Researchers from Berkeley Lab at the University of California have been applying deep learning and analytics to electronic health record data to help the Veterans Administration (VA) tackle medical and psychological challenges of returning service members.

Part of a collaboration between the U.S. Department of Energy (DOE) and the VA, the team has worked with a publicly available dataset containing medical record information on about 40,000 patients from one Boston hospital intensive care unit to search for patterns that might point to suicide risk.

"The VA has been collecting medical records and genomic data from some 700,000 veterans, and they need help from DOE to interpret all of this information to improve healthcare for these individuals," said Silvia Crivelli, lead for Berkeley Lab’s involvement in the suicide prevention project, said in a statement.

The goal is to improve identification of patients at risk for suicide through new patient-specific algorithms that produce tailored and dynamic suicide risk scores, and make those resources available to VA caregivers and patients. The program is also focused on prostate cancer and cardiovascular disease.

Suicide is the 10th leading cause of death in the U.S., and it is significantly higher in the veteran population, with 20-22 deaths per day – an alarming statistic the VA’s Million Veteran Program Suicide Prevention Exemplar project is designed to help address.

Early efforts focused primarily on finding patterns in a diverse and complex pool of data, such as building a deep learning network that can distinguish and classify patients at high risk for suicide from discharge notes and physicians’ notes found in these datasets.

According to researcher Rafael Zamora-Resendiz, the neural network was trained to classify between patients who are at high risk for suicide and those who are not to find patterns in the doctors’ language.

“We then applied some techniques on the trained network to find which words contributed most to the final prediction,” he explained. “The real challenge is figuring out a way of tracing how these words are combined internally within the network. This will help provide better insight on common motifs found between suicidal patients.”

DOE and the VA are also working to apply supercomputing, software development and networking to medical data sets and EHR data from another 22 million vets.