According to new research, deep learning and artificial intelligence models can be used to detect autism and determine the severity of the condition. All the AI is a photo of the person's retina.
See also: "Brainoware": Human brain cells enhanced with AI
Previous studies have linked changes in the retina to changes in brain structures and, in turn, to Autism Spectrum Disorder (ASD). The evidence suggests that the eye truly is a window to the brain, through the interaction of the central nervous system.
“Individuals with Autism Spectrum Disorder (ASD) exhibit structural changes in the retina that likely reflect alterations in the brain, including abnormalities in the visual pathway through embryological and anatomical connections,” the researchers report in their new article.
“It remains uncertain whether deep learning algorithms can help objectively diagnose autism and assess the severity of symptoms through the analysis of retinal photographs.“
In this study, a team from Yonsei University of Medicine in the Republic of Korea wanted to see if artificial intelligence could detect autism in retinal reflex patterns. Initially, the model was trained with images where the AI was “informed” whether the person had autism or not.
See also: COVID-19 patients showed long-term changes in the brain
Subsequently, the artificial intelligence was asked to analyze the retinas of 958 children and adolescents with an average age of 7.8 years, with half of them diagnosed with autism. It managed to achieve a perfect score in recognizing those who had autism and those who did not.
The AI wasn't as accurate at predicting symptom severity from retinal images – it was only between 48% and 66%. However, there's a lot of potential here to help children get a useful assessment at an earlier age, with less waiting.
In this case, the researchers focused data on children and adolescents aged 4 to 18. However, they believe that expanding this approach to even younger children is a possibility worth exploring in future studies.
In a study published last year, researchers managed to link the retina's response to light with cases of Attention Deficit Hyperactivity Disorder (ADHD) and ASD, another example showing how eyes can act as a mirror of a person's brain activity.
See also: Brain implant helps woman regain cognitive function after injury
In every 36 people, it is estimated that one is autistic, based on its occurrence in eight-year-old children. Awareness of this diagnosis from an early age can make a huge difference in how young people perceive themselves and move forward in the world.
The research has been published in JAMA Network Open.
Source: sciencealert
