Of the 5 percent of North Americans who develop diabetes, 4-15 percent develop macular edema about 8000 per annum register as legally blind. It is simple for the clinician to diagnosis localized (rather than diffuse) edema that has a marked (2 fold or higher) increase in retinal thickness. However, sensitivity to small increases and diffuse edema is not possible. These limits prevent early diagnosis, follow-up monitoring and the accuracy and completeness of laser photocoagulation treatment. A software algorithm is proposed which, in conjunction with scanning laser tomography, will help the physician detect subtle or diffuse increases in thickness by providing and absolute measure of it and by providing a mapping of it over the retinal field of high x, y resolution (about 10 microns). This will help the physicians to see edema early and improve the utility and completeness of laser treatment. Profiles of reflected laser-light intensity, which represent the reflectance of tissue along the z axis (depth into the retina), are collected within a series of optically sectioned images. These profiles are affected by the retinal thickness and have information about the thickness contained within them. They will be processed using a special maximum likelihood parameter estimation algorithm to extract the retinal thickness across the retina, and especially in the region in the around the macula. The objective of this research is to implement prototype software and test this software for feasibility of its principles. The feasibility criterion is based upon showing that the resulting retinal thickness measurements agree with benchmark measurements taken by optical coherence tomography.
