Skip to main content

Table 2 Results of AMD classification based on geostatistical features

From: Semivariogram and Semimadogram functions as descriptors for AMD diagnosis on SD-OCT topographic maps using Support Vector Machine

Feature Fold Sensitivity (%) Specificity (%) Accuracy (%) AUROC Kappa
TR (SV) Average 76.0 42.3 65.8 0.576 0.180
Std 5.7 10.2 4.7 0.071 0.101
Max acc 84.5 61.1 78.9 0.306 0.439
Max kappa 84.5 61.1 78.9 0.306 0.439
TR (SM) Average 91.9 26.0 71.7 0.756 0.204
Std 5.7 13.9 5.3 0.052 0.123
Max acc 96.7 50.0 86.8 0.846 0.541
Max kappa 91.2 63.2 84.2 0.821 0.564
NSR (SV) average 88.8 65.7 81.9 0.802 0.554
Std 4.1 9.6 3.8 0.060 0.091
Max acc 96.5 80.0 92.2 0.861 0.791
Max kappa 96.5 80.0 92.2 0.861 0.791
NSR (SM) Average 90.0 84.3 88.3 0.950 0.724
Std 4.0 7.0 3.2 0.023 0.076
Max acc 98.2 95.5 97.4 0.993 0.936
Max kappa 98.0 96.3 97.4 0.996 0.943
RPEDC (SV) Average 94.2 97.5 95.2 0.989 0.886
Std 3.1 3.2 2.3 0.010 0.054
Max acc 100.0 100.0 100.0 1.000 1.000
Max kappa 100.0 100.0 100.0 1.000 1.000
RPEDC (SM) Average 90.2 91.6 90.5 0.977 0.780
Std 4.5 7.5 3.5 0.014 0.080
Max acc 100.0 95.7 98.7 0.999 0.969
Max kappa 98.0 100.0 98.7 0.992 0.971
RPEDC en face (SV) Average 100.0 0.0 70.2 0.505 0.000
Std 0.0 0.0 4.7 0.008 0.000
Max acc 100.0 0.0 84.4 0.542 0.000
Max kappa 100.0 0.0 71.4 0.500 0.000
RPEDC en face (SM) Average 90.2 41.5 75.5 0.778 0.347
Std 4.7 10.1 4.3 0.051 0.098
Max acc 95.1 62.5 88.3 0.844 0.619
Max kappa 95.1 62.5 88.3 0.844 0.619
  1. The data in italics represents the best values obtained