Premium
P4‐588: END‐TO‐END 3D‐CONVOLUTIONAL NEURAL NETWORK FOR PREDICTING CONVERSION FROM MILD COGNITIVE IMPAIRMENT TO ALZHEIMER'S DEMENTIA
Author(s) -
Bae Jinhyeong,
Stocks Jane,
Heywood Ashley,
Jung Youngmoon,
Karteek Popuri,
Beg Mirza Faisal,
Wang Lei
Publication year - 2019
Publication title -
alzheimer's and dementia
Language(s) - English
Resource type - Journals
SCImago Journal Rank - 6.713
H-Index - 118
eISSN - 1552-5279
pISSN - 1552-5260
DOI - 10.1016/j.jalz.2019.08.136
Subject(s) - dementia , cognitive impairment , convolutional neural network , neuroimaging , artificial intelligence , cognition , psychology , audiology , pattern recognition (psychology) , computer science , neuroscience , medicine , disease
Results: Those classified as NAT as compared to PAT had a significantly greater hippocampus volume (p1⁄4.01), prefrontal volume (p1⁄4.03), and precuneus volume (p1⁄4.008). A higher score on the BNTwas positively correlated with a greater hippocampus volume (p<.001). A better performance on the (SI) was positively correlated with a greater hippocampus volume (p1⁄4.001), prefrontal volume (p1⁄4.004), and posterior cingulate volume (p1⁄4.015). Hippocampus volume was determined to be the best predictor of SI score (p1⁄4.023). Conclusions: NAT as compared to PAT classification serves as a valuable predictor for hippocampus, prefrontal and precuneus volumes. An increased score on the BNT has predictive value for greater hippocampus volume and increased SI performance has predictive value for greater hippocampus, prefrontal and posterior cingulate volume.