Table 6.
Confusion matrix following an SVM based feature selection + MLP classification combination using features from multiple MRI modalities.
| Combination of All Imaging Features (Morphology, Brain Iron Content Marker, Diffusion) | |||||||||||
|---|---|---|---|---|---|---|---|---|---|---|---|
| Class | TP Rate | FP Rate | Precision | Recall | F-Measure | MCC | ROC AUC | Confusion Matrix | Accuracy | ||
| HC | PD | PSP-RS | |||||||||
| HC | 1.000 | 0.000 | 1.000 | 1.000 | 1.000 | 1.000 | 1.000 | 38 | 0 | 0 | 95.1% |
| PD | 0.978 | 0.069 | 0.917 | 0.978 | 0.946 | 0.904 | 0.986 | 0 | 44 | 1 | |
| PSP-RS | 0.800 | 0.012 | 0.941 | 0.800 | 0.865 | 0.840 | 0.983 | 0 | 4 | 16 | |
TP, True Positive; FP, False Positive; MCC, Matthews Correlation Coefficient; ROC AUC, Area under the receiver operating characteristic curve; HC Healthy Controls; PD, Parkinson's disease; PSP-RS, Progressive supranuclear palsy Richardson's syndrome.