Author:
Sharma Simran,Kundal Kavita,Chandok Ishsirjan Kaur,Kumar Neeraj,Kumar Rahul
Abstract
AbstractAlzheimer’s disease (AD) is acknowledged as one of the most common types of dementia. Various brain regions were found to associated with AD pathology. Precuneus and fusiform gyrus are two notable regions whose role has been implicated in cognitive function. However, a thorough investigation was lacking to link these regions with AD pathology. In this study, we conducted a comprehensive radiomic based investigation using magnetic resonance imaging (MRI) scans to link precuneus and fusiform gyrus with AD pathology. We obtained T1 weighted MR scans of AD (n=133), MCI (n=311) and CN (n=195) subjects from ADNI database at three different time points (i.e., 0, 6 and 12 months). Then, we conducted statistical analysis to compare these features among AD, MCI and CN subjects. We found significant decline in gray matter volume (GMV) and cortical thickness of both precuneus and fusiform gyrus in AD as compared to the MCI and CN subjects. Further, we utilized these features to develop machine learning classifiers to classify AD from MCI and CN subjects and achieved accuracy of 97.78% and 94.41% respectively. These results strengthen the connection of precuneus and fusiform gyrus with AD pathology and opens a new avenue of AD research.
Publisher
Cold Spring Harbor Laboratory