Alzheimer’s disease (AD) is the leading cause of dementia and is alone responsible for 60-70% of all cases of dementia. Though sharing clinical symptoms with other types of dementia, the hallmarks of AD are the abundance of extracellular depositions of β-amyloid (Aβ) plaques, intracellular neurofibrillary tangles of hyper phosphorylated tau proteins and synaptic depletion. The onset of the physiological hallmarks may precede clinical symptoms with a decade or more, and once clinical symptoms occur it may be difficult to separate AD from other types of dementia based on clinical symptoms alone. Since the introduction of radiolabeled Aβ tracer substances for positron emission tomography (PET) imaging it is possible to image the Aβ depositions in-vivo, strengthening the confidence in the diagnosis. Because the accumulation of Aβ may occur years before the first clinical symptoms are shown and even reach a plateau, Aβ PET imaging may not be feasible for disease progress monitoring. However, a negative scan may be used to rule out AD as the underlying cause to the clinical symptoms. It may also be used as a predictor to evaluate the risk of developing AD in patients with mild cognitive impairment (MCI) as well as monitoring potential effects of anti-amyloid drugs.Though currently validated for dichotomous visual assessment only, there is evidence to suggest that quantification of Aβ PET images may reduce inter-reader variability and aid in the monitoring of treatment effects from anti-amyloid drugs.The aim of this thesis was to refine existing methods and develop new ones for processing, quantification and visualization of Aβ PET images to aid in the diagnosis and monitoring of potential treatment of AD in clinical routine. Specifically, the focus for this thesis has been to find a way to fully automatically quantify and visualize a patient’s Aβ PET image in such way that it is presented in a uniform way and show how it relates to what is considered normal. To achieve the aim of the thesis registration algorithms, providing the means to register a patient’s Aβ PET image to a common stereotactic space avoiding the bias of different uptake patterns for Aβ- and Aβ+ images, a suitable region atlas and a 3-dimensional stereotactic surface projections (3D SSP) method, capable of projecting cortical activity onto the surface of a 3D model of the brain without sampling white matter, were developed and evaluated.The material for development and testing comprised 724 individual amyloid PET brain images from six distinct cohorts, ranging from healthy volunteers to definite AD. The new methods could be implemented in a fully automated workflow and were found to be highly accurate, when tested by comparisons to Standards of Truth, such as defining regional uptake from PET images co-registered to magnetic resonance images, post-mortem histopathology and the visual consensus diagnosis of imaging experts.
| Identifer | oai:union.ndltd.org:UPSALLA1/oai:DiVA.org:uu-317688 |
| Date | January 2017 |
| Creators | Lilja, Johan |
| Publisher | Uppsala universitet, Radiologi, Uppsala |
| Source Sets | DiVA Archive at Upsalla University |
| Language | English |
| Detected Language | English |
| Type | Doctoral thesis, comprehensive summary, info:eu-repo/semantics/doctoralThesis, text |
| Format | application/pdf |
| Rights | info:eu-repo/semantics/openAccess |
| Relation | Digital Comprehensive Summaries of Uppsala Dissertations from the Faculty of Medicine, 1651-6206 ; 1322 |
Page generated in 0.0025 seconds