At the Department of Veterans Affairs, the goal of increasing AI use cases means improving the health outcomes of veterans and increasing job satisfaction for VA employees who serve veterans. Identify veterans who may benefit from specific interventions through programs such as the Stratification Tool for Opioid Risk Reduction (STORM) and Recovery Activities and Coordination for Health – Veterans Enhanced Treatment (REACH VET). Data is being analyzed to proactively identify. Following the Executive Order on Artificial Intelligence, VA was able to identify dozens of use cases for AI. Radiology is particularly at the forefront, with more than 75% of FDA-approved AI devices coming from radiology departments, but the department notes that AI has the ability to improve functionality across the practice. Masu.
“Generally, at VA, we think AI has the potential to contribute in the near term in a very broad range of areas. And I think that falls into three main categories: provider burnout; Syndrome mitigation is one category. In that vein, we are actively conducting an AI Technology Sprint that is part of the AI Executive Order. We currently have two systems in operation. One is a record of clinical encounters. and the other is extracting information from paper medical records,” said Kimberly McManus, deputy chief technology officer for artificial intelligence at the Department of Veterans Affairs. Federal Monthly Insights – Operationalizing AI. “Another area similar to this is improving the veteran and staff experience more broadly. It's about how we can augment… And the third area is improving patient care, and that's where I see a lot of FDA-approved medical devices, like radiology medical devices. , there is also something new in pathology and dermatology.”
One way the VA is addressing the staff experience is by using optical character recognition to extract handwritten notes and convert them into computer-written notes. Once these notes are transformed, AI can be used to compile and extract information that caseworkers may be looking for.
“Generative technology may be used to better summarize and identify information. This type of search and summarization is useful for veterans, whether healthcare providers are trying to find information in electronic health, or It really applies across the agency, but also to many other companies, whether our benefit judges are looking to identify that information as they process claims,” McManus said.
VA's top priority is suicide prevention among veterans, and AI is playing a big role in that effort. The REACH VET program uses an algorithm to stratify veterans as high or low risk for suicide and identify those in need of intervention.
“We have programs in place that allow us to reach out to veterans in one area and provide support services, but we're also looking at different areas for the future,” McManus said. “Broadly speaking, it involves using factors already recorded in electronic medical records that are known to be clinical risk factors for suicide and combining them to identify a risk score.”
With the health and well-being of veterans at risk and the vast amount of veteran data at our fingertips, VA takes the security of veteran information to the forefront. They have developed six guiding principles to follow to harness the trust of data and AI for themselves and their partners.
- Does it have a purpose?
- Is it effective and safe?
- Is it secure and private?
- Is it fair and just?
- Is it transparent and explainable?
- Are you accountable and monitored?
“We have an enterprise data platform called Summit, which includes much of our EHR electronic health record data and other types of data,” McManus said. Tom Temin and Federal Drive. “We store a lot of our healthcare data in cloud platforms that we store on our network. We also have organizations like Oak Ridge National Laboratory that are doing deeper research and development related to data science and AI. We also collaborate with other organizations.”
As the VA further adopts the use of AI, concerns about drift and bias will become more of an issue for the organization. VA's approach to this is to keep humans in the loop and design use cases around human end users. The human factor also allows agencies to assess whether AI is the right solution.
“The biggest key to the success of AI, machine learning, and AI algorithms is to really get all the end users, medical clinical experts, and workflow experts at the table from the beginning. We understand that, but we need everyone at the table from the beginning to figure out how the algorithms actually fit into the workflow and actually impact the end user,” McManus said. “I'm really excited and optimistic about this space. We're always focused on trustworthy AI and safety. There are many areas that have the potential to positively impact our mission, so I'm optimistic and think we've made a lot of progress. We still have a long way to go, but I'm very excited. ”
Copyright © 2024 Federal News Network. All rights reserved. This website is not directed to users within the European Economic Area.
