Table 3 Comparison of actual and predictive outcomes.

From: Development and validation of CNN-MLP models for predicting anti-VEGF therapy outcomes in diabetic macular edema

Outcomes

Architectures

Statics

P-value

Confidence Interval (CI)

BCVA

CNN-MLP

0.263

0.792

[0.013, 1.307]

CNN

-0.802

0.422

[0.524, 0.876]

CST

CNN-MLP

0.004

0.997

[123.246, 483.494]

CNN

0.523

0.603

[257.535, 329.584]

CV

CNN-MLP

1.314

0.194

[9.579, 16.361]

CNN

1.676

0.099

[11.309, 14.131]

CAT

CNN-MLP

0.283

0.778

[230.208, 472.032]

CNN

9.187

< 0.001

[190.984, 304.076]

Back to article page