Abstract
The goal of Positron Emission Tomography (PET) is to provide accurate in-vivo information about biological processes. PET uses radio-labelled compounds to visualise cell-specific uptake corresponding to these processes. Animal models provide well-understood and valuable opportunities to learn about human biology, to develop diagnostic tools and to test potential therapies.The aim of this thesis is to investigate quantification of radiotracer uptake by a state-of-the-art commercial preclinical PET-CT scanner.
First, a performance evaluation of the scanner was carried out based on the recommendations of the National Electrical Manufacturers Association both of the prototype and a production version of the scanner. The scanner was found to have the highest resolution among the commercial scanners in conjunction with high sensitivity. Secondly, one of the major problems with quantification, the partial-volume effect, was studied. Partial-volume correction was applied using different methods: deconvolution post-reconstruction, modelling a point-spread function in the reconstruction and modelling the detector in the reconstruction. Simulation of small objects, and subsequently phantom measurements were carried out, and the best of the investigated methods to correct for partial-volume effect was suggested. Thirdly, on the production scanner, other corrections and their contribution to decreasing errors in quantification were investigated by measuring a mouse-sized and a rat-sized phantom with a wide range of activities and contrast ratios. Errors of quantification were estimated and linked to fundamental characteristics of the scanner. Fourthly, animal experiments were carried out to test the quantification in-vivo comparing the results obtained with the scanner and by ex vivo biodistribution measurements. Corrections previously tested in were applied to the images. Images and biodistribution data matched well, though with some limitations.
As a conclusion, investigation of image quantification of a preclinical PET scanner suggest that quantitative radionuclide imaging is possible.
Date of Award | 2013 |
---|---|
Original language | English |
Awarding Institution |
|
Supervisor | Paul Marsden (Supervisor) & Gregory Mullen (Supervisor) |