Optimization of novel developments in Positron Emission Tomography (PET) imaging

Positron Emission Tomography (PET) is a widely used imaging modality for diagnosing patients with cancer. Recently, there have been three novel developments in PET imaging aiming to increase PET image quality and quantification. This thesis focuses on the optimization of PET image quality on these three developments.;The first development is the fully 3D PET data acquisition and reconstruction. 3D Acquisitions are not constrained in collecting events in single 2D planes and can span across different planes. 3D acquisition provides better detection since it can accept more events. Also it can result in lower radiation dose to the patient and shorter imaging times. With the application of 3D acquisition, a fully 3D iterative reconstruction algorithm was also developed. The aim of the first project in this thesis is to evaluate the PET image and raw data quality when this fully 3D iterative reconstruction algorithm is applied.;The second development in PET imaging is the time-of-flight (TOF) PET data acquisition and reconstruction. TOF imaging has the ability to measure the difference between the detection times, thus localize the event location more accurately to increase the image quality. The second project in this thesis focuses on optimizing the TOF reconstruction parameters on a newly developed TOF PET scanner. Then the improvement of TOF information on image quality is assessed using the derived optimal parameters. Finally the effect of scan duration is evaluated to determine whether similar image quality could be obtained between TOF and non-TOF while using less scan time for TOF.;The third development is the interest in building PET / magnetic resonance (MR) multi-modality scanner. MR imaging has the ability to show high soft tissue contrast and can assess physiological processes, which cannot be achieved on PET images. One problem in developing PET/MR system is that it is not possible with current MR acquisition schemes to translate the MR image into an attenuation map to correct for PET attenuations. The third project in this thesis proposed and assessed an approach for the attenuation correction of PET data in potential PET/MR systems to improve PET image quality and quantification.