A serious and permanent eye condition called myopic maculopathy is becoming more common.
Myopia, often referred to as nearsightedness, is increasingly prevalent, particularly among children.
Experts forecast that by 2050, around 50% of the global population will be affected by myopia. Researchers attribute part of this rise to an increase in “near work,” which involves focusing on close objects like smartphones and screens.
For many individuals, the inability to see distant objects can be managed easily with glasses or contact lenses. However, for some, this condition can escalate into a more severe issue known as myopic maculopathy.
A research team at Arizona State University’s School of Computing and Augmented Intelligence is working on innovative diagnostic tools that utilize artificial intelligence (AI) to improve screening for this condition. Their findings were recently published in the peer-reviewed journal JAMA Ophthalmology.
Myopic maculopathy occurs when the central part of the eye, responsible for sharp, direct vision, becomes elongated and damaged. Over time, the eye takes on a more football-like shape rather than remaining spherical, leading to visual distortion.
This serious illness is the foremost cause of severe vision impairment or blindness. In 2015, myopic maculopathy was accountable for visual impairment in 10 million individuals. If trends continue, more than 55 million people are expected to experience vision loss, and around 18 million may become blind from this condition by 2050.
Due to the irreversible nature of myopic maculopathy, early intervention is critical. Identifying the disease at an early stage can improve outcomes, which is especially important for children. Ophthalmologists have the option to prescribe specialized contact lenses or eye drops that can slow the disease’s progression.
Yalin Wang, a professor in computer science and engineering at Fulton Schools, emphasizes that technological advancements can provide crucial solutions.
“AI is leading a revolution that uses global knowledge to enhance diagnostic accuracy, particularly in the initial stages of the disease,” he explains. “These developments will lower medical expenses and enhance the quality of life across communities.”
A challenge to see things in a new light
To address this issue, the Medical Image Computing and Computer Assisted Intervention (MICCAI) Society launched a challenge in 2023. This organization aims to foster innovation in biomedical research and invited experts to enhance computer-aided screening systems for retinal imagery.
Currently, myopic maculopathy is diagnosed through optical coherence tomography scans, which use reflected light to capture images of the back of the eye. An ophthalmologist usually manually examines these scans, which can be a lengthy process requiring specialized skills.
Wang and his team from the Geometry Systems Laboratory responded to this challenge and were among the winners.
In the initial phase of their project, Wang and his colleagues—including computer engineering PhD student Wenhui Zhu and neurologist Dr. Oana Dumitrascu—focused on classifying myopic maculopathy. The disease has five classifications that indicate its severity. Accurate classification aids ophthalmologists in providing tailored and efficient treatment options for patients.
The researchers developed new AI algorithms named NN-MobileNet. These algorithms help software effectively analyze retinal images and accurately predict the classification of myopic maculopathy.
Then, they shifted their focus to harnessing a type of AI called deep neural networks to estimate the spherical equivalent from retinal scans. The spherical equivalent is crucial for determining a patient’s refractive error, which is needed for glasses or contacts prescriptions. By utilizing deep neural networks, the researchers enable computers to analyze extensive data sets and apply AI-driven algorithms for meaningful insights.
With enhanced accuracy in measuring the spherical equivalent, doctors can provide better treatment recommendations. Therefore, the team devised new algorithms concentrating on data quality and relevance, achieving remarkable results while minimizing computing requirements. The outcomes were also published in JAMA Ophthalmology.
Lastly, Wang collaborated with other winning teams from the MICCAI challenge on a third research paper that summarized their findings, published in JAMA Ophthalmology in September. Researchers from various universities shared their discoveries to inspire further breakthroughs in the early diagnosis of myopic maculopathy and to enhance global health care outcomes.
A better vision for global health
Wang shares that one of the driving forces behind his work is addressing health disparities.
“Individuals in rural areas often face challenges accessing advanced imaging tools and healthcare providers,” he notes. “Once AI-enabled technologies become more accessible, they will significantly elevate the quality of life globally, particularly in developing nations.”
Ross Maciejewski, director of the School of Computing and Augmented Intelligence, highlights that Wang’s project exemplifies the impactful research being conducted by faculty in the medical field.
“With the rising incidence of both myopia and myopic maculopathy, it’s crucial to find solutions to prevent vision loss and assist healthcare professionals in providing optimal treatment options for their patients,” Maciejewski states. “Yalin Wang’s innovative research is a thoughtful application of artificial intelligence to tackle this urgent medical challenge.”
