Premium
P1‐387: PRINCIPAL AXES OF PHENOTYPIC VARIABILITY IN ALZHEIMER'S DISEASE DERIVED FROM AN FDG‐PET BASED UNSUPERVISED MACHINE LEARNING ALGORITHM
Author(s) -
Jones David T.,
Lowe Val J.,
Graff-Radford Jonathan,
Botha Hugo,
Murray Melissa E.,
Parisi Joseph E.,
Josephs Keith A.,
Machulda Mary M.,
Therneau Terry M.,
Przybelski Scott A.,
Senjem Matthew L.,
Kantarci Kejal,
Boeve Bradley F.,
Knopman David S.,
Petersen Ronald C.,
Jack Clifford R.
Publication year - 2018
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.2018.06.395
Subject(s) - artificial intelligence , dimensionality reduction , principal component analysis , neurodegeneration , phenotype , multivariate statistics , pattern recognition (psychology) , disease , machine learning , medicine , computer science , pathology , biology , genetics , gene
performed to test: 1) the effect of time on pairwise FC for each group separately and 2) the effects of baseline PiB uptake and lacune number (measure of CeVD burden) on pairwise FC in aMCI and svMCI. Nuisance covariates included baseline age, sex and education. Results:Amyloid-b burden was associated with longitudinal declines in DMN FC, with PiB+ patients showing longitudinal DMNFC declines (Figure 1A). In contrast, CeVD burden was associated with longitudinal changes in ECN FC, with svMCI patients showing longitudinal ECN FC increases (Figure 1B). This divergent effect was also observedwhen examining the effect of baseline PiB uptake and lacune number on longitudinal DMN/ECN FC changes in aMCI and svMCI subjects (Figure 2). Conclusions:Our findings suggest that amyloid-b and CeVD burdens have divergent effects on longitudinal functional network changes in aMCI and svMCI patients.