Alzheimer's disease detection via machine learning.

Abstract: 

In this talk we discuss the potential of two machine learning techniques (support vector machines and neural networks) in aiding medical diagnosis of Alzheimer's disease. In particular, we see whether these two techniques are capable of distinguishing between data from patients with Alzheimer's disease and data from healthy controls, obtained from MRI scans of two longitudinal studies.

A brief description of the neuropathology of Alzheimer's disease will be followed by an introduction to both techniques. Our method for extracting cortical thickness will also be outlined. The main results of the thesis will then be presented and their significance evaluated.

James Hortle is an Honours student under the supervision Dr Quoc Thong Le Gia.