Bench to Bedside (B2B)
Pictured from left: Alicia Du (B2B organizer), Pebble AI team members, Coy Steel, Jaden Johnson, Amelia Nelson, Rylie Gagne, and Mike Woodruff (judge) (photo provided by University of Utah Center for Medical Innovation)
On April 14, the University of Utah Center for Medical Innovation hosted its annual Bench to Bedside (B2B) competition, uniting 140 student innovators from over 30 academic majors and disciplines. A total of 31 teams each aimed to address critical issues in healthcare through novel technologies and collaboration. Taking home the $7,500 cash prize, sponsored by the Patient Safety Technology Challenge, was Pebble AI, a groundbreaking solution applying machine learning and mobile video analysis to neonatal care.
Pebble AI is tackling a pressing challenge in neonatal medicine: accurately identifying hypoxic-ischemic encephalopathy (HIE), a condition that accounts for 25% of newborn deaths globally. The team notes that traditionally, the current method of diagnosing HIE relies heavily on subjective interpretation by neonatologists using tools like the SARNAT scoring system to determine whether a newborn requires treatment. This subjectivity can lead to misdiagnosis, delayed treatment, and even irreversible brain damage or death. Pebble AI aims to transform this process by introducing an objective, accessible, data-driven method to improve diagnostic accuracy and speed.
“Our objective, data-backed approach moves NICU decisions from guesswork toward precision, effectively shifting the game from 'opinion' to hard data and even harder science,” explained the Pebble AI team. Their AI tool uses mobile cameras and real-time facial video analysis to identify signs of HIE, providing clinicians with immediate, non-invasive assessments that can be used in any hospital setting, including resource-limited areas.
The Pebble AI team’s progress is nothing short of impressive. Thus far, the team has completed numerous milestones, including but not limited to designing a study and assessment, securing IRB approval, defining a regulatory path, demonstrating usability in NICUs, and collecting 1,000 videos from 400 newborns. Additionally, they developed a clinical-grade tracking solution and are expanding their data sources while refining their market strategy.
Dr. Mike Woodruff, who judged the Patient Safety Technology Challenge, praised the team for their clinical insight and practical impact. “They were one of the few teams focused on avoiding misdiagnosis,” he said. “We loved how they were leveraging AI to complement human knowledge and diagnostic skills to prevent harm. A very scalable solution that makes perinatal expertise available to remote locations and should improve outcomes for newborns.”
B2B organizer Amanda LeMatty emphasized how the involvement of the Patient Safety Technology Challenge mentors elevated the event. “Their guidance helped student teams better understand the real-world impact of their innovations on patient safety,” she said. “We’re grateful for their support in shaping the next generation of healthcare innovators!”
Jose Emiliano Urciaga Juarez, a Pebble AI team member, shared his personal motivation. “A family member received delayed care in a hospital, and it made me realize how technology can help doctors act faster and save lives,” he said. As Pebble AI continues to refine its model, pursue clinical testing, and navigate the regulatory landscape, it stands as a promising example of how student innovation, guided by real-world mentorship and driven by a passion for impact, can shape the future of patient safety. The team hopes to bring Pebble AI to NICUs around the world.
