Artificial intelligence-driven diagnosis for age-related macular degeneration bridging pathology and engineering: a survey

Age-related macular degeneration (AMD) is the primary reason for severe visual impairments, making early diagnosis critically important. This paper provides a comprehensive review of the methods used to support screening and diagnostic decisions, focusing on four categories: early, intermediate, and advanced stages of AMD, in addition to AMD across all stages. In this regard, a reference framework is initially proposed to describe research perspectives in pathology. Utilizing this framework, a literature review is conducted to identify the most reliable demographic, environmental, and comorbidity-related risk factors, clinical symptoms, and various aspects of AMD pathology, setting the necessary prerequisites for subsequent sections. The potential application of risk factors is also explained for personalized medicine. While phenotypic risk factors and genetic variants play a crucial role in predicting the progression of AMD, it is more vital to examine demographic and environmental factors at earlier stages for developing effective prevention plans. Therefore, the selection of appropriate risk factors emerges as a critical area of research. Afterward, we present a comparative analysis of different screening and diagnostic methods pertinent to AMD from an industrial engineering perspective. This analysis brings attention to the suite of artificial intelligence (AI) to describe, analyze, and evaluate diagnostic models, thereby providing a reference outline for clinical practice. AI methods can automate the interpretation of retinal images, serving as a supportive tool for clinical decision-making to improve the management of disease progression. In general, this survey highlights the necessity of developing more integrated methods to support decisions at different planning levels.