A Predictive Model for Identifying Possible MCI to AD Conversions in the ADNI Database

Alzheimer´s disease (AD) is one of the most common forms of dementia and has become a serious issue among the elderly in the aging society. Since AD is incurable and degenerative, early diagnosis is essential, which can give patients and their family more opportunities to arrange their lives. In the meantime, histopathologic studies have found that MCI (mild cognitive impairment) subjects usually have intermediate levels of AD pathology. In this paper, a predictive model is developed for identifying possible conversions from MCI to AD based on the ADNI (Alzheimer´s disease neuroimaging initiative) database. It is shown that, with the help of a range of advanced data mining techniques, the developed model can achieve promising performance with AUC around 0.88.