Skip to main content
. 2018 Mar 15;2018:1247430. doi: 10.1155/2018/1247430

Table 5.

Comparison in classification accuracy between FDG-PET and AV45-PET using combination of baseline and second-visit timepoints for all diagnostic pairs.

FDG-PET results AV45-PET results
RBF-SVM Random forest P value RBF-SVM Random forest P value
AD versus NC 91.7% 91.2% 0.010 90.8% 87.9% 0.024
AD versus EMCI 85.7% 85.7% 0.015 80.0% 88.0% 0.024
AD versus LMCI 87.5% 79.2% 0.016 88.9% 81.5% 0.027
EMCI versus NC 63.3% 56.7% <0.01 57.7% 59.7% 0.0203
LMCI versus NC 63.5% 65.4% 0.018 61.2% 55.7% <0.01
EMCI versus LMCI 53.9% 64.1% <0.01 52.2% 56.5% 0.017