The NEI 40th Anniversary Symposia Series
Imaging Macular Degeneration with OCT
Current OCT instruments have given us the capability to acquire very large datasets containing information about extended retinal regions. The availability of this information has the potential to lead to a better understanding of disease processes and therapeutic responses. One crucial step in this direction is the development of algorithms to analyze the information in an efficient, quantitative manner. This means algorithms that can recognize clinically relevant feature in the OCT datasets and produce meaningful descriptors with well understood characteristics (in particular properties such as reproducibility, robustness, and accuracy).
We discuss this problem, illustrating some possible approaches to producing clinically useful descriptions of macular degeneration from OCT images. In particular we focus in some depth on measurements of the geometry of drusen in patients with non-exudative age-related macular degeneration (AMD). In this context we report clinical data on the application of a new algorithm that can automatically characterize drusen geometry in three dimensional space, in particular drusen area and volume, on a set of circular regions centered at the fovea. Our results show that drusen area and volume measurements were both qualitatively and quantitatively reproducible. An interesting fact is that considerable changes in the values of these parameters were observed over the follow-up time. These data suggest that an appropriate analysis of OCT datasets has the potential to greatly improve our understanding of drusen and that drusen may reveal a more dynamic natural history than assumed based on en face fundus photographic assessment.
Giovanni Gregori, Ph.D.
Research Assistant Professor of Ophthalmology
Bascom Palmer Eye Institute
University of Miami Miller School of Medicine
Giovanni Gregori, Ph.D. is a mathematician trained in optics at Stanford. He did post-doctoral work in the United States and Europe. He has been a Research Assistant Professor in the Bascom Palmer Eye Institute (BPEI) at the University of Miami Miller School of Medicine since 2002 (http://bascompalmer.org/site/current/current_centers.asp#center). Dr. Gregori has worked extensively on image processing and segmentation algorithms. The main focus of Dr. Gregoriís research is on the development of quantitative tool for the analysis of medical images, in particular those acquired using Optical Coherence Tomography (OCT) technology. Dr. Gregori has been a leader in the development of spectral domain OCT. His group introduced several new powerful mathematical algorithms for processing large OCT datasets. These techniques were used to generate the first accurate description of the retinal geometry in vivo. A noteworthy application of this work is the use of OCT datasets to derive a novel quantitative understanding of the effects of drusen on the Retinal Pigment Epithelium and the photoreceptors. Dr. Gregoriís work in OCT segmentation is the basis of several patents and commercial licenses. A number of clinical studies that utilize the advanced OCT software developed by Dr. Gregori are currently underway.