Detection and quantification of hyperfluorescent leakage by computer analysis of fundus fluorescein angiograms

Abstract

Digital imaging systems can be used for direct acquisition of images of the ocular fundus or for their indirect acquisition from fundus photographs or transparencies. Computerised image processing techniques can then be used to manipulate and quantify features of interest. We describe a method for the detection and quantification of macular leakage in fluorescein angiograms. The rate of change in fluorescence over time is examined on a pixel-by-pixel basis and used to provide a gradient threshold that discriminates pixels displaying leakage from normal pixels. A region-growing technique is then used to enhance the detection of leakage missed by gradient thresholding alone. This report discusses the potential applications of the technique and highlights the methodology required to obtain reproducible results.