Researchers from the Republic of Korea have discovered a convenient and accurate way to detect autism, simply by analyzing a photo of a child’s eye. Technologically advanced artificial intelligence models are used to examine the photo and diagnose the disorder.

According to the Irish Star, the new study published in JAMA Network Open builds upon previous research that found links between brain structures and retinal nerve changes. These changes are typical of those with autism spectrum disorder (ASD).

“Individuals with ASD have structural retinal changes that potentially reflect brain alterations, including visual pathway abnormalities,” said the report, written by researchers at the Yonsei University College of Medicine in the Republic of Korea. They added that while the retinal imaging was accurate in diagnosing ASD, it was not as helpful in assessing the severity of the disorder.

The retinas of 958 children and teenagers were assessed by the AI model that had been programmed to find subjects who had autism and those who did not through retinal patterns. Half the studied subjects, whose average age was 7.8, were correctly diagnosed with autism based solely on the images of the eye.

However, the researchers acknowledged that the technology was only between 48% and 66% accurate when it came to assessing the extent of autism through retinal images, says the Irish Star. And they suggested further studies would benefit even younger children, as their study focused only on those between the ages of 4 and 18.

“Retinal alterations in individuals with ASD may manifest even before retinal maturation,” they wrote. Previous studies have linked the retina’s response to light with ASD, as well as attention deficit hyperactivity disorder, or ADHD, says the Ophthalmology Times.

Research from Flinders University and the University of Australia last year found that recordings from the retina could identify distinct signals for both ADHD and ASD. Using an electroretinogram or ERG, researchers found that the electric activity of the retina in response to light stimulus was higher for those with ADHD and lower for those with ASD.

It’s estimated that one out of 36 people suffer from autism, and early detection and intervention can be critical in treating young people.

“The findings of this study suggest that retinal photographs may serve as a viable candidate for an objective method to screen for autism and possibly for symptoms severity,” the Korean researchers wrote.