Diagnosis of Retinal Diseases Using Artificial Intelligence and Deep Learning Approaches

Abstract

Abstract In the past decade, the power and speed of the artificial intelligence (AI), machine learning (ML), and deep learning (DL) techniques in medical image recognition are evolving rapidly, and will become more integrated into ophthalmic care for data analysis, segmentation, early diagnosis and timely treatment of ocular disorders. AI and DL technologies have provided ophthalmologists with automatic tools for early diagnosis and confirmation, image reading, corneal topography mapping, intraocular lens calculations, and timely treatment of eye disorders. Generally, ophthalmic image analysis is hard, expensive, and time-consuming. To overcome these difficulties, in recent years, several computational techniques in different ophthalmic disorders were developed. In this regard, here, we have provided a comprehensive review on the existing computational techniques, a number of different available software packages and toolkits, and some available DBs in retina diseases diagnosis include methods for diabetic retinopathy, age-related macular degeneration, glaucoma, retinopathy of prematurity, other ophthalmic disorders. Additionally, an overview of the potential impacts of the current AI, ML, and DL methods and their applications in the early detection and treatment of the four diseases and some of the potential challenges that may happen have been described. The aim of the describing computational approaches and tools is that ophthalmologists and ophthalmic scientists might take advantage of these computational methods to early diagnosis and timely treatment of ocular disorders, ophthalmic image analysis, and to find other methods to gain complementary information for their question of interest.