The National Artificial Intelligence Institute is developing an approach to bring uniformity in AI projects across the Veterans Health Administration and its various regions and medical centers. Veterans Affairs Secretary Denis McDonough recently announced that big data principles will be integrated with trustworthy AI framework at the AI Summit for Health. Due to issues that have been reported, the agency is also focusing on several priorities developing AI capabilities and enabling operational use cases, including an AI touchscreen that was also announced at the summit that will center around health care worker burnout.
The AI tech sprint centered around health care worker burnout will focus on specific areas. One will be spoken language converted into text through AI and natural language processing when a patient is with a provider. The second will be using AI through optical character recognition where the information is turned into text so the provider can find information quicker.
“So all of these are really ways to make the systems work faster. They augment what the provider is doing. And of course, the provider then has the ability to then leverage that to decide what to do moving forward,” Gil Alterovitz, VA chief AI officer and director of NAII, said on Federal Monthly Insights – Artificial Intellience. “AI, like any other technology, it really depends on essentially what is your use case and what you want to do.”
NAII’s mission through iteration and piloting is to move forward on AI in different priority areas.
“There’s a lot of really exciting work happening right now, and we are focused on a number of different priorities and developing AI capabilities to allow for both translating from research and development, and moving forward to enable operational use cases of artificial intelligence. In doing so, we have created a network of National Artificial Intelligence Institute centers across the United States,” Alterovitz told the Federal Drive with Tom Temin.
One of the agency’s key points is getting input from veterans to tackle issues considered high priority. Alterovitz said they have VA medical centers that have different use cases that they’re involved in, across research, development and operational, as well as a community of people who are interested in AI and have ideas.
“There’s an approach being developed around governance within the Veterans Health Administration as well, to ensure and bring uniformity across the different parts of the different medical centers across the country and different regions,” Alterovitz said. “And so work is ongoing there. And it’s really interesting that sometimes it may be an idea that comes from a single person in one site that then informs us and allows us to learn about what potentially could scale.”
“In some senses we are able to also engage and get that input from really across the world through our AI tech sprints, which is a way to engage externally around specific use cases to see if there are solutions out there that may work for specific cases,” Alterovitz added.