Two new studies suggest that tracking toddlers' eye movements while they watch videos of social interactions could help diagnose autism years earlier than current methods. The research, published in JAMA and JAMA Network Open, found that measuring where children focus their gaze can serve as an objective biomarker for early signs of autism.
In the first study, researchers at Children's Healthcare of Atlanta monitored over 1,000 children aged one to two-and-a-half who had autism symptoms or genetic risk factors. The children watched 14 short videos of social interactions while specialized cameras tracked their eye movements 120 times per second. A computer program analysed the data and detected autism with 86% accuracy compared to clinical diagnosis.
The second study evaluated the same eye-tracking method and found it was 78% accurate at diagnosing the condition. Currently, the average age of autism diagnosis in the US is around five, and in the UK it is over five, with families often waiting over a year for a diagnosis. The researchers believe this technique could enable diagnosis between just over one and two-and-a-half years old.
Lead author Professor Warren Jones from Emory University said the results show that how young children look at social information can be an effective biomarker for early autism signs. Co-author Dr Ami Klin added that children who currently face years of waiting could be diagnosed much earlier, allowing for earlier intervention.
