VA Aims To Reduce Administrative Tasks With AI, Machine Learning
A wide-reaching program looks for fresh machine learning technology to complete menial procedures and tasks, allowing VA clinicians and partners to focus on care.
Officials at the Department of Veterans Affairs are looking to increase efficiency and optimize their clinicians’ professional capabilities, featuring advanced artificial intelligence and machine learning technologies.
In a November presolicitation, the VA seeks to gauge market readiness for advanced healthcare device manufacturing, ranging from prosthetic solutions, surgical instruments, and personalized digital health assistant technology, as well as artificial intelligence and machine learning capabilities.
Dubbed Accelerating VA Innovation and Learning, or AVAIL, the program is looking to supplement and support agency health care operations, according to Amanda Purnell, an Innovation Specialist with the VA
“What we are trying to do is utilize AI and machine learning to remove administrative burden of tasks,” she told Nextgov.
The technology requested by the department will be tailored to areas where a computer can do a better, more efficient job than a human, and thereby give people back time to complete demanding tasks that require human judgement.
Some of these areas the AI and machine learning technology could be implemented include surgical preplanning, manufacturing submissions, and 3D printing, along with injection molding to produce plastic medical devices and other equipment.
Purnell also said that the VA is looking for technology that can handle the bulk of document analyses. Using machine learning and natural language processing to scan and detect patterns in medical images, such as CT scans, MRIs and dermatology scans is one of the ways the VA aims to digitize its administrative workload.
Staff at the VA is currently tasked with looking through faxes and other clinical data to siphon it to the right place. AVAIL would combine natural language processing to manage these operations and add human review when necessary.
Purnell said that the forthcoming technology would emphasize “streamlining processes that are better and faster done by machines and allowing humans to do something that is more kind of ‘human-meaningful,’ and also allowing clinicians to operate to the top of their license.”
She noted that machines are highly adept at scanning and analyzing images with AI. The VA procedure would likely have the AI technology to do a preliminary scan, followed by a human clinician to make their expert opinion based on results.
With machine learning handling the bulk of these processes along with other manufacturing and designing needs, clinicians and surgeons within the VA could focus more on applying their medical and surgical skills. Purnell used the example of a prosthetist getting more time to foster a human connection with a client rather than oversee other health care devices and manufacturing details.
“It is making sure humans are used to their best advantage, and that we’re using technology to augment the human experience,” she said.
The AVAIL program also stands to improve the ongoing modernization effort of the VA’s beleaguered electronic health record (EHR) system, which has suffered deployment hiccups thanks to difficult interfaces and budget constraints.
The AI and machine learning technology outlined in the presolicitation could also support new EHR infrastructure and focus on an improved user experience, mainly with an improved platform interface and other accessibility features.
Purnell underscored that having AI manage form processing and data sharing capabilities, including veteran claims and benefits, is another beneficial use case.
“We’re alleviating that admin burden and increasing the experience both for veterans and our clinicians, in that veterans are getting more facetime with our clinicians and clinicians are doing more of what they are trained to do,” Purnell said.