Research Article
Inclusion of Neuropsychological Scores in Atrophy Models Improves Diagnostic Classification of Alzheimer’s Disease and Mild Cognitive Impairment
Table 3
Optimal accuracy obtained for each classification pair.
| Classification | Number of features for maximum accuracy | Accuracy | Sensitivity | Specificity | Precision | Measure |
| CN versus AD | 7 | 0.939 ± 0.012 | 0.963 ± 0.021 | 0.895 ± 0.011 | 0.938 ± 0.026 | 0.948 ± 0.012 | CN versus EMCI | 5 | 0.856 ± 0.023 | 0.966 ± 0.012 | 0.741 ± 0.063 | 0.808 ± 0.062 | 0.871 ± 0.036 | CN versus LMCI | 14 | 0.908 ± 0.031 | 0.968 ± 0.024 | 0.837 ± 0.044 | 0.882 ± 0.054 | 0.919 ± 0.029 | EMCI versus LMCI | 10 | 0.736 ± 0.024 | 0.743 ± 0.055 | 0.727 ± 0.061 | 0.777 ± 0.082 | 0.758 ± 0.085 | EMCI versus AD | 8 | 0.945 ± 0.032 | 0.946 ± 0.022 | 0.950 ± 0.023 | 0.870 ± 0.016 | 0.951 ± 0.012 | LMCI versus AD | 5 | 0.901 ± 0.024 | 0.907 ± 0.044 | 0.893 ± 0.012 | 0.923 ± 0.033 | 0.914 ± 0.038 |
|
|