New Amazon AI can transcribe a doctor's natural speech to text

Amazon announces a new speech recognition service for clinical settings

Amazon's newly introduced Amazon Transcribe Medical is an automatic speech recognition technology that understands medical lingo -- abbreviations and all -- and can transcribe a doctor's speech to text.

Two years after Amazon announced the Amazon Transcribe service, a tool that automatically converts speech to text complete with natural formatting and punctuation, the company announced Amazon Transcribe Medical, a similar tool that has been optimized for clinical settings.

This tool would automatically convert the speech of doctors reading aloud their clinical notes to text in real time. According to Amazon, clinicians don't need to feed the tool punctuation by saying things like "comma" or "full stop" -- these details are understood by the technology.

Once complete, these text files can then be sent to electronic health record systems or AWS language services like Amazon Comprehend Medical, which pulls out specific medical information from unstructured texts.

Matt Wood, the VP of AI at Amazon Web Services, told CNBC that the purpose of this technology, "is to free up the doctor, so they have more attention going to where it should be directed. And that's to the patient."

Likewise, this could potentially save resources in clinical settings, as human scribes and third-party transcription services will be less frequently needed. Amazon states that the service is HIPAA eligible meaning that it protects the privacy of individuals' medical data.

Right now, Amazon Transcribe Medical is available in the East and West regions of the US.