Bench to Bedside | November 2016

Smart Robotic Toy Gym Could Identify Early Signs of Babies’ Developmental Delay

The wriggling supine baby reached up toward the jungle-themed play gym that arced above her and grabbed a dangling elephant by the trunk. As she squeezed it, the elephant squeaked. She smiled, released, and grabbed the elephant again, this time joining in with a happy squeak of her own.

“It was beautiful to watch her discover the relationship between action and noise and elicit it, and that was the kind of thing we want to test,” said Michelle Johnson, PhD, an assistant professor of physical medicine and rehabilitation at the Perelman School of Medicine at the University of Pennsylvania and in Bioengineering at the School of Engineering and Applied Sciences (SEAS). “If, in comparison to an age-matched infant who we suspect might experience delays in development, does that child even discover the elephant toy?”

Such comparisons between healthy babies playing in the gym and babies with suspected motor delays are all part of the design in this play environment, a high-tech setup built in Dr. Johnson’s Rehabilitation Robotics Lab. Dr. Johnson is collaborating with Laura Prosser, PT, PhD, a physical therapist and rehabilitation scientist at Children’s Hospital of Philadelphia and assistant professor of Pediatrics at the Perelman School of Medicine at Penn, in a research effort to help identify babies at risk for motor delays much earlier than is currently possible.

Babies who are born extremely premature or with conditions that affect their brain development are at elevated risk of motor, cognitive, and other types of delays as they grow. Some babies benefit from early intervention services to help make up for delays.

“The problem is being able to correctly predict which babies will have delays and which ones won’t,” Dr. Prosser said.

Typically, when evaluating an infant who is at risk for motor delays, clinicians like Dr. Prosser observe babies moving and monitor their progress toward major milestones such as sitting up, crawling, and walking. They use standardized tests to determine a clinical score that compares their development to the typical development of infants their age. But no quantitative tools exist to compare detailed information about how babies learn to move and control their movements, such as exactly how babies hold toys when starting to reach, or how small movements of their trunk support more mature kicking patterns. This type of information is needed to identify babies with likely motor delays months before they miss the obvious milestones, when the impact of early rehabilitation might be greater.

That is what led Dr. Prosser and Dr. Johnson to stuff a squeaky elephant toy (and a monkey and a lion) with grasp sensors and inertial motion sensors consisting of a gyroscope and accelerometer.

“We conceived together this concept of a smart gym environment with systems we could build in the lab for earlier diagnosis of infants,” said Dr. Johnson, whose past work has involved development of robotic and other technological rehabilitation tools for adults recovering from stroke and adults with cerebral palsy. “We want to determine whether our engineering rehab tools could support the process of detecting motor deficits in infants before any of the clinical tools that are out there can do that, and we are especially excited because if we are successful, then we could get infants who need rehab the help they need earlier.”

They are piloting their creation and study of the SmarToyGym with funding from the Eunice Kennedy Shriver Institute for Child Health and Human Development. The project is a collaborative effort whose other key players include Sam Pierce, PT, PhD, from Widener University; Daniel Bogen, PhD, a professor of Bioengineering in Penn SEAS; Frances Shofer, PhD, from the department of Emergency Medicine at the Perelman School of Medicine; and Helen Loeb, PhD, a bioengineer in the Center for Injury Research and Prevention at CHOP.

During the first year of the two-year project, they built several prototype versions of the gym, which entailed design challenges distinct from the robotic building efforts in Dr. Johnson’s previous efforts. Ordinarily, she noted, people would develop a more constrained experimental setup to test a limited type of skill or motion so there would be few variables and simpler measurements. Instead, the gym allows babies to play naturally and collects vast amounts and types of data from built-in sensors and video monitoring.

As a result, the data analysis of this gym will be complex, integrating multiple types of data from multiple toys, cameras, and the mat itself, including movements and compression of the toys, the baby’s rolling or crawling motion, and other natural interactions between the baby and the gym.

“Our vision was, we wanted something that ultimately could be placed in a home, in a doctor’s office, or in a day care, a product that would in an unobtrusive way capture this information while the babies play naturally,” Dr. Johnson said. “That means we need to use more modern data analysis tools to try to put our arms around the complicated data we get about babies’ movements.”

Now in its second year, the project has moved to a pilot study of the potential to use the gym and its data to detect meaningful differences in motor skills between 12 typically developing babies and 12 babies who are at risk for motor delays.

“What we hope to learn from this first round of clinical testing is, can these types of sensors and this type of instrumentation identify that babies who are at risk move differently from babies who are typically developing?” Dr. Prosser said. “If that’s the case, then the next steps will be refining the algorithms to quantify those measured differences, and display the metrics in a user friendly interface so parents can see immediately how their baby is moving.”

If the measurement aspect of the study is successful, the researchers hope to receive funding to expand their testing to improve the measurements. They also want to expand future testing into community settings such as homes and daycares, and to test enough babies to define and differentiate both typical and impaired abilities by age and diagnosis, and to consider expanding beyond their initial measurements of motor development to also measure cognitive development.

A long-term vision includes refining the gym so that it is not just a measurement tool, but also a rehabilitation tool that can be individualized to each baby’s specific needs — placing a toy that lights up above her weaker right arm to encourage her to reach in that direction, or placing that squeaky elephant near her left hand to help her to build her grip strength, for example.

Making the toy natural, unobtrusive, and easy for babies to use, but also inexpensive, further appealed to Dr. Johnson because of her lab’s focus on developing affordable rehabilitation tools for developing countries. Many developing countries lack the clinical tools and expertise to diagnose motor delays in young babies, but a toy that could both detect problems and help fix them through natural play might offer benefits to more people in more places.

The researchers are still actively recruiting babies for the pilot tests of the gym, including both typically developing and at-risk babies age 3 to 11 months. Clinicians, parents, childcare professionals, or others who are interested in learning more can contact Dr. Prosser or Dr. Johnson for details.


www.research.chop.edu

Produced by The Children’s Hospital of Philadelphia Research Institute.

© 2017 by The Children’s Hospital of Philadelphia, All Rights Reserved.