The use of AI-based tools is growing in healthcare, partly in response to interest spurred by ChatGPT. While many healthcare leaders are skeptical about the ethics of using generative AI in healthcare, others believe it can be useful for creating more empathetic patient notes and quickly answering health questions for those who may not otherwise engage with the healthcare industry.
According to a press release, AWS HealthScribe is powered by Amazon Bedrock, which enables interoperability between HealthScribe and digital health apps focused on general medicine and orthopedics. As a result, healthcare software providers don’t need to train their own large language models specific to those specialties.
The program provides security and privacy by encrypting data both in transit and at rest and by giving healthcare organizations control over where their data is stored. In addition, AWS does not train its models based on healthcare organizations’ patient-physician conversations.
Physicians can easily review clinical notes before entering them into the EHR, as AWS HealthScribe cites the source of each line of generated text from the original transcript. This helps ensure that clinical notes are accurate.
“Documentation is a particularly time-consuming effort for healthcare professionals, which is why we are excited to leverage the power of generative AI in AWS HealthScribe and reduce that burden,” Bratin Saha, vice president of machine learning and AI services at AWS, said in a press release. Saha said that the announcement “builds on AWS’ 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.”