Automated Video Tracking of Autistic Children's Movement During Caregiver-Child Interaction: An Exploratory Study.
A quick video and a laptop can give you valid autism severity scores in minutes.
01Research in Context
What this study did
Researchers filmed 30 autistic kids playing with their moms for 10 minutes.
A computer tracked every tiny move the child made. No humans coded anything.
They wanted to see if simple motion numbers could stand in for long clinical tests.
What they found
Kids who moved faster and longer were younger, scored lower on IQ tests, and showed more autism traits.
When kids stayed closer to mom, they entered shared play more often.
The computer’s motion scores lined up with standard severity ratings.
How this fits with other research
Jabbar et al. (2026) also used computer vision but looked for hand-flapping and spinning. Both studies prove cameras can spot autism signs without a person watching.
Mastrogiuseppe et al. (2015) counted gestures by hand and found fewer gestures in autism. Greenlee et al. (2024) shows machines can now do that counting for us.
Cholemkery et al. (2016) strapped actigraphy sensors on kids with ADHD to catch fidgeting. Greenlee et al. (2024) gets the same motion data from plain video, no extra gear needed.
Why it matters
You can run a 10-minute play clip through free tracking software and get numbers that match hours of testing. Use it to flag kids who need deeper assessment or to track if your treatment is calming excess movement and boosting shared play.
Want CEUs on This Topic?
The ABA Clubhouse has 60+ free CEUs — live every Wednesday. Ethics, supervision & clinical topics.
Join Free →Film your next client during free play, run the clip through open-source pose software, and compare the motion heat-map to their last ADOS score.
02At a glance
03Original abstract
Objective, quantitative measures of caregiver-child interaction during play are needed to complement caregiver or examiner ratings for clinical assessment and tracking intervention responses. In this exploratory study, we examined the feasibility of using automated video tracking, Noldus EthoVision XT, to measure 159 2-to-7-year-old autistic children's patterns of movement during play-based, caregiver-child interactions and examined their associations with standard clinical measures and human observational coding of caregiver-child joint engagement. Results revealed that autistic children who exhibited higher durations and velocity of movement were, on average, younger, had lower cognitive abilities, greater autism-related features, spent less time attending to the caregiver, and showed lower levels of joint engagement. After adjusting for age and nonverbal cognitive abilities, we found that children who remained in close proximity to their caregiver were more likely to engage in joint engagement that required support from the caregiver. These findings suggest that video tracking offers promise as a scalable, quantitative, and relevant measure of autism-related behaviors.
Journal of autism and developmental disorders, 2024 · doi:10.1093/ptj/pzz114