Table 5 Random Forest performance measures for AD progression prediction of MCI subjects based on CFA, MRI, PET, genetics, and MH modalities (\(MS2\) test dataset; Second Layer).
Modalities used | Precision (%) | Recall (%) | Accuracy (%) | F1-score (%) | AUC |
|---|---|---|---|---|---|
All | 87.50 | 87.50 | 87.76 | 87.75 | 0.953 |
CFA | 91.30 | 87.50 | 89.80 | 89.81 | 0.926 |
MRI | 56.67 | 70.83 | 59.18 | 59.66 | 0.691 |
PET | 72.73 | 66.67 | 71.43 | 71.44 | 0.812 |
Genetics | 76.00 | 79.17 | 77.55 | 77.58 | 0.787 |
MH | 57.14 | 50.00 | 57.14 | 57.07 | 0.562 |
CFA + MRI | 90.91 | 83.33 | 87.76 | 87.86 | 0.903 |
CFA + PET | 95.45 | 87.50 | 91.84 | 91.95 | 0.955 |
CFA + genetics | 84.00 | 87.50 | 85.71 | 85.75 | 0.926 |
CFA + MH | 91.30 | 87.50 | 89.80 | 89.81 | 0.918 |
CFA + PET + MRI | 88.00 | 91.67 | 89.80 | 89.83 | 0.949 |
CFA + PET + genetics | 91.67 | 91.67 | 91.85 | 91.83 | 0.956 |
CFA + PET + MH | 87.50 | 87.50 | 87.76 | 87.75 | 0.943 |
CFA + PET + genetics + MRI* | 91.70 | 91.70 | 91.86 | 91.84 | 0.963 |
CFA + PET + genetics + MH | 87.50 | 87.50 | 87.76 | 87.75 | 0.948 |
