Neuroimaging has provided relevant information on the diagnostic status and disease progression of AD and MCI. In quantifying patterns of structural change during early stages of AD, several neuroimaging initiatives have discovered biological markers associated with AD and MCI based on brain images and machine learning. To our knowledge, the AD classification accuracies from existing literature ranged between 86-93% using the volumetric morphology of a subset of the T1-weighted images from the Alzheimer’s Disease Neuroimaging Initiative (ADNI). Up to date, there is no study that demonstrates the discriminative accuracy of brain imaging markers in predicting incident dementia a few years after imaging assessment. It also remains unclear whether imaging markers in early life is associated with cognitive performance later.
This project aims to employ brain imaging prior to dementia incidence and genetic data for the prediction of Alzheimer’s disease. We will develop novel machine learning approaches for this purpose. We will demonstrate 1) the prediction accuracy of incident dementia 4 years after brain imaging; 2) the association of brain imaging markers with cognitive performance and genetic risk for AD; 3) the relation of brain imaging markers with family dementia risk using both UK Biobank and ADNI data.
