Quantification of leakage by FAF imaging is usually performed by qualitative scoring. Some authors (Wigg et al.) have proposed quantitative methods. Using randomly invented data from lesions, a reinterpretation of the method and a framework is proposed in this Python script. This a first aproximation to that approach.
Laser induced choroidal neovascularization is a widely used animal model for recapitulating wet AMD disease. Lesions are caused in the animal using an ophthalmic laser. The main idea is to simulate the neovascularization process which takes place during the natural course of the disease.
Fondus fluorescence angiography (FAF) is one the most used method for addressing the neovascularization. This type of images allows the measurement of a wide number of parameters. One of them is, when present, the leakage. This parameter is defined as an hyperfluorescent area of the image that usually correspond to a liquid extravasation from retinal blood vessels.
Usually, qualitative approaches are done using a scale grade (from 0 to 3). This scale is based on subjective observation from several graders. The limitations are the inability to quantify as well as the fact that important factors such as area are overlooked.
Quantitative method has been suggested in some recent publications. This is my first python approach to one of those method.