Photo credit: Executive Mosaic
As the federal government deploys artificial intelligence, the technology’s impact could manifest itself in multiple areas, including healthcare.
Chief Information Officer, Department of Health and Human Services Dr. Karl Mathias In my keynote at , I highlighted three key areas of AI advancement we’re seeing in healthcare. Potomac Officers Club4th Annual CIO Summit.
NIH grant analysis
Dr Matthias said the National Institutes of Health has received “a huge number of grants.” 2023 Wash100 Award Winner Screening the thousands of grant applications became a major challenge for the NIH.
“What they did was use AI to analyze grants. Think of it as an enhanced resume scanner. And we’re not going to review it and decide whether to approve it – we probably will soon – but we will decide which team to send it to. [It] It showed an accuracy rate of 92%, so it basically solved the problem of getting through,” explained Dr Matthias.
AI is also used further down the grant lifecycle. Each year, each grant he receives one audit. That means the HHS Inspector General’s Office receives between 35,000 and 45,000 audits per year.
“They have an AI system that looks at these things and looks at them and highlights some red flags. It is.” For both American taxpayers and healthcare workers, the difference.
Fraud detection in CMS
Medicare and Medicaid Service Centers use AI to screen millions of transactions and detect potential fraud.
“They have developed an AI system that combs and analyzes potentially problematic trends and highlights them for investigation by investigators,” said Dr Matthias. But as threats continue to evolve, healthcare organizations must evolve as well.
“The bad guys adjust their tactics…they know what they’re looking at. Now the tree-based model [CMS uses] Also, adjust your tactics and ditch the old in favor of the new model. They’re getting pretty good, but still, people committing crimes will change their tactics, we have to change their tactics,” he urged.
Detection of counterfeit drugs
HHS has a device called CD5 that shines light on drugs at different wavelengths and takes pictures. Second, detecting counterfeits requires human analysis and visual comparison. This process typically takes up to 30 minutes, but with AI built into the tool, it can be done in seconds.
“These are important,” Dr. Matthias told the audience directly. “And the reason they matter is because we are talking about fraud. It can become something.”
In his closing keynote, Dr. Matthias touched on several high-priority focus areas such as data and cybersecurity. If you’re into cyber, don’t miss the Potomac Officers’ Club event. 2023 Cyber Summit June 8th! register here.
