Quantification of parafoveal capillary network using a semi-automated algorithm

The quantification of the morphology of the parafoveal capillary network (PCN) in fluorescein angiography (FA) images using a novel semi-automated computerized method.

Using the MatLab R2011 a software we developed an algorithm that detects automatically the parafoveal capillary bed and its branch points as depicted in FA images creating simultaneously an one-pixel-wide skeleton of it. The detection process starts after delineating manually the foveal avascular zone in a cropped 1500μm*1500μm subimage resulting from the original FA image. Thereafter the algorithm calculates the capillary density and the branch points in a circle area with 1000μm radius. The method was also applied on FA images from subjects without diabetes mellitus, diabetics without diabetic retinopathy (DR) signs, patients with non-proliferative DR and patients with proliferative DR in order to assess the PCN morphology metrics for the studied groups.

The PCN density and the parafoveal capillary branch points were estimated for the mentioned subject groups and any significant differences among them were assessed as well.

The described method could serve as a potential tool for the diagnosis and monitoring of PCN diseases and subclinical abnormalities. The assessed metrics reflect the capillary abnormalities in the central 1000μm area across different DR stages.