This High Schooler Developed an A.I. Tool to Diagnose Autism and ADHD Using the Retina

Edward Kang’s RetinaMind analyzes patients’ retinal images and accurately diagnoses neurodevelopmental disorders 89 percent of the time

Ramsha Waseem | Freelance writer

May 11, 2026 12:22 p.m.

Share

Copy linkEmailSMSFacebookXRedditLinkedInBlueskyPrint

Add as preferred source

Three years ago, Edward Kang was sifting through research papers for ideas for a school project when he came across an intriguing study by researchers at the Chinese University of Hong Kong that used retinal images to diagnose autism.

“I thought it was fascinating and really unintuitive that you can use something like the eye to understand what’s happening in the brain,” says Kang, now a 17-year-old high school senior at Bergen County Academies in Hackensack, New Jersey.

The teenager set out to make the researchers’ model even more accurate. Ultimately, these efforts led him to develop RetinaMind—an A.I. tool that diagnoses autism spectrum disorder and attention deficit hyperactivity disorder using retinal images.

Diagnosing neurodevelopment disorders

Affecting 1 in 54 children in the United States, autism spectrum disorder (ASD) is the fastest-growing neurodevelopmental disorder in the country. Attention deficit hyperactivity disorder (ADHD) is one of the most common childhood disorders, experienced by nearly seven million children in the U.S.

“Both are neurologically based conditions that are described by development of skills or by unusual or problematic behaviors,” says Paul Lipkin, a neurodevelopmental pediatrician at Kennedy Krieger Institute and professor of pediatrics at Johns Hopkins Medicine. Considered to be “behavioral phenotypes” with no biomarkers, the two disorders are derived from brain functions. They are also usually accompanied by other brain-based developmental or behavioral issues. “In the case of development,” explains Lipkin, “those affected by autism and/or ADHD frequently have intellectual or learning as well as language disabilities and motor coordination problems.”

Though studies have found that early intervention—especially in the case of autism—can result in better long-term effects for children, early diagnosis of these disorders is difficult. Currently, with no physical tests to diagnose autism and ADHD, medical professionals often use developmental and behavioral tests like the American Psychiatric Association’s Diagnostic and Statistical Manual, the Autism Diagnostic Observation Schedule and Conners Rating Scales to diagnose the two respective disorders.

“My hope is that RetinaMind will enable earlier diagnoses for neurodevelopmental disorders than currently possible, unlocking earlier treatment and, therefore, a higher quality of life for the millions of patients of autism and ADHD around the world,” says Kang.

The road to developing RetinaMind

To begin his project, Kang taught himself how to code and the basics of machine learning. “I don’t really come from a programming background,” he confesses. “I looked at a lot of different tutorials online.” He also enrolled in a few online classes.

The first iteration of his model was a basic version of the Convolutional Neural Network (CNN), a type of deep-learning model that is primarily designed to classify images, and a replica of the one used in the study he had found. The process, says Kang, boiled down to “trying to replicate what they’ve done and creating a very simplistic model that is just taking the image, getting the diagnosis and training the model based on how well it can predict that diagnosis.” This model became a baseline against which he measured new and improved versions.

In his next prototype, Kang decided to add ADHD to his model. A diagnostic tool, he says, should be able to differentiate between various disorders instead of simply detecting whether a person has one. “Distinguishing between neurotypical individuals and those with autism is not very difficult, and existing studies have already achieved close to 100 percent accuracy,” he says. Identifying distinct disorders is a much harder task and one that is clinically important.

Kang also employed a few more advanced computational techniques to improve the model’s accuracy and effectiveness. For example, he used ensemble learning, a technique in which different models are given the same task. “You feed them the same retinal image and ask them to predict autism or ADHD, and then you take their predictions and combine them,” says Kang. He then calculates an average from their prediction. Using multiple models and a voting approach, says Kang, means that the results are more reliable. “It tends to be more accurate, and performance can improve,” he explains.

