Oak Ridge National Lab, VA Using Machine Learning to Improve Health Care for Veterans

Oak Ridge National Laboratory announced that its researchers, working with counterparts in the Department of Veterans Affairs, Harvard’s TH Chan School of Public Health, Harvard Medical School and Brigham and Women’s Hospital have developed a novel, machine learning-based technique to explore and identify relationships among medical concepts. The new technique involves using electronic health record data across multiple health care providers, ORNL said Thursday.

In a statement, ORNL said the newly-developed technique, called Knowledge Extraction via Sparse Embedding Regression, was recently published in Nature Digital Medicine. The process integrates electronic health record data sourced from the VA and Boston-based Partners Healthcare then “provides automated feature selection that leads to phenotype identification algorithms and knowledge discovery.”

Dr. Katherine Liao, a principal investigator of KESER at VA Boston and associate professor of medicine at Harvard Medical School, said the technique provides a high-level view of the relationships between clinical knowledge that doctors could fail to see when caring for patients at the individual or group level.

ORNL said that in 2016, it began collaborating with the VA on MVP-CHAMPION, a big-data initiative to create a large, precision-medicine platform to host medical records covering some 24 million veterans. The laboratory’s team then set out to automate the identification of phenotypic relationships while providing visibility into the underlying machine learning assumptions and decision processes.

It was further explained that the KESER methodology involves four steps. These include converting data into a structured format, constructing a low dimensional vector representation of each medical code, selecting features to attribute importance and mapping attributed relationships as a network.