Automatic CDR Estimation for Early Glaucoma Diagnosis

Table 2

Bagged tree model performance evaluation under Dataset2 database. The three classes are background (class 1), optic disc (class 2), and cup (class 3). (a) Without postprocessing, (b) smoothed with AC, and (c) smoothed and ellipse fitting.

Indicator	a (%)	b (%)	c (%)
Accuracy	94.54	94.75	94.61
Class-wise-averaged accuracy	79.66	81.95	81.19
Sensibility class 1	97.01	96.63	96.34
Sensibility class 2	73.07	83.02	83.18
Sensibility class 3	81.45	76.20	78.75
Specificity class 1	90.30	93.61	94.47
Specificity class 2	96.36	95.73	95.56
Specificity class 3	99.01	99.43	99.33
Positive predictive value class 1	98.68	99.13	99.24
Positive predictive value class 2	62.37	61.61	60.75
Positive predictive value class 3	77.93	85.10	83.58
Negative predictive value class 1	80.11	78.74	77.47
Negative predictive value class 2	97.74	98.55	98.57
Negative predictive value class 3	99.20	98.98	99.09