Study by Feng Xie, PhD, Published in High-Impact Journal

A study by Assistant Professor Feng Xie, PhD, in the Division of Computational Health Sciences was recently published in The Lancet Digital Health.

Dr. Xie's research focuses on using artificial intelligence to better understand doctors' and nurses' written notes about newborn babies in the hospital. By analyzing these routine clinical notes, the team of researchers aims to identify early warning signs of serious neonatal conditions more quickly and accurately. Their goal is to support clinicians with timely, data-driven insights during critical early stages of care.

In doing this work, the team found that unstructured clinical notes contain important predictive information that is often missed by traditional risk models. However, when language models are specifically adapted to neonatal care and evaluated under real-world clinical conditions, they can reliably predict neonatal morbidities and generalize across multiple hospitals.

The impact this research could have on neonatal care is significant. 

"This work could help clinicians identify high-risk newborns earlier, enabling closer monitoring or earlier intervention when it matters most. Over time, such tools may improve outcomes for vulnerable infants while supporting more consistent and equitable care across health systems," said Dr. Xie.

This study is part of a broader research program focused on building trustworthy, interpretable, and robust AI for medical applications (Dr. Xie's lab's current core research focus). The team's next steps include expanding multi-center collaborations, integrating language models with other clinical data sources, and evaluating their use in different clinical settings. 

Additionally, they plan to apply these advanced methods to a wide range of clinical questions, including surgical care, emergency and critical care, and chronic disease management

"We are grateful to our multidisciplinary team and clinical partners and welcome collaboration with others interested in advancing responsible AI for the vulnerable patient populations," said Dr. Xie.

Congratulations, Dr. Xie!