity while other regions decreased. Information of the spatial location of these images was provided in the form of a mask. The method was applied to experimental data. These data were acquired using a dPET system and re-binned to the gcPET geometry. The results, obtained from dynamic phantoms, showed that the characteristic behaviour could be recovered and that the code produced satisfactory dynamic images. The method was also applied to data from a patient with a tumour. Again, the reconstructed image showed good results compared to the dPET reconstruction. Time activity curves showed a significant difference between the uptake of tumour and myocardium. Finally, we presented a method to deal with the situation where the activity in certain pixels decreases and then increases during the acquisition.
| Identifer | oai:union.ndltd.org:bl.uk/oai:ethos.bl.uk:421358 |
| Date | January 2004 |
| Creators | Omar, Ahmed M. |
| Publisher | University of Aberdeen |
| Source Sets | Ethos UK |
| Detected Language | English |
| Type | Electronic Thesis or Dissertation |
Page generated in 0.0019 seconds