abstract
- © 2018 IEEE. Predicting who will convert from mild cognitive impairment (MCI) to Alzheimer's disease would have large clinical implications, and stratifying this process could enhance its accuracy. This study aimed at identifying features from the Alzheimer's Disease Neuroimaging Initiative database associated with such a conversion whose temporal evolution significantly differs between men and women. We gathered information from 59 women and 109 men who were diagnosed with MCI at baseline and eventually converted to AD. Information included features from laboratory assays, MRI and PET analyses, and neuropsychological tests. Statistical longitudinal models identified features with significant differences in longitudinal behavior between genders. We found 20 features with such differences at either the time at which subjects were first diagnosed with AD or related to the speed at which they evolved. Our results show that the way in which male and female subjects evolve throughout the MCI to AD conversion process is different, hence, predicting who will go through such a conversion could be better predicting by taking into account these differences.