An investigation into the PCV subtype

Research Overview
Age-related macular degeneration (AMD) is a prominent cause of vision loss in the United States. The disease can be classified into several subtypes. One of these subtypes, polypoidal choroidal vasculopathy (PCV), is characterized by the pathologic formation of polyps and a branching vascular network (BVN). Patients with PCV are at increased risk for vision-affecting developments such as pigment epithelium detachments and massive subretinal/submacular hemorrhage. In part due to this increased risk, studies have indicated that PCV may be more chronic and result in a higher incidence of vision loss than other AMD subtypes. At the same time, the disease’s pathogenesis and prognosis remain inadequately characterized.
There are two major difficulties that complicate the study of PCV. The first is that imaging of PCV is conventionally reliant on an invasive, dye-injection based method. Invasive techniques are ill-suited as routine imaging tools. This means not only that it is difficult to accrue sufficient data to reveal PCV evolution, but that in many cases the disease is not diagnosed until it has reached an advanced state. Second, PCV is far more prevalent in populations of Asian descent than European.

Research Objectives
Develop tools to transform PCV imaging by enabling non-invasive, automated feature detection
Train separate networks to segment polyps and the branching vascular network, and to automatically diagnose PCV from OCTA data volumes
Investigate the prognostic value of PCV features
Develop automated diagnostic approaches and feature characterization