Researchers develop jumpsuit with sensors that can track an infant’s motor skills

Key takeaways:

• Researchers developed a wearable suit with sensors and an AI-based algorithm to track gross motor milestones in infants.
• Their measurements were comparable to parental report and internationally established benchmarks.

Researchers developed a wearable jumpsuit with sensors that can objectively measure an infant’s gross motor development and, with the help of AI, identify milestones with more than 90% accuracy, according to findings reported in Pediatrics.

The Motor Assessment of Infants with a Jumpsuit (MAIJU) technology would allow clinicians to evaluate infants’ gross motor development in their home environment, rather than relying on office visits or parental reports, the researchers said.

“The presently used MAIJU wearable adds a much-needed accuracy to the conventional approach of assessing gross motor development with reaching of milestones,” Manu Airaksinen, DSc, senior scientist at the Baby Brain Activity (BABA) Center in the department of physiology at the University of Helsinki, told Healio. “It is well known to clinicians that the doctor’s appointment is not optimal for assessing the true performance of the infant, whereas the parents may not always find it easy to assess their own child’s behavior in objective terms.”

The researchers input data from the MAIJU suits into an AI-based algorithm that was trained to identify which movement milestones babies have reached. They used WHO guidelines to determine the benchmarks for each milestone.

They developed the algorithm with data from 97 infants (46% boys; age range, 4 to 19 months) and validated it with 37 infants (73% boys; age range, 4 to 22 months). They instructed parents to put the suit on their infant and let them play for at least 1 hour. Infants wore the suit between one and eight times over the course of the study. Overall, the researchers gathered 620 measurements from the MAIJU.

Airaksinen and colleagues compared their algorithm with parent reports and a WHO reference study from 2004 that relied on human experts.

The researchers found that the algorithm identified gross motor movement milestones at the same time as parental reports. Compared with WHO data, the infants met milestones during the same age windows, except for sitting without support, which the infants in their study reached later. The authors noted this was likely because they measured spontaneous movement, whereas the WHO study set infants in a sitting position first.

Compared with parent reports, the MAIJU identified most milestones with more than 90% accuracy, the authors wrote. They also found a strong association between parent reports of time spent in each posture and actual measured time.

“The biggest surprise was the general finding: that an infant’s motor performance can be assessed reliably with the MAIJU method without seeing the infant, just by measuring their normal behavior at home,” Airaksinen said. “It was also surprising to see that the assessment compares so well with human experts and the parental assessments, and that the objective MAIJU-based measures can actually track infants’ skills over infancy.”

Airaksinen said the researchers plan to use the measurements from the MAIJU to develop “motor growth charts” for global use. He said his team is offering the MAIJU for other researchers for use in pediatric studies without commercial ties.

“We believe that it is important to allow other people to do their independent MAIJU-based studies, which would eventually support acceptance of this conceptually new high-tech methodology as a mainstream solution in health care,” Airaksinen said.

For more information:

Manu Airaksinen, DSc, can be reached at manu.airaksinen@helsinki.fi.

References:

• Airaksinen M, et al. Pediatrics. 2025;doi:10.1542/peds.2024-068647.
• Wijnhoven TMA, et al. Food Nutr Bull. 2004;doi:10.1177/15648265040251S106.