Meet Edward Kang, 2nd Place Winner of the 2026 Regeneron STS

Watch on

How retinal development is different in neurodivergent people

Since late 2024, the inventor has been working to understand the underlying biological mechanisms and foundations that cause retinal differences in people with autism and those with ADHD. “I really began working more on the cell biology side,” he says, “creating an in-vitro or cell-based model of autism and studying what kinds of genes may be involved in why autism patients have retinal differences that can be detected to begin with.”

Kang used gradient-weighted class activation mapping, or GradCAM, an explainable A.I. technique that identifies the specific regions of an image that are most useful to the model in making a prediction. The technique allowed him to explore the inner workings of the CNN and to identify which region of the initial input image the model took into consideration to complete its task. “In this case, that would mean which part of the retina was important for making a diagnosis of autism and ADHD,” Kang explains.

Researchers have previously identified many retinal features that differ, on average, in people with autism or ADHD. Specialized tools, like optical coherence tomography scans, can detect differences in the length, thickness and depth of the macula, retinal nerve fiber layers and other regions. The challenge is that these differences are very subtle and overlap heavily with the normal range seen in neurotypical individuals. So a clinician alone cannot look at a retinal image and diagnose autism or ADHD.

Computer models like RetinaMind can simultaneously detect and combine extremely subtle retinal patterns that are far too complex for humans to recognize, which makes them powerful enough to make diagnoses.

In his research, Kang identified a dozen potential candidate genes linking autism and retinal development. “One potentially interesting gene I identified is ABCA4, which encodes a protein responsible for detoxifying the retina,” the young scientist shares. “My retinal cell autism model showed less ABCA4 expression compared to the control. This suggests that autistic patients may have less of this detoxifying protein, potentially leading to increased retinal toxicity and degradation, which could explain some of the observed retinal differences.” He hopes that the list of genes he has identified can help answer the complex question of why retinal development differs in people with neurodevelopmental disorders.

An award-winning invention

The somewhat counterintuitive goal of RetinaMind is to use a retinal image to predict something unrelated to the eye. Once the A.I. tool has the image, it will analyze it and share the result. It breaks down percentages of its confidence that a retinal image indicates that a patient is neurotypical or has autism or ADHD. “The diagnosis with the highest confidence becomes the official diagnosis of the model,” Kang says. To support a diagnosis, the model produces a heat map visualization of the retinal image, highlighting in red key parts that led to the diagnosis. RetinaMind has an accuracy rate of about 89 percent.

Kang’s invention won second place and an award of $175,000 at the 2026 Regeneron Science Talent Search, the oldest and most prestigious science, technology, engineering and math competition for high school students in the United States. The competition recognizes the most promising students developing ideas to solve urgent global challenges.

“Edward’s project stood out for combining A.I. with lab-based biology, which gave it both computational sophistication and biological depth,” says Maya Ajmera, president and CEO of Society for Science, a nonprofit dedicated to expanding scientific literacy, which hosts the competition. “He focused on real-world challenges—on autism and ADHD.” At a time when getting a diagnosis can take months or even years, notes Ajmera, early screening could make a major difference for a lot of families.

“He didn’t just build a model,” she adds. “He also explored the underlying gene changes, which strengthened the scientific rigor and helped explain why the patterns might exist.”

The future of RetinaMind

Lipkin is excited at the potential of a retinal image to lead to earlier diagnosis, but he is quick to caution that autism and ADHD are developmental and behavioral conditions rooted in the brain with much overlap between them and other conditions. “Any retinal differences identified may not be specific for these conditions, but instead of some brain-based neurologic condition generally,” he says.

Kang agrees with Lipkin’s concerns.

“Right now, my model just makes a blanket diagnosis of either autism spectrum disorder or attention deficit hyperactivity disorder,” says Kang. “But within these kinds of disorders, it’s a very wide spectrum of different kinds of conditions.”

While he’s happy with RetinaMind as a proof of concept, he’s already thinking of next steps to train the model to distinguish between mild, moderate and severe autism.

“The more specific information we can get out of the model,” he explains, “the more effective it is in terms of guiding treatment and making sure that the child is getting the right amount of support that they need.”

Kang adds, “I think that can be something powerful for the future.”

Ramsha Waseem | Read More

Ramsha Waseem is a freelance writer. Her work has appeared in BBC Travel, Dwell, Atlas Obscura and other media outlets.