Amazon’s Newest Machine Learning Product Makes Sense of Unstructured Medical Text

By Dave Muoio, MobiHealth News | November 28, 2018

Amazon Comprehend Medical is a HIPAA-eligible service able to automatically identify and process ​patient diagnoses, symptoms, medical test details, treatments and other relevant medical information.

Amazon Web Services has unveiled a new machine learning tool that looks to help healthcare industry developers process bodies of unstructured medical text.

Called Amazon Comprehend Medical, the HIPAA-eligible service is able to pull out medically-relevant information such as patient diagnoses, symptoms, medical test details, treatments and dosages, while simultaneously highlighting any protected health information.

While the service can directly be fed medical notes, prescriptions, audio transcripts and other reports directly, an API is also available so that the machine learning tool can be integrated within an organization’s existing systems. Amazon also stressed in the announcement that the service does not require an organization to manage any servers, and does not require any data to be stored or saved for model training.

The Comprehend Medical service has already been cutting its teeth with some real-world application — according to Amazon, both the Fred Hutchinson Cancer Research Center and Roche Diagnostics have been previewing the service to, respectively, identify patients applicable to specific cancer therapies and inform decision support portfolios.

“With petabytes of unstructured data being generated in hospital systems every day, our goal is to take this information and convert it into useful insights that can be efficiently accessed and understood,” Anish Kejariwal, director of software engineering for Roche Diagnostics Information Solutions, said in an Amazon blog post. “Amazon Comprehend Medical provides the functionality to help us with quickly extracting and structuring information from medical documents, so that we can build a comprehensive, longitudinal view of patients, and enable both decision support and population analytics.”

What’s the impact

In the absence of an automated machine learning tool, processing unstructured medical texts would require manual data entry by trained staff or the in-house development of custom code. As such, Amazon is pitching its new service as a faster and more cost-effective way to support clinical decision making, revenue cycle management, clinical trial management and population health platform construction.

“The process of developing clinical trials and connecting them with the right patients requires research teams to sift through and label mountains of unstructured medical record data,” Matthew Trunnell, CIO at Fred Hutchinson Cancer Research Center, said in the blog post. “Amazon Comprehend Medical will reduce this time burden from hours per record to seconds. This is a vital step toward getting researchers rapid access to the information they need when they need it so they can find actionable insights to advance lifesaving therapies for patients.”

Although Comprehend Medical is currently designed for providers, insurers, pharmas and medical researchers, Amazon noted in a blog post that the technology “may be able to one day help consumers with managing their own health, including medication management, proactively scheduling care visits, or empowering them to make informed decisions about their health and eligibility.”

What’s the trend

It hasn’t been more than a few weeks since Amazon Web Services showed off another group of HIPAA-eligible machine learning tools: Amazon Translate, Amazon Comprehend and Amazon Transcribe. Along with older offerings such as Amazon Polly, Amazon SageMaker and Amazon Rekognition, the company is now sitting on a number of tech-driven tools and services for healthcare organizations.

And that’s not to mention Amazon’s growing appetite for healthcare consumers — last month the company announced that it will now exclusively sell Arcadia’s new line of consumer-use medical devices, called the Choice brand.

Original Article
2018-11-29T11:56:17+00:00