Berkeley using a new deep learning program to assess risk of suicide amongst veterans

Identifying patterns of risk within patients often involves a massive amount of data interpretation and algorithmic examination. New computer resources through Berkeley are today being dedicated to producing tailored algorithms for dynamic risk scores for VA patients and caregivers.

Researchers from the Berkeley lab at the University of California have developed a deep learning approach with advanced analytics that is going through recorded data to the Veterans Administration. This task can be used to tackle a series of psychological challenges and medical challenges for returning service members.

The publicly available data set includes learning from 40,000 patient profiles admitted to the Boston hospital intensive care unit. By looking into patterns that could point to suicide risk is possible to identify patients that are at a higher risk for suicide and make sure that resources available for their caregivers as well as on the patient side.

As suicide is currently the 10th leading cause of death in the United States, new initiatives need to be created in order to reduce risks. The veteran population has a significantly higher rate of suicide. The neural network associated with this learning initiative can classify patients that are at a higher risk of suicide as well as find patterns within the previous diagnosis.

Berkeley’s lab is continuing to use a strategic initiative within machine learning and core AI technologies. The Center for clinical artificial intelligence is interested in developing these applications for machine learning and AI in healthcare.

The VA is continuing to update its records to provide more resources to machine learning. With records now available for over 700,000 veterans, teams like the Berkeley Lab will have an even more significant data set that they can use for suicide prevention in deep learning programs.