Figure 2
From: Differentiating a pachychoroid and healthy choroid using an unsupervised machine learning approach
Choroidal thickness and vascularity index measurements. Haller layer (solid arrow) and the Sattler-choriocapillaris (dotted arrow) thickness manual measurements in enhanced depth imaging-optical coherence tomography of a healthy subject (A) and a patient with central serous chorioretinopathy (B). The Niblack’s autolocal thresholding method was used for binarization of the image. Then, subfoveal choroidal area was selected using the polygon tool (yellow dots). The dark pixels were selected using the color threshold tool and the area occupied by the dark pixels was defined as the luminal area. (C).
