ETRI Develops AI for Early Autism Spectrum Disorder Screening in Children

Researchers at the Electronics and Telecommunications Research Institute (ETRI) have developed a groundbreaking artificial intelligence (AI) technique designed for the early screening of autism spectrum disorder (ASD) in infants and toddlers. This new technology, known as "social interaction recognition AI," aims to identify early signs of ASD more efficiently, thus expanding opportunities for early screening and intervention, which are crucial for improving developmental outcomes in affected children. ASD is a neurodevelopmental disorder characterized by impaired social communication skills and restricted, repetitive behaviors. Early detection and intervention can significantly enhance a child's quality of life, but current diagnostic methods often face significant challenges, including a scarcity of experts, limited social awareness, and resource constraints. As a result, the average time from symptom detection to diagnosis can range from two to six years, despite the fact that key symptoms may appear as early as 12 to 24 months of age. To address these issues, ETRI researchers partnered with Professor Yoo Hee-Jung's team from the Department of Psychiatry at Seoul National University Bundang Hospital. They analyzed data from 3,531 cases of infants and toddlers aged 42 months or younger, focusing on the sensitivity of various screening indices for ASD. Based on this analysis, the team created a scenario for observing social interactions that can be used to apply AI technology effectively. One of the key components of this scenario is the world's first "social interaction-inducing content" designed to elicit and observe a variety of social responses in young children. These responses include showing interest in objects, reacting to their name, imitating behaviors, making pointing gestures, and maintaining eye contact. The AI technology then uses cameras to capture these interactions and performs a comprehensive analysis, including personal attribute recognition, emotional state prediction, gaze tracking, response to name detection, and identification of pointing and repetitive behaviors. The living lab for ASD screening was established at the Korea Institute of Robotics and Technology Convergence (KIRO) Seoul Center in 2020, where the researchers have conducted observational tests and collected data over the past five years. This prolonged period of study and data collection has allowed the technology to evolve and become more sophisticated, leading to the creation of a multidisciplinary, convergence-based AI system that overcomes the limitations of traditional screening tools. By enabling more objective and quantitative evaluations, the AI technology reduces barriers to mental health services, making it practical for use in preschools, childcare centers, and even at home. This accessibility is crucial, as it can significantly shorten the time between symptom detection and diagnosis, thereby facilitating early intervention and improving outcomes for children with ASD. Moreover, the technology holds the potential to raise social awareness of ASD, encouraging more preventive testing and early intervention. Dr. Yoo Jang-Hee, a principal researcher at ETRI's Social Robotics Research Section, expressed optimism about the technology's impact. "We hope that this will help shorten the time between symptom detection and diagnosis, along with changing societal perceptions of autism. While it is important for our research to tackle complex problems, we are equally excited about its broader contributions to addressing critical issues like autism." Industry insiders are lauding the development as a significant breakthrough in the field of ASD screening. The integration of advanced AI technologies with clinical insights promises to revolutionize the diagnostic process, making it faster, more accurate, and more accessible to those who need it. ETRI, a leading South Korean research institute, has a strong track record in developing innovative technologies, and this project further cements its reputation in the global AI and healthtech landscape. Seoul National University Bundang Hospital, a renowned institution in the healthcare sector, has played a crucial role in validating the effectiveness of the AI algorithm through extensive clinical testing. Together, ETRI and SNUBH demonstrate how collaboration between technology and healthcare professionals can lead to transformative solutions for major health challenges. The successful implementation of this AI-driven screening tool is expected to have far-reaching implications for the early detection and management of ASD, potentially improving the lives of thousands of children and their families.