Division researcher Kari Kretch has earned a new federal grant that will use wearable technology and machine learning to measure how infants’ everyday movements are related to optimal motor development.

BY KATHARINE GAMMON

WHEN BABIES MOVE, CRAWL, SIT AND STAND, they are essentially “hitting the gym” — training their little bodies to be able to do those motor skills easily in the future. But researchers haven’t been able to untangle which movement experience lead to future motor skills. That’s because it can be hard to peer into babies’ everyday lives without their living in a lab setting.

Assistant Professor Kari Kretch is hoping new technology will help answer these questions.

Kretch recently received a federal grant — along with colleagues from the University of California-Riverside and the University of Georgia — for a new study that will collect data on 175 infants’ daily activities for a week, to better understand how their patterns of movement change over time.

The data, collected by wearable sensors combined with artificial intelligence technology, will also allow clinicians to understand which patterns lead to later motor skills — critical information that can help understand both typically developing infants as well as those with delayed development.

Unfettered Access, Thanks to Wearable Technology

The researchers will ship the equipment to families all over the United States. It’s novel that they’re able to expand their participant pool to those who are located outside the immediate area, they say. And the equipment itself? “It just looks like a pair of pants,” Kretch says — pants embedded with sensors at the hip and ankle that can determine if a baby is seated in a chair, being held by a parent or moving about freely.

After the wearable technology is returned, the lab will use machine learning to analyze the collected data, identifying body positions and restraint status.

“There are not a lot of methods that can classify infant activity across a whole day or a whole week,” says Kretch, who is both a physical therapist and a developmental psychologist. “The sensors and machine learning allow us to capture that without having to have a human watch and annotate a video and provide more than a small slice of what their day is like.”

After the initial data is collected, the team will survey the families to understand how infants’ opportunities to move relate to caregivers’ beliefs about infant movement and physical activity.

Then, using video conferencing, they will assess the same infants’ motor skills at a later age.

The study’s goal is to discover how motor experiences — such as sitting and standing — are distributed throughout a day or week to predict specific later improvements in skills, such as sitting and standing. The infants will be tested three times — at seven months, nine months and 11 months.

More Effective Treatments

Kretch’s lab, the Learning, Development and Rehabilitation Lab, focuses on early motor development in infants, including those with developmental disabilities like cerebral palsy or Down syndrome.

The study’s next steps include understanding the relationships between everyday movement experience and motor skills in populations of infants with motor impairments.

Ultimately, Kretch hopes this study will improve patient outcomes — especially with early intervention for infants and toddlers.

For example, the study might suggest that several short periods of motor practice is more effective than one longer period each day — or the opposite.

That knowledge could help future clinicians make treatments more effective “We know that the actual intervention is not happening in the 45 minutes that a child is with a therapist every week,” she explains. “It’s happening throughout childrens’ lives. This study will help us understand what type of guidance and interventions we can give that can promote particular experiences and outcomes.”