Automated movement tracking of young autistic children during free play is correlated with clinical features associated with autism

Abstract

Autistic children’s play provides insights into social, communication, and other skills; however, methods for measuring these observations can be labor-intensive and rely on subjective judgment. This study explored whether children’s movement and location during play measured via automated video tracking correlates with clinical features. Movement tracking metrics of 164 autistic children (27–96 months old) during free play were analyzed in relation to standard assessments of cognitive, language, social-communication, and adaptive skills. Children with higher cognitive and language abilities were more likely to spend time in the center toy region and more slowly or never approached the region without toys. Children with higher autism-related features spent less time in the center. Children with lower daily living skills spent more time near the caregiver and those with lower overall adaptive and language skills approached the caregiver more quickly. Over 90% of autistic children representing a range of ages and skills provided analyzable movement data during play and those with higher cognitive, language, and adaptive skills displayed movement tracking patterns that reflect more sustained focus on toy play and independence from the caregiver. Results suggest that automated movement tracking is a promising complementary, objective method for assessing clinical variation during autistic children’s play. Lay Abstract Play-based observations allow researchers to observe autistic children across a wide range of ages and skills. We recorded autistic children playing with toys in the center of a room and at a corner table while a caregiver remained seated off to the side and used video tracking technology to track children’s movement and location. We examined how time children spent in room regions and whether or not they approached each region during play related to their cognitive, social, communication, and adaptive skills to determine if tracking child movement and location can meaningfully demonstrate clinical variation among autistic children representing a range of ages and skills. One significant finding was that autistic children who spent more time in the toy-containing center of the room had higher cognitive and language abilities, whereas those who spent less time in the center had higher levels of autism-related behaviors. In contrast, children who spent more time in the caregiver region had lower daily living skills and those who were quicker to approach the caregiver had lower adaptive behavior and language skills. These findings support the use of movement tracking as a complementary method of measuring clinical differences among autistic children. Furthermore, over 90% of autistic children representing a range of ages and skills in this study provided analyzable play observation data, demonstrating that this method allows autistic children of all levels of support needs to participate in research and demonstrate their social, communication, and attention skills without wearing any devices.