Amazon is expanding its range of health-focused apps and services with a platform that offers AI tools to help clinicians transcribe and analyze their conversations with patients.

At its annual AWS Summit conference in New York, Amazon unveiled AWS HealthScribe, an API to create transcripts, extract details and create summaries from doctor-patient discussions that can be entered into an electronic health record (EHR) system. The transcripts from HealthScribe can be converted into patient notes by the platform’s machine learning models, Amazon says, which can then be analyzed for broad insights.

“Documentation is a particularly time-consuming effort for healthcare professionals, which is why we’re excited to leverage the power of generative AI in AWS HealthScribe and reduce that burden,” Bratin Saha, VP of machine learning and AI services at AWS, said in a blog post shared with TechCrunch via email. “Today’s announcement builds on AWS’s commitment to the healthcare and life sciences industry and our responsible approach to technologies like generative AI to help reduce the burden of clinical documentation and improve the consultation experience.”

HealthScribe identifies speaker roles and segments transcripts into categories based on clinical relevance, like “small talk,” “subjective comments” or “objective comments.” In addition, HealthScribe delivers natural language processing capabilities that can be used to extract structured medical terms from conversations, such as medications and medical conditions.

The notes in HealthScribe, augmented by AI, include details like the history of the present illness, takeaways and reasons for a visit.

Amazon says that the AI capabilities in HealthScribe are powered by Bedrock, its platform that provides a way to build generative AI-powered apps via pretrained models from startups as well as Amazon itself. This might be cause for alarm, given generative AI’s tendency to exhibit biases, confidently invent facts and generally go off the rails.

Speech recognition algorithms, too, often contain biases. One recent study published in The Proceedings of the National Academy of Sciences showed that speech recognition systems from leading tech companies were twice as likely to incorrectly transcribe audio from Black speakers as opposed to white speakers.

As a piece in Scientific American points out, in normal conversations, we might choose to “code-switch” depending on the audience. But there’s no code-switching with automated speech recognition programs — either you assimilate, or you’re not understood.

So is HealthScribe consistent? Can it be trusted, particularly when it comes to deciding whether to label a part of a discussion as “subjective” or “objective” or identifying medications? And can it handle the wide array of different accents and vernaculars that patients and providers might use?

The jury’s out on all that.

But perhaps in an effort to prevent some of the more major potential mistakes, HealthScribe can only create clinical notes for two medical specialties at present, general medicine and orthopedics. And the platform offers clinicians a chance to review notes before finalizing records in their EHR, providing references to the original transcript for sentences used in the AI-generated notes.

Given the reluctance of some companies to let generative AI apps field sensitive data, it’s not surprising that Amazon’s also highlighting HealthScribe’s security and privacy aspects in its marketing materials. HealthScribe doesn’t retain customer data after processing requests and encrypts data at transit and at rest, and Amazon says it doesn’t use the inputs and outputs generated by HealthScribe to train any of its AI models.

Healthcare software providers using HealthScribe also have control over where they want to store transcriptions and preliminary clinical notes. But the service isn’t compliant out of the box with HIPAA, the U.S. law that provides protections for personal health information.

HealthScribe is “HIPAA eligible,” however — meaning customers who work with Amazon to meet HIPAA requirements can ultimately reach compliance. Those customer must sign a contract known as a business associate addendum, which AWS’ documentation covers in detail here.

Amazon says that 3M Health Information Systems, Babylon Health and ScribeEMR are among the companies already using HealthScribe.

Alongside HealthScribe, Amazon today announced AWS HealthImaging, a service designed to make it easier to store, transform and analyze medical imaging data “at a petabyte scale”

With HealthImaging, customers can run medical imaging apps from a single copy of each medical image in the AWS cloud. As an added benefit of running on AWS infrastructure, HealthImaging enables dynamic pricing for active and archive data as well as “subsecond” image access latencies from AWS’ Frequent Access or Archive Instant Access storage tiers.

Amazon claims the average organization can reduce the total cost of ownership of their medical imaging storage up to 40% — a bold statement, to be sure.

HealthImaging is now generally available in the US East (N. Virginia), US West (Oregon), Asia Pacific (Sydney), and Europe (Ireland) AWS regions.