The use of Artificial Intelligence for Diagnosing Retinopathy of Prematurity – A Systematic Review

Abstract

Purpose: Retinopathy of prematurity (ROP) is a vaso-proliferative disease of the retina associated with prematurity and is well known to be the leading cause of childhood blindness worldwide. Given the prevalence of ROP and the increasing demand for efficient screening solutions, this systematic review aims to update the current development of Artificial Intelligence (AI) technologies for ROP diagnosis and screening, considering the appropriate AI types that align with the specific needs and workloads of ROP screening programs. Methods: We performed a systematic literature review of the English online literature databases by applying a general search strategy on April 20, 2024.  Search phrases included multiple variants of terms including "ROP", "retinopathy of prematurity", "artificial intelligence", "diagnosis", "sensitivity analysis", "specificity", "area under the curve".  Findings: A total of 12 studies were identified from varied countries. Summary: Review of the published literature demonstrate high sensitivity across different studies, indicating their strong potential for early detection of ROP but demonstrating variability in specificity. The review also underscores the importance of domain adaptation and validation across diverse populations to ensure generalizability. AI integration in clinical practice, especially in telemedicine, can enhance early detection, standardize diagnoses, and alleviate the burden on healthcare professionals, particularly in low-resource settings.