Premium
P1‐313: Prediction of MCI converters in the ADNI cohort using patterns of cortical thinning
Author(s) -
Eskildsen Simon,
Fonov Vladimir,
Coupe Pierrick,
Ostergaard Lasse,
Collins Louis
Publication year - 2011
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.2011.05.592
Subject(s) - mcnemar's test , linear discriminant analysis , pattern recognition (psychology) , artificial intelligence , converters , parametric statistics , random forest , medicine , mathematics , computer science , statistics , physics , quantum mechanics , voltage
malized using the standardized uptake value ratio (SUVR) method. In the subject cohort, 29 patients were AD, while the other 90 were HC; 33 of whom were PiB-positive and 57 were PiB-negative. We used a surface based approach to identify the regions where the cortical thickness in AD patients was significantly lower than PiB-negative HC. Individual mean cortical thicknesses were calculated in the significant patch of each AAL region. T-test analysis was performed between the 3 groups (AD, PiB-negative HC, PiB-positive HC) in each tessellated region while controlling for age. Results: In the discriminating regions of AD/PiB-negative (Figure 1), we showed that PiB-positive HC had significantly lower cortical thickness than PiB-negative HC in the hippocampus and amygdala region, in the precuneus and in the superior temporal gyrus in the left hemisphere Table 1). The regions in the right hemisphere showed the same patterns but were not significant.Conclusions: The results showed that atrophy patterns in PiB-positive HC groupwere similar to AD group but to a lesser extent, suggesting an effect of amyloid plaques early on in asymptomatic individuals.