Skip to main content
. 2021 Apr 10;11(4):685. doi: 10.3390/diagnostics11040685

Figure 10.

Figure 10

Real FF-OCT image used to test our models. (a) Total average of 64 frames of cropped volumetric image. (b) Expert-annotated ground truth. (1–16) are averages of every four frames of cropped volumetric image, forming 16 layers as input images for our models.