Fig. 4

Image enhancement and en-face image creation using OCT B-scan data. (a) Original volumetric-scan image. (b) Denoised image. Deep-learning-based noise and shadow reduction have been performed for image enhancement. (c) Flattened image based on the choroid–sclera interface generated from (b). (d) En-face image generated from the flattened images. Hypo-reflective spots correspond to the cross-sections of perforating scleral vessels in the OCT B-scan image (c) (yellow arrows). OCT, optical coherence tomography.