Building disease progression models for Alzheimer's disease Lead Investigator: Jiayu Zhou Institution : Michigan State University E-Mail : jiayuz@msu.edu Proposal ID : 570 Proposal Description: The main direction of my research is to build disease progression models for Alzheimer's disease by developing novel machine learning algorithms. For the last five years, I have designed machine learning models to identify key biomarkers and image markers that signal the Alzheimer's progression, from which I have been able to build disease progression models for early stage Alzheimer's disease and validated on the ADNI data. More recently I would like to build models with higher performance using deep learning architecture, with the hope that the flexibility provided by deep learning can leverage finer interactions among the biomarkers and image signals. The NACC data is ideal for this research goal because it has a cohort that is much larger than the ADNI and the huge amount of information in NACC data can be used to train sophisticated deep models. I would also like to validate the deep models built from NACC on the ADNI data.