Positron emission tomography (PET) is an imaging modality widely used in the study of neurodegeneration to image the accumulation of the misfolded proteins such as amyloid-β (Aβ) and tau. The aims of this thesis were to investigate Aβ accumulation in Alzheimer’s disease (AD) using mathematical modelling and perform kinetic modelling of the novel tau radiotracer [18F]-AV1451 in corticobasal degeneration (CBD) and traumatic brain injury (TBI). The main contributions of this thesis are:
Aβ - Modelling the spatiotemporal distribution of Aβ in AD. The spatiotemporal distribution of Aβ in AD was modelled using a network diffusion model and a four-parameter logistic growth model. Results suggest the spatiotemporal distribution of Aβ in AD results from properties of the local tissue environment rather than long-term spreading from seed regions.
Aβ - Amyloid Load - Using parametric images derived from the logistic growth model, a new out- come measure, amyloid load AβL was defined for a single [18F]-AV45 Aβ PET scan. It had greater sensitivity than the currently used mean cortical SUVr out- come measure and will improve the power to detect signal changes in future trials of novel anti-Aβ therapeutics.
Tau - Kinetic modelling of [18F]-AV1451
Dynamic [18F]-AV1451 PET data from a CBD study was effectively modelled with a reversible two-tissue compartmental model, the simplified reference tis- sue model (SRTM) and standardised uptake value ratio (SUVr). SRTM was used to quantify [18F]-AV1451 uptake in a TBI study consisting of 32 subjects. There were no differences between patients and controls at the ROI level and the voxel level analysis revealed heterogeneous radiotracer uptake amongst patients. It is impossible to tell whether increased PET signal is due to tau or off-target binding.