Posted On: Apr 2, 2020

Amazon Transcribe Medical is HIPAA eligible automatic speech recognition (ASR) service that makes it easy for developers to add speech-to-text capabilities to their healthcare and life science applications. Up until now, Amazon Transcribe Medical has been available as a streaming API whereby users can pass realtime audio for transcription. Starting today, customers of the service can conveniently transcribe recorded medical audio files too.

Use cases for batch medical speech-to-text include helping clinicians transcribe their pre-recorded patient notes. Applications designed to drive clinical documentation efficiency can now leverage Amazon Transcribe Medical’s state-of-the-art speech recognition engine to efficiently transcribe large quantities of medical recordings with high accuracy. In other applications, such as drug-safety monitoring (pharmacovigilance), pharmaceutical companies can employ Amazon Transcribe Medical in their contact center analytics to transcribe patient and clinician phone calls about medicines and their associated effects at scale.

Using batch transcription processing is easy. Simply upload audio files to an Amazon S3 bucket and then create transcription jobs by pointing the service to that bucket. Batch medical transcription comes with many useful capabilities out-of-the-box. For instance, if you have multi-channel medical audio, as is common for contact centers, Channel Identification is available to tease out the caller and the agent speech for separate transcription. Or, if you have recorded audio of a clinician or patient conversation on a single channel, you can use Speaker Identification to tease out which person is saying what. As usual, all transcripts are accompanied by word-level timestamps that make tracking and use of output transcripts convenient.

Support for Amazon Transcribe Medical batch mode is available in all regions where the service is available. To learn more about the batch transcription feature, see this technical documentation and FAQs.