Table 1 Confusion matrix describing performances of the 30° × 30° FAF based DL classifier on detection of EMAP cases.

From: Deep learning model for automatic differentiation of EMAP from AMD in macular atrophy

30° × 30° FAF

CNN

 

Positive

Negative

Ground truth

Positive

110

20

Negative

31

180

  1. CNN, convolutional neural network; EMAP, extensive macular atrophy and pseudodrusen-like appearance; FAF, fundus autofluorescence.
Back to article page