WVU occupational therapy student helps advance AI tools for infant motor development research

Aiden Aiello, a Master of Occupational Therapy (MOT) student, is working alongside a team of faculty researchers to explore how artificial intelligence (AI) can assess head and neck control in infants during tummy time, an essential component of early motor development.

Aiello is collaborating with Ketaki Inamdar, PT, Ph.D., Valeriya Gritsenko, Ph.D. and Breanna Adkins, Ed.D., MOT, OTR/L, on this interdisciplinary project that connects technology and pediatric occupational therapy.

The goal of the project is to see how well AI compares to motion capture systems when assessing infant motor skills development, and also improve its accuracy so it can be used in clinical settings. The team is comparing the AI’s tracking data to the motion capture output to evaluate how closely they align. 

“Currently, systems like motion capture are the gold standard of capturing 3D data; however, these systems are extremely expensive and not present in many clinical spaces,” Aiello said. “As AI emerges as a potential alternative, we want to explore its uses in a clinical setting.”

As part of the project, Aiello helps set up and operate the motion capture system, monitors and comforts the infant participants during the trials and processes the resulting videos by editing out extraneous data. 

The motion capture system uses eight cameras and LED markers placed on key points of the infant’s body, including the head, shoulders and hips, to record 3D movement. The AI program is trained to use the video footage to recognize and track those same body parts by analyzing pixel movement.

Once the footage is cleaned, it is run through the neural network, an AI program with the ability to detect patterns in head and neck movement.

So far, the research has found that while AI shows promise, it’s not yet accurate enough to replace motion capture. “Preliminary results have shown that AI is not yet ready to take the place of the gold standard motion capture, and many kinks and variables will need to be addressed to boost the AI’s accuracy,” Aiello said.

Dr. Inamdar, the primary faculty researcher on the project, explained that even if the research “doesn’t work,” it’s still important to share their findings to help change current standards.

“As AI tools gain momentum in healthcare, it’s essential to understand where they fall short, especially in sensitive areas like infant development,” she said. “Studies like this help us refine the technology for real clinical impact. It’s just as important to share what doesn’t work yet as what does.”

Aiello’s involvement in the project began when Dr. Adkins, a faculty member in the Division of Occupational Therapy, announced the opportunity to collaborate with the Division of Physical Therapy. Already familiar with research from his undergraduate experience in the Research Apprenticeship Program (RAP), Aiello jumped at the chance to contribute to a clinically meaningful project.

“I have loved every minute of all my research opportunities at WVU,” Aiello said. “These opportunities have given me plenty of chances to develop my skills as a researcher and a clinician, as well as to get out and disseminate my research through poster presentation.”

Inamdar emphasized the importance of students participating in research.

“For students, projects like this are also a chance to refine their clinical assessment skills, infant handling skills and communication with families,” Inamdar said. “Having students involved not only strengthens our research but also helps shape the next generation of evidence-informed clinicians.”

For more information on WVU’s occupational therapy program or pathways to this degree, visit https://medicine.wvu.edu/ot/.