AI-based methods for detecting and classifying age-related macular degeneration: a comprehensive review

Abstract

This paper explores the advancements and achievements of artificial intelligence (AI) in computer vision (CV), particularly in the context of diagnosing and grading age-related macular degeneration (AMD), one of the most common leading causes of blindness and low vision that impact millions of patients globally. Integrating AI in biomedical engineering and healthcare has significantly enhanced the understanding and development of the CV application to mimic human problem-solving abilities. By leveraging AI-based models, ophthalmologists can improve the accuracy and speed of disease diagnosis, enabling early treatment and mitigating the severity of the conditions. This paper presents a comprehensive analysis of many studies on AMD published between 2014 and 2024, with more than 80% published after 2020. Various methodologies and techniques are examined, particularly emphasizing utilizing different retinal imaging modalities like color fundus photography and optical coherence tomography (OCT), where 66% of the studies used OCT datasets. This review aims to compare the efficacy of these AI-based approaches, including machine learning and deep learning, in detecting and diagnosing different stages and grades of AMD based on the evaluation of different performance metrics using different private and public datasets. In addition, this paper introduces some suggested AI solutions for future work.