MCI versus normal aging Unit Odds Ratio (95%CI) Significance
Non-PVC model
gender 1 1.855 (0.995 - 3.462) 0.0521
education 2.9586 0.881 (0.655 - 1.185) 0.4027
age 6.5536 0.597 (0.439 - 0.812) 0.001
entorhinal 337.3 0.681 (0.478 - 0.970) 0.0335
banks STS 341.8 0.773 (0.558 - 1.071) 0.1214
caudal middle frontal 876.3 0.742 (0.538 - 1.023) 0.0687
inferior temporal 1279 0.744 (0.520 - 1.064) 0.105
superior temporal 1166.3 1.418 (0.948 - 2.122) 0.0894
hippocampus 532.6 0.345 (0.222 - 0.537) <0.0001
Cortical thickness + subcortical volume residuals
gender 1 1.131 (0.645 - 1.985) 0.6677
education 2.9586 0.907 (0.703 - 1.171) 0.4536
age 6.5536 0.815 (0.625 - 1.064) 0.1331
paracentral gyrus 371.1 1.261 (0.949 - 1.674) 0.1094
temporal pole 225.5 1.478 (1.134 - 1.927) 0.0039
amygdala 21.0682 0.667 (0.499 - 0.890) 0.0059
Gray matter mask
gender 1 1.507 (0.869 - 2.612) 0.1444
education 2.9574 0.857 (0.658 - 1.117) 0.2539
age 6.4782 0.798 (0.611 - 1.041) 0.0965
entorhinal 1349.3 0.549 (0.361 - 0.835) 0.0051
fusiform 1235.9 2.439 (1.342 - 4.434) 0.0035
inferior temporal 1301 0.576 (0.341 - 0.975) 0.04
isthmus of the cingulate 2007.9 0.588 (0.386 - 0.895) 0.0133
The top model shows the non-pvc model (c = 0.810), followed by the cortical thickness + subcortical volume residuals (c = 0.678), and the GMM on the bottom (c = 0.688).
Table 3: Stepwise logistic regression models for differentiating between MCI and normal aging groups before and after partial volume correction.
Goto home»