Table 3 Performance of the deep learning models in the dataset from the Zhongshan Ophthalmic Centre.

From: Deep learning for detecting retinal detachment and discerning macular status using ultra-widefield fundus images

 

AUC

Sensitivity (95% CI)

Specificity (95% CI)

Accuracy (95% CI)

The first model

1.000 (95% CI: 0.999–1.000)

99.5% (98.5–100)

99.5% (98.5–100)

99.5% (98.8–100)

The second model

0.934 (95% CI: 0.846–0.995)

93.5% (84.5–100)

95.3% (92.0–98.6)

95.0% (91.9–98.1)

  1. AUC area under the curve, RD retinal detachment. The first model is used to identify RD. The second model is used to identify macula-on RD
Back to article page