Novel Uses and Challenges of Artificial Intelligence in Diagnosing and Managing Eyes with High Myopia and Pathologic Myopia

Abstract

Myopia is a global health issue, and the prevalence of high myopia has increased significantly in the past five to six decades. The high incidence of myopia and its vision-threatening course emphasize the need for automated methods to screen for high myopia and its serious form, named pathologic myopia (PM). Artificial intelligence (AI)-based applications have been extensively applied in medicine, and these applications have focused on analyzing ophthalmic images to diagnose the disease and to determine prognosis from these images. However, unlike diseases that mainly show pathologic changes in the fundus, high myopia and PM generate even more data because both the ophthalmic information and morphological changes in the retina and choroid need to be analyzed. In this review, we present how AI techniques have been used to diagnose and manage high myopia, PM, and other ocular diseases and discuss the current capacity of AI in assisting in preventing high myopia.

Keywords: artificial intelligence; deep learning; diagnosis and management; high myopia; machine learning; pathologic myopia.

Figure 1

General workflow of artificial intelligence analyses of high myopia and pathologic myopia.

Figure 2

Representative fundus photographs showing the different types of lesions of maculopathy in eyes with pathologic myopia. (A) Normal fundus image. (B) Tessellated fundus. (C) Diffuse atrophy around optic disc and posterior fundus (blue arrows). (D) Patchy atrophy fundus (white arrows). (E) A fundus image from a left eye with macular atrophy at the center of posterior fundus (black arrow). Patchy atrophy (white arrow) as well as diffuse atrophy background can also be seen. (F) Fundus image with myopic choroidal neovascularization at the center of fundus (yellow arrow). Reprinted from Deep Learning Approach for Automated Detection of Myopic Maculopathy and Pathologic Myopia in Fundus Images, Vol 5, Pages No. 1235–1244, Copyright (2021), with permission from Elsevier.

Figure 3

Grading samples of myopic maculopathy in ocular coherence tomographic (OCT) images. (A). Myopic eye without myopic maculopathy. Each of retinochoroidal layer is clearly seen. (B). Myopic neovascularization (MNV). Hyperreflective materials can be seen above the retina pigment epithelium (RPE), and this component is attenuated in the tissue coherence signals below. (C). Retinoschisis. The splitting of the inner retina from the outer retinal layers with multiple perpendicularly aligned columnar structures connecting the split retinal layers. (D). Dome-shaped macular (DSM). An inward bulging of the retina pigment epithelium above the baseline connecting the RPE lines on both sides away from the DSM. (E,F). Retinal detachment. The neurosensory retina is detached from the RPE. (G,H) Macular hole. A tear above the RPE layer and an anvil-shaped deformity of the cracked edges of the retina. Reprinted from Validation of Soft Labels in Developing Deep Learning Algorithms for Detecting Lesions of Myopic Maculopathy from Optical Coherence Tomographic Images, Copyright (2021), with permission from Wolters Kluwer Health.

Figure 4

Image-driven artificial intelligence in high myopia and pathologic myopia.