Neural Network Segmentation of Optical Coherence Tomography Angiography for Diabetic Retinopathy

Diabetic retinopathy (DR) is a complication of diabetes which is the most common cause of vision loss among people with diabetes. DR damages and alters the structure of the capillary network (microvasculature) in the retina, a light-sensitive tissue that lines the back of the eye and is responsible for vision. Clinicians can analyze the microvasculature through optical coherence tomography angiography (OCTA), an imaging technique which allows for micrometer-scale examination of its structures. This project is focused on developing a fast, robust, and accurate neural network to segment the retinal microvasculature in acquired OCTA images, which will be deployed in the clinic to be used on images acquired from real patients. Segmentation of the microvasculature will allow for accurate quantitative analyses such as average vessel width and density, which will expedite patient treatment and help with research towards an early diagnosis.

Faculty Supervisor:

Marinko Sarunic

Student:

Partner:

University of Zagreb

Discipline:

Engineering

Sector:

Education

University:

Simon Fraser University

Program: