In an innovative advancement in autism screening, a research group in Dubai has created a pioneering artificial intelligence (AI) tool that harnesses eye-tracking technology alongside developmental data. This comprehensive approach aims to facilitate early identification of children who might require specialized assessments for autism spectrum disorder (ASD).

AI Technology and Eye-Tracking Integration

Led by researchers from the University of Dubai, this AI-powered screening tool aims to deliver an early risk assessment for autism. By analyzing eye movements in conjunction with developmental milestone information, the system offers valuable insights into potential early indicators of ASD. Through this method, the researchers hope to pinpoint children who may need additional evaluation sooner, thereby addressing delays that often accompany access to specialist services.

During the screening, children engage with brief videos featuring social scenarios and geometric shapes. While the children watch, the eye-tracking technology meticulously records their gaze patterns, fixation durations, and visual scanning behaviors. Parents simultaneously complete a questionnaire about their child’s developmental milestones. By collating this data, the AI model generates an early risk indicator that helps prioritize those in need of further assessment.

Efficiency and Accuracy of the New Screening Method

One noteworthy aspect of this innovative tool is its ability to analyze raw eye-tracking data directly. Unlike traditional AI systems that rely on converting visual data into images, this technology simplifies the process, allowing it to function effectively on everyday devices like tablets and smartphones. The screening process itself lasts only around two to three minutes and is completely non-invasive, adding to its practicality for families.

Initial benchmark tests revealed an impressive accuracy rate of approximately 96%. However, the researchers clarified that while the technology holds promise in facilitating early intervention, it is not intended to replace formal diagnostic methods. Ongoing clinical validation studies across diverse healthcare settings will help determine its effectiveness and reliability.

Promoting Early Intervention for Autism

Early identification of autism can significantly impact children’s ability to access timely intervention and support. Most experts agree that signs of autism can often be recognized within the first two years of life, yet many children are not formally diagnosed until around the age of four. By bridging this gap, the newly developed system aims to ensure that children receive the support they need as early as possible.

In contrast to traditional screening methods that largely rely on parent-completed questionnaires, this new system integrates objective measurements of eye movements with developmental information. Through this pioneering approach, researchers are examining whether AI can augment existing practices in autism screening, ultimately broadening access to early intervention services.

The project is financed through the Dubai Research, Development and Innovation Grant Initiative under the Dubai Future Foundation, and it’s a collaboration among the University of Dubai’s College of Engineering and Information Technology, Emirates Health Services, and Al Amal Psychiatric Hospital. Additionally, support extends to research teams from Australian institutions, further enriching the project’s development. As clinical validation progresses, the researchers aspire to make this technology available in clinics and homes, expanding its reach and impact on early autism screening.