Complex Diffusion Brain Imaging-Based Machine Learning Model to Predict MCI to AD Conversion Lead Investigator: Bang-Bon Koo Institution : Boston University School of Medicine E-Mail : koobangbon@gmail.com Proposal ID : 1483 Proposal Description: 1) The goal is to determine whether complex diffusion brain imaging is sufficient in accurately predicting progression to AD for MCI patients in a machine learning model. This will be in conjunction with cognitive scores, brain morphometry, demographic factors, etc. However, we expect that complex diffusion brain profiles will provide significant insight into progressive neurodegeneration from MCI to AD. Development of a highly accurate single-subject computational model will greatly contribute to the identification of effective and efficient imaging biomarkers for early neuropathological developments. 2) To determine whether complex diffusion brain imaging profiles of various MCI time points and AD can be used to develop an accurate machine learning model to predict MCI to Ad conversion.