During its re:Invent 2020 virtual keynote today, Amazon launched Amazon HealthLake, a service that enables health care organizations to store, transform, and analyze up to petabytes of life science data in Amazon Web Services. Amazon says that the HIPAA-eligible HealthLake, which is available in preview starting today, can automatically understand and extract medical information including rules, procedures, and diagnoses in real time.
Health care data is often spread across various systems such as electronic medical lab systems, and it’s challenging to organize because it’s often unstructured. Data in medical records like clinical notes, reports and forms like insurance claims, and image scans needs to be prepped and normalized before analyses can begin.
HealthLake aims to address this challenge by enabling customers to apply intelligence to hundreds of thousands of data points across different siloes in dozens of formats. For example, HealthLake leverages natural language understanding and ontology mapping to identify whether a patient has been properly prescribed a drug, pulling out information from blood glucose monitoring systems, physicians notes, insurance forms and lab reports, and more to inform its conclusions. Data can be loaded on an ongoing basis and queried and searched using standard methods or imported into Amazon SageMaker to train models to forecast metrics such as the number of diabetes cases year over year.
During a livestreamed demonstration, AWS vice president for AI Matt Wood showed HealthLake being used to highlight a subset of patients with uncontrolled diabetes and adjust treatments to avoid severe complications. HealthLake queried data directly from a health care provider and generated a visualization in Amazon QuickSight that could be monitored in the context of other patients.
Want to publish your own articles on DistilINFO Publications?
Send us an email, we will get in touch with you.
Amazon says that HealthLake recognizes interoperability standards including the Fast Healthcare Interoperability Resources (FHIR), a standard format to enable data sharing of health systems in a consistent format. Support for additional standards will arrive further down the line.
As we wrote last year, Amazon views AI in health care as a frontier worth exploring — and perhaps its next major revenue driver. The AI in health care market is anticipated to reach $19.25 billion by 2026, driven in part by a demand for telemedical and remote monitoring services. The launch of HealthLake comes a year after Amazon debuted Transcribe Medical, a service that’s designed to transcribe medical speech for clinical staff in primary care settings. And in 2018, Amazon made three AWS offerings HIPAA eligible — Transcribe, Translate, and Comprehend — following on the heels of rival Google Cloud.