Predicting Retinal Sickling Events using a Machine Learning

Authors: Giancarlo Paternoster, Dr. Umar Mian

Projected Project Completed: May 2025

This project leverages the power of machine learning to enhance the detection of retinal abnormalities in patients with sickle cell disease. By utilizing fundus camera images, the model is trained to identify regions in the retina prone to sickling events, a complication that can lead to severe vision loss if left untreated. The training dataset includes annotated images highlighting early markers of vascular stress and retinal ischemia, enabling the model to recognize subtle patterns that may not be immediately apparent to the human eye.

The resulting machine learning algorithm offers a non-invasive, highly sensitive tool to predict sickling-prone areas in the retina. This innovation holds the potential to improve early intervention strategies, aiding ophthalmologists in tailoring treatments to prevent further complications. By combining advanced imaging with artificial intelligence, the project bridges the gap between cutting-edge technology and patient-centered care in the field of retinal health.