Skip to main content

Table 2 Univariate classifier (logistic regression) of dementia status with different neuropathological features. The cross-validated mean accuracy and error was computed. Plaque score, measures of CAA and Braak stage all performed slightly better than Thal phase as dementia predictors

From: Epidemiological pathology of Aβ deposition in the ageing brain in CFAS: addition of multiple Aβ-derived measures does not improve dementia assessment using logistic regression and machine learning approaches

Features of univariate classifierMean accuracy (+/− error)
Thal Phase0.5957 (+/−  0.0040)
Plaque score0.6166 (+/−  0.0048)
CAA Type0.6243 (+/−  0.0046)
CAA Areas0.6222 (+/−  0.0047)
Braak Stage0.6179 (+/−  0.0051)
Age (One hot encoded)0.6540 (+/−  0.0038)
Brain weight0.5721 (+/− 0.0032)