Skip to main content

Table 4 Comparison of CAA, Thal phase and plaque score importance in classifiers that also use age, brain weight and Braak stage. Adding any one of the three Aβ-related parameters gave a similar prediction performance compared to the best performing multivariable model. Accuracy and confidence intervals are listed for dementia prediction when each neuropathology feature is omitted

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

Classifier (including age, brain weight and Braak score)CAAThal PhasePlaque score
Logistic regression0.6756 (+/−  0.0032)0.6885 (+/−  0.0036)0.6741 (+/−  0.0030)
Decision Tree0.6798 (+/−  0.0042)0.6940 (+/−  0.0043)0.6980 (+/−  0.0044)
LDA0.6705 (+/−  0.0033)0.6748 (+/−  0.0034)0.6737 (+/−  0.0034)