|
Getting your Trinity Audio player ready...
|
A new AI machine learning model has been found to improve the detection of lung disease in newborns. The tool, developed by a team of University of Rochester researchers, can better predict bronchopulmonary dysplasia than existing calculators.
BPD is a disorder that typically affects babies who are born early. A preterm baby has underdeveloped lungs, and some might require mechanical ventilation or oxygen therapy for breathing. These therapies can sometimes damage lung tissue.
“We take great effort in the neonatal intensive care unit to prevent lung damage,” says Andrew Dylag M.D., an associate professor in the Neonatology Division of the Department of Pediatrics at URochester. “Despite this, premature infants still develop BPD. There are BPD ‘calculators’ on the internet that can predict the severity of lung disease while the baby is still in the hospital, but they use a very limited set of data.”
A recent update impacted the accuracy of these online calculators, prompting the URochester team to find another way.
“We thought that using more detailed data from the university’s electronic health record would improve disease predictions and allow us to pinpoint vulnerable times when we might be able to intervene to prevent lung disease in newborns,” Dylag says.
URochester’s Clinical and Translational Science Institute supported the project.
“We hoped that CTSI could help us test the hypothesis that machine learning could improve disease predictions in hospitalized premature newborns,” Dylag says. “The 2023 Digital Health Seedling award was exactly the type of funding we needed to develop new collaborations across the university community and kickstart our team’s academic interactions.”
The pilot award, which is up to $25,000, helped the team to add collaborators including
Albert Arendt, Hopeman professor of engineering, Jiebo Luo from the Department of Computer Science, and professor Xing Qiu from the Department of Biostatistics and Computational Biology. Several students were involved in the project as well.
“The neonatology team brought content and clinical expertise to the work, the computer science team developed and tested the models, and the biostatisticians ensured the rigor and testing of the models and algorithms,” Dylag says.
CTSI’s informatics and analytics group also pitched in.
“We initially got involved by helping the study team pull clinical data from eRecord,” says Jack Chang, associate director of research informatics. “Recognizing the project involved a very large patient population and massive data including protected health information, we identified a need for a more secure analytical workspace.”
The resulting model substantially outperformed static approaches for predicting BPD and death among extremely low gestational age newborns. Integrating machine learning methods into clinical applications holds promise for enhancing real-time BPD trajectory mapping, the team’s study published in the May issue of the Journal of Pediatrics states.
“We want to build clinical decision support tools to identify how disease predictions change in real time,” Dylag says. “If we validate our algorithm and can present the disease prediction to the clinical team, we can test guideline implementation for how to manage or treat infants that may reduce BPD severity.”
Smriti Jacob is Rochester Beacon managing editor.
The Beacon welcomes comments and letters from readers who adhere to our comment policy including use of their full, real name. See “Leave a Reply” below to discuss on this post. Comments of a general nature may be submitted to the Letters page by emailing [email protected].