Early Screening with Infant Autism AI: 6-Minute Video
- Researchers in South Korea have developed an artificial intelligence (AI) system designed to expedite the detection of autism spectrum disorder (ASD) in young children.
- The Korea Electronics and telecommunications Research Institute (ETRI) announced the progress of "social interaction AI" technology. This system analyzes a child's responses while they watch specific short video...
- Autism spectrum disorders are neurological conditions characterized by challenges in social communication, along with restricted and repetitive behaviors.
AI System Aims to Speed autism Detection in Young Children
Table of Contents
- AI System Aims to Speed autism Detection in Young Children
- AI System Aims to Speed Autism Detection in Young Children: Q&A
- What is this new AI technology for autism detection?
- How does this AI system work?
- Why is early detection of autism crucial?
- What are the challenges in detecting autism early?
- How was the AI system developed?
- What kind of recognition has the AI system received?
- What are the other AI models used for autism detection?
- What is the future impact of AI in autism screening?
- Where can I find more information?
- Summary of Key AI Technologies for autism Detection
Published: 2025-04-29
Researchers in South Korea have developed an artificial intelligence (AI) system designed to expedite the detection of autism spectrum disorder (ASD) in young children.
The Korea Electronics and telecommunications Research Institute (ETRI) announced the progress of “social interaction AI” technology. This system analyzes a child’s responses while they watch specific short video content, with assessments completed in under six minutes. Researchers also created content designed to elicit social responses from children.
The Challenge of Early ASD detection
Autism spectrum disorders are neurological conditions characterized by challenges in social communication, along with restricted and repetitive behaviors. Early detection and intervention are critical for improving long-term outcomes. Though, a shortage of specialists, societal stigmas, and logistical constraints frequently enough delay diagnosis by two to six years after symptoms first appear.
Data-Driven Development
Professor Yoo Hee-jung, a mental health medicine specialist at Seoul National University bundang Hospital, analyzed data from 3,531 infants and young children under 42 months old. This analysis led to the development of objective screening indicators and observation scenarios suitable for AI request.
Customized Content for Assessment
Based on this research, customized content was created to encourage social interactions crucial for ASD screening. These interactions include reactions, imitation, pointing, and eye-tracking when presented with engaging objects or when their name is called.
How the System Works
The AI technology analyzes video of a child watching the content, captured via camera, to assess key indicators.
Recognition and Future Impact
The technology received the ‘National R&D Excellence in 2024’ award. The innovation has resulted in more than 50 patent applications, both domestic and international, and over 18 publications in international journals (SCIE).
AI’s Growing Role in Pediatric Screening
The development aligns with a growing trend of using AI to accelerate the early detection of developmental conditions like ASD, perhaps allowing children to begin therapy sooner and significantly improving their long-term outcomes. (See: Screening for Autism Is AI Proving Ground in Pediatrics)
Othre AI models are also showing promise. One model, AutMedAI, can predict autism in children under two with nearly 80% accuracy, according to Neuroscience News. Another system using virtual reality and AI boasts over 85% accuracy in detecting ASD, potentially outperforming customary methods. (See: AI and VR System Detects Autism in Kids with Over 85% Accuracy)
AI System Aims to Speed Autism Detection in Young Children: Q&A
What is this new AI technology for autism detection?
This AI system, developed in South Korea, is designed to expedite the detection of autism spectrum disorder (ASD) in young children. The system analyzes a child’s responses to specific short video content to assess key indicators of autism. The goal is to enable earlier detection and intervention, potentially leading to better long-term outcomes.
How does this AI system work?
The AI technology analyzes video of a child watching specific short video content. This video is captured via camera to assess key indicators of potential autistic traits. The content is designed to elicit social responses from children, such as reactions, imitation, pointing, and eye-tracking. Assessments can be completed in under six minutes.
Why is early detection of autism crucial?
Early detection and intervention are crucial for improving long-term outcomes for children with autism spectrum disorder (ASD). However, delayed diagnosis, often by two to six years after symptoms appear, is a common challenge.
What are the challenges in detecting autism early?
Several factors contribute to delays in autism diagnosis, including:
Shortage of specialists
Societal stigmas
Logistical constraints
How was the AI system developed?
The development of this AI system involved:
Data Analysis: Professor Yoo Hee-jung from Seoul National University Bundang Hospital analyzed data from 3,531 infants and young children under 42 months old.
objective Screening Indicators: This analysis led to the development of objective screening indicators and observation scenarios suitable for AI assessment.
Customized content: Content was created to encourage social interactions (reactions, imitation, pointing, eye-tracking) crucial for ASD screening.
What kind of recognition has the AI system received?
the technology received the “National R&D Excellence in 2024″ award. It has also resulted in:
More than 50 patent applications (domestic and international)
over 18 publications in international journals (SCIE)
What are the other AI models used for autism detection?
Several other AI models are showing promise in the early detection of autism:
AutMedAI: Can predict autism in children under two with nearly 80% accuracy.
VR and AI System: Uses virtual reality and AI and boasts over 85% accuracy in detecting ASD.
What is the future impact of AI in autism screening?
AI has the potential to accelerate the early detection of developmental conditions like ASD. This could allow children to begin therapy sooner and significantly improve their long-term outcomes.
Where can I find more information?
Screening for Autism Is AI Proving Ground in Pediatrics:
Neuroscience News:
Summary of Key AI Technologies for autism Detection
| AI Model | Accuracy | Features |
| —————— | —————— | ——————————————————————————————————- |
| This AI system | N/A | Analyzes video responses to short video content; focuses on social interaction indicators. |
| autmedai | Nearly 80% (Under 2) | Predicts autism in children under two years old. |
| VR and AI System | Over 85% | Uses virtual reality and AI for autism detection. |