Retinal image segmentation

• Grant number: 24/16942-0
• Support Opportunities: Scholarships in Brazil - Scientific Initiation
• Start date: October 01, 2024
• End date: September 30, 2025
• Field of knowledge: Physical Sciences and Mathematics - Computer Science - Computing Methodologies and Techniques
• Principal Investigator: Nina Sumiko Tomita Hirata
• Grantee: João Paulo Menezes Linaris
• Host Institution: Instituto de Matemática e Estatística (IME). Universidade de São Paulo (USP). São Paulo , SP, Brazil
• Associated research grant: 22/15304-4 - Learning context rich representations for computer vision, AP.TEM

Abstract

Certain anomalies in retinal images are crucial for detecting eye diseases. Among these diseases, one that is relatively common is Diabetic Retinopathy, resulting from the impairment of microvasculature due to Diabetes Mellitus. Early detection of diabetic retinopathy can prevent a significantly prevalent outcome, which is vision loss. In this scientific initiation project, we are interested in the segmentation of certain structures in retinal images. Segmentation is a step that precedes the extraction of quantitative metrics about the characteristics of the eye, which are important in the diagnostic evaluation. We aim to study and implement computational methods based on deep learning for the segmentation of these structures, with a focus on blood vessels. The methods will be tested on public image datasets and possibly on images collected by handheld fundus camera.