. 2023 Aug 11;4(2):100380. doi: 10.1016/j.xops.2023.100380

Table 2.

External Validation (Dataset 1) Classification Accuracy, Precision, Recall (sensitivity), F1 Score, and Receiver Operating Characteristic Analysis with the AUC for the SP A1 Parameter, CBI, and TBI Using Naive Bayes Classifier and for the Processed Pseudoimage Slices Using the Adopted DenseNet 121 Model

Parameter	Class	Recall	Specificity	Precision	F1 Score	Accuracy	Mean AUC ± SE
DenseNet121	Normal	0.98	0.78	0.82	0.89	0.88	0.934 ± 0.026
DenseNet121	Keratoconus	0.78	0.98	0.98	0.87	0.88	0.934 ± 0.026
SP A1	Normal	0.92	0.83	0.84	0.88	0.88	0.954 ± 0.020
SP A1	Keratoconus	0.83	0.92	0.91	0.87	0.88	0.954 ± 0.020
CBI	Normal	0.90	0.97	0.97	0.93	0.94	0.982 ± 0.010
CBI	Keratoconus	0.97	0.90	0.91	0.94	0.94	0.982 ± 0.010
TBI	Normal	1.0	1.0	1.0	1.0	1.00	1.000 ± 0.000
TBI	Keratoconus	1.0	1.0	1.0	1.0	1.00	1.000 ± 0.000

AUC = area under the receiver operating characteristic curve; CBI = Corvis biomechanical index; SE = standard error; SP A1 = stiffness parameter at first applanation; TBI = tomography and biomechanical index.