How VA is Applying Artificial Intelligence to Proactively Solve Veterans’ Problems

NatalieSchorr/Shutterstock.com

The agency’s newly appointed artificial intelligence director outlined his vision and approach.

As the Veterans Affairs Department’s inaugural Director of Artificial Intelligence, Gil Alterovitz aims to leverage the emerging technology and the agency’s cornucopia of data to proactively anticipate and tackle problems afflicting veterans like never before. 

In a conversation with Nextgov, Alterovitz detailed his present efforts and future-facing vision to support VA in executing that mission. 

“Nowhere in the country is there such potential for research to be developed and translated into clinical care so quickly. In this case, it’s to help our special population of veterans … and those patients have actually asked us to deal with their needs,” Alterovitz said. “We really want to be the go-to place for veterans through AI research and development—so instead of reacting, we can really anticipate their needs.”

Veterans Affairs has one of the largest integrated medical records systems and access to one of the most robust data collections in the world. The new director highlighted a variety of projects in which the agency is already making new strides through the implementation of AI. 

This week, the department released a paper in Nature, a publication Alterovitz said is widely accepted as one of the most respected biomedical science journals in the world. The study demonstrates how the agency, in collaboration with technology company DeepMind Health, developed an AI system that can forecast a life-threatening kidney disease before it appears. 

The research team applied machine learning to a large population of patient data—over 700,000 from over 100 VA sites—to address severe kidney failure or damage known as Acute Kidney Injury, or AKI. Alterovitz said AKI is incredibly difficult for doctors to detect, noting that they usually see results once the damage is already done or when it’s too late to reverse repair. 

But implementing the AI model enabled researchers to identify more than 90% of the most severe AKI cases 48 hours sooner than usual care.

“This means you can do treatment to prevent acute renal injury from having long-lasting effects, so that’s a huge gain for us,” Alterovitz said. “It really is an example where we can anticipate this ahead of time and be that go-to place for veterans through AI research and development.”

He added that people are already looking to translate the results of the study into other languages.

The AI director also said he is already working to enhance the agency’s technical capacity and accelerate partnerships with stakeholders across industry and academia, which was illuminated as a priority in the White House Office of Science Technology and Policy’s recently unveiled National AI Research and Development Strategic Plan. Alterovitz was a core writer of that national strategy. 

As part of that effort, he is now leading an “AI Tech Sprint” at VA that aims to enable outside organizations to test out data in the VA format to develop AI tools and programs that can lead to new data-driven insights without waiting for drawn out periods to establish partnership agreements. The program is modeled off of a “Health Tech Sprint” that Alterovitz co-led while he was a Presidential Innovation Fellow with the Health and Human Services Department.

“Whether it be around de-identified data or data that’s synthetic, it’s data in the VA format. So as people develop new tools ... that will be useful for everybody, they are using that time to learn about how the VA data structures work,” he said. “So rather than waiting a long time and making agreements without the tools, they are able to start it right away.”

Having only served in this role for about a month, Alterovitz is also working with VA insiders to design a roadmap for success and pursuing a variety of avenues to establish his internal team. He’s taking an inventory of the VA’s wide variety of AI and other innovation-focused projects that are already in the works with the hope of coordinating those activities so people across all of the agency’s research areas can benefit. 

“When you think about how we are preparing ourselves, what we want to do is set a vision for the future and leverage all the strengths that we have—so we’re thinking about the technical talent, the industrial agility, industry and the ability of academia to foster innovation and leverage the strengths of this country to work toward the mission of the veterans,” he said.

And when it comes to measuring success, Alterovitz said it all comes back to the veterans who he serves. To him, it’s really about finding results that can lead to actionable change around problems that veterans really care about and can feel themselves—much like the results of that recent study in Nature that can be implemented in veterans’ critical care. 

It’s work that is catching the eyes of other federal agencies.

“Actually over the weekend, people from a couple other agencies were emailing me and asking me, ‘oh we would like to set up a similar position, can you tell us about the types of things such a person could do, what kind of things would help advance the field forward and how different nations can collaborate?’” he said. “So, yes I think looking forward and really thinking of the future is really a mindset that can allow you to have a different type of perspective and one that I think we are starting to see others wanting to adopt as well.” 

For the new director, it feels like a moment where there’s a perfect “alignment on a number of different stars”: the VA’s mission to help veterans who sacrifice for the country—which particularly attracted him to the role, patients specifically asking the agency to solve problems using their data, and new technologies like AI enabling researchers to look at large data sets and identify revolutionary health solutions in an unprecedented manner. 

“So I’m excited, these are exciting times,” he said. “I feel like it’s a really special moment.”