Diagnosis of Retinal Diseases Based on Bayesian Optimization Deep Learning Network Using Optical Coherence Tomography Images

Table 5

Performance of VGG16.

Class labels	Precision (%)		Recall (%)		F1-score (%)
	Feature extractor	Fine tuner	Feature extractor	Fine tuner	Feature extractor	Fine tuner
AMD	99.41	100	96	99.71	97.67	99.86
CNV	79.74	87.24	69.71	95.71	74.39	91.28
CSR	83.05	97.77	96.57	100	89.3	98.87
DME	73.83	93.49	62.86	90.29	67.90	91.86
DR	81.44	99.14	84	99.14	82.7	99.14
Drusen	67.19	95.72	60.86	83.14	63.89	88.99
MH	95.41	100	77.14	98.29	85.31	99.14
Normal	62.65	89.81	87.71	95.71	73.1	96.67
Macro average	80.34	95.4	79.36	95.25	79.28	95.23
Weighted average