The overall goal of this project is to deliver an automated PET image analysis tool to be used in a clinical setting that quickly evaluates the PET images and reports PET standard uptake utilization rates from the corresponding MRI segmented anatomical locations to determine patients? risk of developing Alzheimer?s disease (AD). While multiple research studies have highlighted the utility of using PET imaging in diagnosis of AD and stratification of patients for the possible conversion from mild cognitive impairment (MCI) to AD, the research methods used are far from being available as a versatile tool that can be utilized in a clinical setting. In addition to the common ligand 18F-FDG, multiple Alzheimer?s relevant ligands have been recently introduced (11C-PiB, 18F-Florbetapir, 18F-Florbetaben, 18F-flutemetamol, as well as THK5317, THK5351, AV-1451, and PBB3). An automated PET image analysis tool will be timely and will make the best use of the newly introduced PET ligands. The advance of Alzheimer?s relevant ligands will likely improve diagnosis of AD and monitoring of disease progression. In phase I of our project, we will develop methods that assist in diagnosis of AD and calculate the odds of MCI to AD conversion using automated processing of PET data (FDG- PET) and MRI atrophy measures. In phase 2 of our proposal, we will extend our methods to include the new PET ligands. The amyloid ? and tau specific ligands present with different brain distributions, and will enhance our understanding of the dynamic details of the Alzheimer?s footprint in the brain. The degree of cortical binding of amyloid agents in patients with AD is variable, and the quantitative evaluation will require calculating normative data for the new ligands. As an additional confounder, the amyloid binding can exist in non-AD patients such as patients with Lewy body dementia or cerebral amyloid angiopathy. Similar to amyloid ? deposition in non-AD entities, various tauopathies in addition to Alzheimer?s disease will show tau ligand depositions in the patients? brain. We will determine normative values for each segmented brain region for multiple PET ligands currently available and calculate MRI atrophy measures. Using all regional PET measurements and MRI atrophy measures we will develop an automatic classification algorithm that will separate AD patients from non-AD controls, and also compute MCI to AD conversion risk for each MCI subject.
The overall goal of this project is to deliver an automated PET image analysis tool which can be used in a clinical setting and quickly evaluates the PET images and report values from the corresponding MRI segmented anatomical locations to determine the Alzheimer?s Disease (AD) risk of patients. We will determine normative values for each segmented brain region for multiple PET ligands currently available, and also calculate MRI atrophy measures. Using all regional PET measurements and MRI atrophy measures, we will develop an automatic classification algorithm that will separate AD patients from controls and also compute the risk of conversion from mild cognitive impairment (MCI) to AD for each patient.
