Machine Learning Predicts Life-Threatening Disease in Infants

Researchers have created an early warning system that uses machine learning to predict necrotizing enterocolitis (NEC), a life-threatening intestinal disease that affects premature infants.

NEC impacts up to 11,000 premature infants in the US annually, researchers noted, and 15 to 30 percent of babies die from NEC. The condition involves sudden and progressive intestinal inflammation and tissue death, and survivors often face long-term intestinal and neurodevelopmental complications.

There is currently no tool to predict which preterm infants will get the disease, and NEC often goes unrecognized until it’s too late to effectively intervene. Researchers don’t yet understand the causes of NEC, but several studies have focused on shifts in the intestinal microbiome, the bacteria in the intestine whose composition can be determined from DNA sequencing from small stool samples.

Researchers from Columbia Engineering and University of Pittsburgh School of Medicine hypothesized that using machine learning to model clinical, demographic, and microbiome data from premature infants may allow identification of patients at high risk for NEC before disease onset. This could enable early intervention and mitigation of serious complications.

“If doctors could accurately predict NEC before the baby actually becomes sick, there are some very simple steps they could take—treatment could include stopping feeds, giving IV fluids, and starting antibiotics to prevent the worst outcomes such as long-term disability or death,” said the study’s lead author Thomas A. Hooven.

The team examined 2,895 stool samples from 161 preterm infants, 45 of whom developed NEC. Because of the complexity of the microbiome data, researchers performed several data preprocessing steps to reduce its dimensionality, and to address the hierarchical nature of this data to harness it to machine learning.

Since human microbiomes are subject to change, the machine learning methods developed by the researchers address the sequential aspects of the problem. For example, in the first 20 days after an infant is born, the infant’s microbiome goes through a drastic change. The team noted that many studies have shown that infants with a higher diversity of microbiome are typically healthier.

“This led us to think that changes in microbiome diversity can help to explain why some infants are more likely to be sick from NEC,” said Adam Lin, a computer science MS student and co-author of the study.

Instead of viewing microbiome samples from an infant as independent, the team represented each patient as a collection of samples and applied machine learning algorithms to understand the complex relationships among the samples. The approach achieved a good balance of specificity and sensitivity in repeated trials, researchers found.

“The Area Under the ROC Curve (AUC) is about 0.9, which demonstrates how good our models are at distinguishing between affected and unaffected patients,” said the study’s co-author, Ansaf Salleb-Aouissi, a senior lecturer in discipline from the computer science department at Columbia Engineering and a specialist in artificial intelligence and its applications to medical informatics.

“Ours is the first effective system for a clinically applicable machine learning model that combines microbiome, demographic, and clinical data that can be collected and monitored in real-time in a neonatal ICU. We are excited about extending its applicability to a new area of predictive monitoring in medicine.”

The research team is now working to develop a noninvasive standalone testing platform for accurate identification of infants at high risk for NEC before clinical onset to prevent the worst outcomes. Once the platform is ready, they plan to conduct a randomized clinical trial to validate their technique’s predictions in a real-time neonatal ICU cohort.

“It’s amazing how we may be able to use machine learning to stop this from happening to babies. We looked at the data and developed a tool that can truly be useful, even life-saving,” said Salleb-Aouissi.

“NEC represents an excellent application from a machine learning perspective. The lessons we’ve learned from our new technique could well translate to other genetic or proteomic datasets and inspire new machine learning algorithms for healthcare datasets.”

The machine learning approach could signal better outcomes for patients and families, the researchers noted.

“For the first time I can envision a future where parents of preterm infants, and their medical teams, no longer live in constant fear of NEC,” said Hooven.