An overview of artificial intelligence use in diabetic retinopathy treatment: a narrative review

Abstract

Background: Artificial intelligence (AI), usually called machine intelligence, is a scientific field that allows robots thinking like people. The most common diseases that cause visual impairment and blindness, such as age-related macular degeneration (ARMD), cataract, glaucoma, and diabetic retinopathy (DR), have recently been subject to deep learning-based on AI screening and prediction models. Aim: to discuss the core ideas of AI and its apply to DR, as well as to evaluate the currently faced issues and the future of ophthalmology. Methods: In this review, English studies from common databases such as Pubmed/MEDLINE, Google Scholar, Web of Science, Scopus, and the Cochrane Library with the keywords "Machine Learning," "Artificial Intelligence," "Deep Learning," combined with keywords, involving "diabetic retinopathy" were involved. The end date for this review is April 2022. Scientific novelty: This paper illustrates the core AI concepts and their application in diabetic retinopathy. The current ophthalmology issues and future opportunities, offering undiscovered knowledge in this field are also analysed. It will raise community knowledge of employing AI and reveal new capabilities in the analysis of ocular disorders to present the fundamental idea of AI regarding its therapeutic applications. Conclusion: Medical professionals can make quick and precise decisions using AI technologies in orded to analyse massive volumes of data, such as physiological imaging and clinical presentations. It is believed that over time, AI systems will become more precise and successful at predicting the onset and course of DR.