Usage of deep learning algorithms for Ocular Toxoplasmosis tracking on retina images

• Grant number: 22/11378-3
• Support Opportunities: Scholarships in Brazil - Program to Stimulate Scientific Vocations
• Start date: January 04, 2023
• End date: February 23, 2023
• Field of knowledge: Health Sciences - Medicine
• Principal Investigator: Rubens Belfort Mattos Junior
• Grantee: Yanka Raíssa Ribeiro da Silva
• Host Institution: Escola Paulista de Medicina (EPM). Universidade Federal de São Paulo (UNIFESP). Campus São Paulo. São Paulo , SP, Brazil

Abstract

Toxoplasmosis is an infectious disease whose etiology is Toxoplasma gondii protozoa, withsignificant world prevalence and social morbidity. Ocular toxoplasmosis is an important cause of non-reversible blindness, with typical retinochoroiditis lesions in 90% of cases. The antimicrobial treatment leads to disease remission, but it does not eliminate theprotozoa from the organism, which culminates in more than half of the ocular toxoplasmosis patients presenting disease recurrence. The ophthalmologist distribution is scarce worldwide, especially in low and mid-incomecountries, with uveitis and retina specialists even rarer. A computer vision artificial intelligence algorithms enable an automated ocular toxoplasmosis screening through color retinal fundus photos, guiding follow-up, promoting educational measures, and the need for specialist follow-up in selected cases. (AU)