Table 3:
Diagnostic discrimination of iNPH patient group from other subject groups
| Variable | Sensitivity | Specificity | Positive Predictive Value | Negative Predictive Value | Cutoff Value of Each Measure | Area Under Curve |
|---|---|---|---|---|---|---|
| FA value | ||||||
| Compared with AD | 94 | 82 | 89 | 90 | 0.59 | 0.92 |
| Compared with PDD | 94 | 72 | 85 | 89 | 0.59 | 0.89 |
| Compared with control | 94 | 84 | 85 | 94 | 0.59 | 0.91 |
| Compared with AD, PDD, and control | 94 | 80 | 68 | 97 | 0.59 | 0.91 |
| Axial eigenvalue | ||||||
| Compared with AD | 94 | 82 | 89 | 90 | 1.52 | 0.95 |
| Compared with PDD | 100 | 82 | 90 | 100 | 1.48 | 0.96 |
| Compared with control | 100 | 84 | 86 | 100 | 1.50 | 0.98 |
| Compared with AD, PDD, and control | 94 | 83 | 71 | 97 | 1.52 | 0.97 |
Note:—Data (except for the cutoffs of FA values and axial eigenvalues) are shown as percentages. Receiver operating characteristic analysis was used to determine the optimum cutoff values for evaluating the usefulness of FA values and axial eigenvalues in differentiating iNPH from AD, PDD, and controls.