Towards fully automated MR-independent binding potential estimation for PET

The estimation of parametric maps of binding potential (BP) for brain imaging with positron emission tomography (PET) is an essential stage of analysis when radioligands are used for neuroreceptor-related studies. A popular approach is the simplified reference tissue model (SRTM), as this avoids the need for arterial blood sampling. Furthermore, there has been a recent advance in PET image reconstruction towards direct (4D) reconstruction methods, which ideally should estimate parameters such as BP without user intervention. However, depending on the radioligand, there are a number of potential issues to be addressed to achieve automated 4D reconstruction: 1) the accurate definition of an appropriate reference region (usually reliant on the availability of an MR), 2) variability in BP results due to variations in the definition of the reference region, and 3) removal of the need for user intervention in the analysis. We are addressing these issues by developing a methodology to automatically identify candidate reference regions for the analysis (based purely on the available PET data - without the need for an MR), and to generate BP maps in a fully automated fashion, with uncertainty estimates reflective of the variability in reference region selection. The initial goal is to provide a provisional kinetic analysis automatically for every PET scan (just as at present both the image reconstruction and motion correction are performed in an automated and standard way) prior to ultimately proceeding towards fully automated direct (4D) BP parametric map estimation. In this initial work we demonstrate the impact of reference region size and location on the BP estimates, as a means of establishing the requirements and tolerance of a fully automated analysis procedure.     «

The estimation of parametric maps of binding potential (BP) for brain imaging with positron emission tomography (PET) is an essential stage of analysis when radioligands are used for neuroreceptor-related studies. A popular approach is the simplified reference tissue model (SRTM), as this avoids the need for arterial blood sampling. Furthermore, there has been a recent advance in PET image reconstruction towards direct (4D) reconstruction methods, which ideally should estimate parameters such as...     »