Amazon Transcribe Medical – Real-Time Automatic Speech Recognition for Healthcare Customers

In 2017, we launched Amazon Transcribe, an automatic speech recognition service that makes it easy for developers to add speech-to-text capability to their applications: today, we’re extremely happy to extend it to medical speech with Amazon Transcribe Medical.

When I was a child, my parents – both medical doctors – often spent evenings recording letters and exam reports with a microcassette recorder, so that their secretary could later type them and archive them. That was a long time ago, but according to a 2017 study by the University of Wisconsin and the American Medical Association, primary care physicians in the US spend a staggering 6 hours per day entering their medical reports in electronic health record (EHR) systems, now a standard requirement at healthcare providers.

I don’t think that anyone would argue that doctors should go back to paper reports: working with digital data is so much more efficient. Still, could they be spared these long hours of administrative work? Surely, that time would be better spent engaging with patients, and getting a little extra rest after a busy day at the hospital?

Introducing Amazon Transcribe Medical
Thanks to Amazon Transcribe Medical, physicians will now be able to easily and quickly dictate their clinical notes and see their speech converted to accurate text in real-time, without any human intervention. Clinicians can use natural speech and do not have to explicitly call out punctuation like “comma” or “full stop”. This text can then be automatically fed to downstream applications such as EHR systems, or to AWS language services such as Amazon Comprehend Medical for entity extraction.

In the spirit of fully managed services, Transcribe Medical frees you from any infrastructure work, and lets you scale effortlessly while only paying for what you actually use: no upfront fees for costly licenses! As you would expect, Transcribe Medical is also HIPAA eligible.

From a technical perspective, all you have to do is capture audio using your device’s microphone, and send PCM audio to a streaming API based on the popular Websocket protocol. This API will respond with a series of JSON blobs with the transcribed text, as well as word-level time stamps, punctuation, etc. Optionally, you can save this data to an Amazon Simple Storage Service (Amazon S3) bucket.

Amazon Transcribe Medical In Action
Let’s do a quick demo with medical text from MT Samples, a great collection of real-life anonymized medical transcripts that are free to use and distribute.

I’m using a streaming application modified for Transcribe Medical, and you’ll be able to do the same in the AWS console. You can view a video recording of this demo here.

Now Available!
You can start using Amazon Transcribe Medical today in the US East (N. Virginia) and US West (Oregon) regions.

Give it a try, and please share your feedback in the AWS forum for Amazon Transcribe, or with your usual AWS support contacts.