The leading US-based hyperscalers are tapping into generative AI to help address some of the biggest issues in healthcare, with respect to process inefficiencies and patient diagnostics.
The technology models, which tap into neural networks to spot patterns and structures in data to create new insights, look promising on paper as mechanism to address both healthcare administrative and diagnostic challenges, though there are some very vocal critics.
The US healthcare sector is a primed for digital help. Notoriously slow to invest in information technology and widely known for administrative inefficiencies, healthcare institutions could potentially benefit from generative AI in several ways.
Hyperscalers step in
Amazon Web Services introduced its AI-based HealthScribe service which gives developers a mechanism to build clinical applications that apply speech recognition, artificial intelligence, and advance machine learning algorithms to automate the transcription of clinical data. A 2016 American Medical Association study reported that the ratio in hours of time spent on patient care to administrative duties is 1:2, of which medical transcriptions is a significant portion of the administrative function.
Google Cloud is teaming up with healthcare IT provider CareCloud to create apps that apply generative AI technology to help medical professionals make better informed clinical decisions and improve overall process efficiency, targeted specifically to small and mid-sized providers, the first byproduct of the partnership is expected to be available to clinics and doctors’ offices in the next few months.
Microsoft is also taking the partner route with respect to healthcare and generative AI. The company is working with electronic health record (EHR) giant Epic Systems to integrate Microsoft generative AI technology into Epic’s EHR. The goal is to expedite documentation. Several healthcare systems are piloting the software including UNC Health, UW Health, and Stanford Health Care.
Generative AI misinformation warnings
While the early use cases look promising, there is considerable concern about how accurate the content generated from the models is. Critics have also sounded the alarm that generative AI could be used to spread misinformation.
Generative AI proponents insist that with enough research and extensive testing, the technology can be used to help organizations across different industries improve operating models and realize better returns. Time, and trials, will show how effective generative AI is as a tool for healthcare.