A Comparative Analysis of Machine Learning Methods for Predicting Progression
from Mild Cognitive Impairment to Alzheimer???s Disease. 
   Lead Investigator:    Emeka Okonji
   Institution      :    Harvard T.H. Chan School of Public Health
   E-Mail           :    eokonji@hsph.harvard.edu
   Proposal ID      :    1230



   Proposal Description:
   Alzheimer?s disease (AD) is the most common cause of dementia and is a major public health issue across the globe. Given that current treatments can only decelerate the progression of AD and have the greatest effect at the earliest disease stage, early identification of subjects at high risk of conversion to full AD is crucial to lessening morbidity. The earliest clinically detectable stage of this disease is mild cognitive impairment (MCI) however, only 10 to 15 of MCI patients from referral sources such as memory clinics and AD centres will develop AD annually. Therefore, the aim of this study is to compare several machine learning classification techniques in order to build an optimal model for identifying MCI subjects at high risk of conversion to full AD within 5 years using a restricted set of non-invasive and easily collected clinical, neuropsychological and sociodemographic predictors. This model will facilitate early intervention in high-risk individuals with MCI and aid the enrichment of clinical trials by identifying candidates likely to develop AD in the time frame of the trial.