Table 3.
Accuracy of new Corvis ST parameters in identifying SKC and KC eyes.
Parameters | AUC (95%CI) | Youden index | Cut off | Sensitivity (%) | Specificity (%) |
---|---|---|---|---|---|
SKC versus control | |||||
Max inverse radius | 0.626 (0.555, 0.693) | 0.23 | > 0.19 | 50 | 73 |
DA ratio max (2 mm) | 0.684 (0.615, 0.748) | 0.33 | > 4.47 | 81 | 52 |
Pachy slope | 0.538 (0.466, 0.608) | 0.16 | > 44.30 | 46 | 70 |
DA ratio max (1 mm) | 0.673 (0.603, 0.737) | 0.31 | > 1.60 | 80 | 51 |
ARTh | 0.618 (0.547, 0.686) | 0.30 | ≤ 434.02 | 51 | 79 |
Integrated radius | 0.668 (0.598, 0.733) | 0.33 | > 8.94 | 81 | 52 |
SP-A1 | 0.753 (0.687, 0.811) | 0.40 | ≤ 107.30 | 66 | 74 |
CBI | 0.703 (0.635, 0.766) | 0.38 | > 0.05 | 56 | 82 |
KC versus SKC | |||||
Max inverse radius | 0.889 (0.837, 0.929) | 0.65 | > 0.20 | 87 | 78 |
DA ratio max (2 mm) | 0.903 (0.853, 0.940) | 0.69 | > 5.29 | 80 | 89 |
Pachy slope | 0.908 (0.859, 0.944) | 0.69 | > 55.49 | 73 | 96 |
DA ratio max (1 mm) | 0.843 (0.785, 0.891) | 0.61 | > 1.67 | 79 | 82 |
ARTh | 0.980 (0.949, 0.994) | 0.88 | ≤ 319.83 | 89 | 99 |
Integrated radius | 0.932 (0.887, 0.962) | 0.72 | > 10.86 | 81 | 91 |
SP-A1 | 0.893 (0.841, 0.932) | 0.64 | ≤ 82.03 | 72 | 92 |
CBI | 0.992 (0.967, 0.999) | 0.93 | > 0.87 | 97 | 96 |
KC versus control | |||||
Max inverse radius | 0.923 (0.878, 0.956) | 0.71 | > 0.19 | 95 | 76 |
DA ratio max (2 mm) | 0.953 (0.913, 0.978) | 0.77 | > 5.13 | 83 | 94 |
Pachy slope | 0.919 (0.873, 0.953) | 0.69 | > 48.46 | 87 | 82 |
DA ratio max (1 mm) | 0.914 (0.866, 0.949) | 0.71 | > 1.66 | 86 | 85 |
ARTh | 0.992 (0.967, 0.999) | 0.96 | ≤ 364.89 | 97 | 99 |
Integrated radius | 0.965 (0.929, 0.986) | 0.79 | > 10.50 | 86 | 93 |
SP-A1 | 0.973 (0.939, 0.991) | 0.84 | ≤ 91.64 | 87 | 97 |
CBI | 0.999 (0.980, 1.000) | 0.99 | > 0.50 | 99 | 100 |
SKC Subclinical keratoconus, KC Keratoconus, DA Deformation amplitude, ARTh Ambrósio’s relational thickness horizontal, bIOP Biomechanical corrected intraocular pressure, SP-A1 Stiffness parameter at the first applanation, CBI Corvis biomechanical index.