Amazon Web Services Inc. said today during its re:Invent 2020 virtual conference that it’s upping its medical chops with the launch of a new service that healthcare providers can use to store, transform and analyze life science data in the cloud at petabyte scale.

The service is called Amazon HealthLake, and it’s essentially a data lake tailored for the healthcare industry that’s compliant with Health Insurance Portability and Accountability Act that governs the use of medical data.

The debut followed a flurry of other new machine learning-related announcements today during a keynote address by Swami Sivasubramanian, vice president of Amazon machine learning. In particular, Amazon announced nine new capabilities for its Amazon SageMaker managed machine learning service, mostly aimed at making it easier for those not expert in AI to do things such as data preparation, manage edge devices and reduce bias in machine learning models.

Like other data lakes, Amazon HealthLake serves as a centralized repository that enables organizations to store data in its natural, or raw format and analyze it without preparing it first. Available in preview in Amazon’s US East (N. Virginia) region starting today, Amazon HealthLake is designed to give healthcare providers an easier way to analyze the stacks of unstructured medical data they accumulate from their patients.

In a keynote presentation, AWS Vice President for AI Matt Wood explained that healthcare providers struggle to apply intelligence to their data because it’s usually spread across numerous repositories in various formats such as clinical notes, reports and image scans. It can take months to prepare, stage and transform that data for analysis, he said.

“It is a Herculean effort to analyze that data,” Wood said. “We can bring it all together in minutes.”

With Amazon HealthLake, there’s no need to do that. The service leverages various types of artificial intelligence, including natural language understanding to reach doctor’s notes, and ontology mapping to read medical images.

Wood showed how it’s possible to use Amazon HealthLake to quickly analyze a massive database and highlight a subset of patients who are struggling to manage their diabetes properly. The service then made recommendations on how to adjust each person’s treatment to avoid further complications. After querying the data available, it generated a visualization in Amazon QuickSite that can be monitored in the context of other patients.

Another benefit of Amazon HealthLake is that data can be offloaded to a service such as Amazon SageMaker to train machine learning models to make forecasts, for example showing which patients are likely to develop diabetes within the next year.

The service also supports common interoperability standards such as Fast Healthcare Interoperability Resources, which provides a consistent format for healthcare data sharing.

Amazon HealthLake is another sign that AWS sees healthcare as an industry that has massive growth potential for its cloud services. Last year, the company announced a service called Amazon Transcribe Medical that enables medical professionals to dictate notes and record patient conversations. In 2018 it debuted Amazon Comprehend Medical, a service that uses AI to mine medical records for information that can be used to improve patient treatment and reduce costs.

Constellation Research Inc. analyst Holger Mueller said today’s news underscores once again that 2020 is the year of vertical cloud announcements. He said Amazon has been using re:Invent 2020 to showcase the ways in which it’s providing more depth to its traditional services with packaged solutions.

“Amazon HealthLake fits this new direction perfectly, it’s a software-as-a-service offering of data lake that’s specialized for the healthcare industry,” Mueller said. “Amazon loves to relieve customer pain points and data lakes have always been a challenge for enterprises, so a managed, secure and industry-oriented service like this will something healthcare executives will definitely be keen to look into.”

Images: Amazon