Healthcare companies are using artificial intelligence and machine learning, and then the cloud to better ingest, manage and leverage a variety of data – whether it is structured data, unstructured data or streams, to break down silos and enable data liquidity in support of collaborative research and care coordination.

"The cloud enables healthcare providers to scale up during peak demand periods, like flu season, and scale back down again when demand has ebbed," said Shez Partovi, director of worldwide healthcare and life sciences, business and market development at Amazon Web Services. "They can then process the data, apply deep learning, and visualize the data in order to make insightful decisions throughout a patient's care journey – or even throughout the research pipeline."

For instance, Orion Health hosts data for 50 million users on the AWS Cloud, enabling its customers to access patient information ranging from clinical information and genomics to claims and reimbursement data, Partovi added. In turn, providers can identify personalized treatment and prevention strategies and optimize clinical decision making.

"Additionally, AI and machine learning are providing the tools to process and analyze the increasing amount of data generated by doctors, hospitals, researchers and organizations, including structured data like EHR forms as well as unstructured data, such as emails, text documents and even voice notes," said Patrick Combes, technology leader, healthcare and life sciences, at AWS.

To that end, AWS recently announced Amazon Comprehend Medical, a machine learning service that can help process unstructured data such as medical notes, prescriptions, audio interview transcripts and radiology reports – as well as identify information such as patient diagnosis, treatments, dosages, systems and signs.

"Machine learning is being used in a variety of tasks such as analyzing medical images to advancing precision medicine," Combes added. "Tools that leverage natural language processing, pattern recognition and risk identification also are fueling new models for predictive, preventive and population health with great potential to help providers identify gaps in care and help improve the health of individuals and communities."

One example is Philips' HealthSuite digital platform, a cloud-based trove with more than 21 petabytes of data from 390 million medical images, medical records and patient inputs – giving providers, clinicians, data scientists and software developers access to both quality data and AI tools to deliver a more personalized care experience, he explained.

Healthcare providers globally are facing increasing internal and external pressure to incorporate data into their decision making to help improve care quality, reduce costs and drive better patient experience and outcomes.

Additionally, there is a growing amount of unstructured data resulting from the shift from structured forms to text and voice notes – driving an opportunity for AI and machine learning.

"Processing this data creates a complex, expensive and timely coding process for medical billers and higher dissatisfaction from providers as they are forced to spend more time responding to inquiries to clarify and identify segments of their notes, and less time spent on patient care," Combes said.

Combes added that AWS is seeing significant interest in machine learning and AI across the healthcare industry to help mine both structured and unstructured data in clinical settings.

Fred Hutchinson Cancer Research Center in Seattle, for instance, is using Amazon Comprehend Medical to evaluate millions of clinical notes to extract and index medical conditions, medications and choice of cancer therapeutic options, reducing the time to process each document from hours to seconds, he explained.

"As healthcare companies from startups to established multinationals look to AI and machine learning, there are several essential ingredients that are key to success," Combes said. "Large quantities of carefully curated, high-quality data; optimized systems that comply with industry standards and regulations; machine learning services that eliminate the heavy lifting of building, training and deploying models; and the cloud."

While curating high-quality data can be especially challenging in the healthcare industry, which is plagued with highly complex and unstructured data, it is essential to operate AI- and machine learning-driven data sets, Partovi said.

"After successfully establishing the foundational elements,” Partovi said, “healthcare organizations can unlock the power of AI and machine learning with the potential to enhance decision making, drive greater value for patients and providers, and reduce time to discovery and insight.”

Amazon Web Services will be in Booth 5058.

Twitter: @SiwickiHealthIT
Email the writer: bill.siwicki@himssmedia.com