Promising Artificial Intelligence-Machine Learning-Deep Learning Algorithms in Ophthalmology
Lokman Balyen,Tunde Peto
DOI: https://doi.org/10.22608/APO.2018479
Abstract:The lifestyle of modern society has changed significantly with the emergence of artificial intelligence (AI), machine learning (ML), and deep learning (DL) technologies in recent years. Artificial intelligence is a multidimensional technology with various components such as advanced algorithms, ML and DL. Together, AI, ML, and DL are expected to provide automated devices to ophthalmologists for early diagnosis and timely treatment of ocular disorders in the near future. In fact, AI, ML, and DL have been used in ophthalmic setting to validate the diagnosis of diseases, read images, perform corneal topographic mapping and intraocular lens calculations. Diabetic retinopathy (DR), age-related macular degeneration (AMD), and glaucoma are the 3 most common causes of irreversible blindness on a global scale. Ophthalmic imaging provides a way to diagnose and objectively detect the progression of a number of pathologies including DR, AMD, glaucoma, and other ophthalmic disorders. There are 2 methods of imaging used as diagnostic methods in ophthalmic practice: fundus digital photography and optical coherence tomography (OCT). Of note, OCT has become the most widely used imaging modality in ophthalmology settings in the developed world. Changes in population demographics and lifestyle, extension of average lifespan, and the changing pattern of chronic diseases such as obesity, diabetes, DR, AMD, and glaucoma create a rising demand for such images. Furthermore, the limitation of availability of retina specialists and trained human graders is a major problem in many countries. Consequently, given the current population growth trends, it is inevitable that analyzing such images is time-consuming, costly, and prone to human error. Therefore, the detection and treatment of DR, AMD, glaucoma, and other ophthalmic disorders through unmanned automated applications system in the near future will be inevitable. We provide an overview of the potential impact of the current AI, ML, and DL methods and their applications on the early detection and treatment of DR, AMD, glaucoma, and other ophthalmic diseases.