Speech and voice recognition are machine learning-based technologies that have become increasingly popular in the past few years owing to the widespread adoption of voice-enabled devices such as Amazon Alexa, Google Home and Apple Pod. The exponential level of development of these technologies has allowed for their numerous applications – from simply providing convenience and entertainment to helping doctors provide better patient service.
The difference between speech and voice recognition is that the first involves recording spoken words using either microphone or a telephone, while the latter uses AI software to identify the person who is speaking. Speech and voice recognition also differ in the way they’re used. For example, speech recognition is used for general dictation, transcribing and medical transcription, while voice recognition tech is used to confirm the identity of the speaker or determine the identity of an unknown individual.
At Amazon Web Services’ (AWS) latest annual conference re:Invent that took place in December 2019, the company launched a service which transcribes doctor-patient interactions and puts them directly into patients’ medical files. Called Amazon Transcribe Medical, the service is equipped with voice transcription technology and offers an API that can integrate with voice-enabled applications and any device with a microphone. Output transcripts will support word-level time stamps and confidence scores.
How does Amazon Transcribe Medical work?
Transcribe Medical uses machine learning to provide highly accurate automatic speech recognition (ASR) for the medical industry. It will allow physicians to quickly dictate their clinical notes and speech into accurate text in real time, without any human intervention. The tool even has built-in punctuation, meaning that doctors won’t have to indicate when there’s a “comma” or a “full stop” in a sentence but can speak normally.
The technology is trained to understand the terminology and style of clinical language, meaning that it can be used to capture physician-patient conversations in text for later analysis or for entry into electronic health record systems.
What about data privacy?
The tool is compliant with privacy rules and regulations under the Health Insurance Portability and Accountability Act (HIPAA). This not only ensures that Amazon is accountable for the tons of customer data it stores, but also customers are responsible for ensuring that they’re compliant with patient privacy laws before they use the technology.
Transcribe Medical builds on 2017’s debut of Amazon Transcribe – an automated transcribing services that runs on speech recognition. The service lets you turn an audio file stored on your Amazon Simple Storage Service (S3) account into a grammatically correct text.
For more information on healthcare tech, read some of our other posts on the topic: