Deep Learning for Prediction of AMD Progression

We are delighted to announce that our abstract, Deep Learning for Prediction of AMD Progression was accepted for presentation at the 2018 Annual Meeting of the Association for Research in Vision and Ophthalmology (ARVO).   The research was performed in collaboration with Professor Sobha Sivaprasad’s group at the UCL Institute of Ophthalmology and Moorfields Eye Hospital.

Age-related macular degeneration (AMD) is a leading cause of vision loss for people in their 50’s and older. AMD progresses in distinct stages from early, to intermediate, to advanced. In advanced AMD, blood vessel growth in the eye causes damage leading to irreversible vision loss. Patients can progress to advanced AMD without showing any symptoms or measurable change, underscoring the importance of determining which patients are at the highest risk for AMD progression. Optical coherence tomography (OCT) is an inexpensive and noninvasive way of imaging the eye and looking at exactly the retinal structure that changes through the course of the disease. We have developed a deep learning-based artificial intelligence system to help determine which patients will progress before they do. The system analyzes images of the retina from an OCT system, focusing on specific regions of interest, and offers a prediction of disease progression. Accurate prediction of progressing patients can lead to earlier intervention which, with treatment, can help prevent severe loss of vision.