Related Experiment Video
Updated: Aug 1, 2025
Gaze in Action: Head-mounted Eye Tracking of Children's Dynamic Visual Attention During Naturalistic Behavior
Published on: November 14, 2018
Computer Vision Analysis of Caregiver-Child Interactions in Children with Neurodevelopmental Disorders: A Preliminary
Dmitry Yu Isaev1, Maura Sabatos-DeVito2, J Matias Di Martino3
1Department of Biomedical Engineering, Duke University, Durham, NC, USA. dmitry.isaev@duke.edu.
Insights
Caregiver responsiveness during play, analyzed by computer vision, showed unexpected links to child development. Higher caregiver responsiveness correlated with less developed language and social skills in children with autism and ADHD.
Area of Science:
- Developmental Psychology
- Child Psychiatry
- Computer Vision in Healthcare
Background:
- Caregiver-child interactions are crucial for child development.
- Understanding these dynamics is key for diagnosing and treating neurodevelopmental disorders.
- Automated analysis offers objective measures of interaction patterns.
Purpose of the Study:
- To analyze caregiver-child interactions using computer vision.
- To investigate 'reaching to a toy' as a measure of engagement.
- To explore associations between interaction patterns and developmental outcomes in children with autism, ADHD, and controls.
Main Methods:
- Computer vision analysis of free play interactions.
- Micro-analytic coding of 'reaching to a toy' behavior.
- Dyadic analysis to identify interaction clusters based on caregiver responsiveness.
Main Results:
- Two distinct interaction patterns were identified.
- Higher caregiver responsiveness was linked to poorer child language, communication, and socialization skills.
- These interaction patterns were not specific to diagnostic groups (autism, ADHD, or neurotypical).
Conclusions:
- Computer vision can objectively quantify caregiver responsiveness.
- Unexpectedly, high responsiveness may indicate developmental challenges.
- Automated analysis shows potential for clinical assessment and monitoring in trials.
Abstract:
We report preliminary results of computer vision analysis of caregiver-child interactions during free play with children diagnosed with autism (N = 29, 41-91 months), attention-deficit/hyperactivity disorder (ADHD, N = 22, 48-100 months), or combined autism + ADHD (N = 20, 56-98 months), and neurotypical children (NT, N = 7, 55-95 months). We conducted micro-analytic analysis of 'reaching to a toy,' as a proxy for initiating or responding to a toy play bout. Dyadic analysis revealed two clusters of interaction patterns, which differed in frequency of 'reaching to a toy' and caregivers' contingent responding to the child's reach for a toy by also reaching for a toy. Children in dyads with higher caregiver responsiveness had less developed language, communication, and socialization skills. Clusters were not associated with diagnostic groups. These results hold promise for automated methods of characterizing caregiver responsiveness in dyadic interactions for assessment and outcome monitoring in clinical trials.
