Skip to main content

Analysis

Amazon Machine Learning Award to Accelerate and Scale UPMC-Affiliated Innovations

By Mandy Roth  
   August 13, 2019

Consortium researchers will have access to Amazon's cloud-based platform and machine learning tools, providing early-stage innovation advantages.

Thanks to a Machine Learning Research Award from Amazon Web Services (AWS) to a research alliance supported by UPMC Enterprises, a seed has been planted to accelerate the consortium's medical research initiatives, help participating entrepreneurs more rapidly scale their innovations, and, in some small fashion, contribute to positioning the Pittsburgh area as a healthcare technology innovation hub.

The award provides researchers access to Amazon's cloud-based platform and machine learning tools, enabling them to incorporate sophisticated technology into innovations at an early stage of the development process.

These innovations "will be able to be deployed more easily in the real world," says Rob Hartman, PhD, director of translational science, UPMC Enterprises.

The Amazon award was made to the Pittsburgh Health Data Alliance (PHDA), which was formed four years ago by UPMC, the University of Pittsburgh, and Carnegie Mellon University. UPMC Enterprises, the health system's innovation arm, funds PHDA and commercializes its breakthroughs.

Accelerates Research Initiatives and Builds Scale Into Innovation Process
 

PDHA uses “big data” generated in health care—including patient information in the electronic health record, diagnostic imaging, prescriptions, genomic profiles, and insurance records—to transform the way that diseases are treated and prevented, and to better engage patients in their own care, according to a news release.

New machine learning technologies and advances in computing power, such as those available through AWS, make it possible to "rapidly translate insights discovered in the lab into treatments and services that could dramatically improve human health," the release states.

Researchers without such tools may store data stored on a local server for training, optimization, and validation, says Hartman. However, he says, these university- or consortium-based resources have limitations.

"By building your infrastructure in best-of-breed, cloud-based computing, those research prototypes can be updated and shared across institutions," Hartman says. In addition, they can easily be deployed in other platforms and software services.

The robust infrastructure enables researchers to develop models that build scalability into the solution from the outset. "That would allow for scale both within the research or academic world, as well as getting into the actual clinical environment," says Hartman.

"Using these cloud resources gives the investigators a lot more flexibility because they're designed with [scalability] in mind," says Zariel Johnson, PhD, program manager-technical for UPMC Enterprises.

The new Amazon resources will impact eight PDHA projects in development including initiatives to:

  • Create an individual risk score for every cancer patient, enabling doctors to better predict the course of a person’s disease and response to treatment
     
  • Use a patient’s verbal and visual cues to diagnose and treat mental health symptoms
     
  • Reduce medical diagnostic errors by mining all data in a patient’s medical record
     
  • Improve the diagnosis and treatment of abdominal aortic aneurysms
     
  • Better understand the origin and evolution of tumor cells

Award Helps Fuel an Even Bigger Plan for Pittsburgh
 

The AWS award is one small element in a much greater plan to position the Pittsburgh area as a "premier healthcare technology hub," says Hartman. With new resources to accelerate the innovation engine of the research alliance, PDHA will serve as a catalyst to bring academic talent and funds to academic innovation at participating institutions, he adds. According to the vision, that dynamic will spawn a growing entrepreneurial ecosystem that will then attract bigger players to the market.

Before the Machine Learning Research Award, UPMC already had a relationship with AWS, contracting with the company for its cloud-based services.

"One of the reasons we decided to work with AWS is that they have a good track record of finding interesting, compelling science that has the ability to have high impact," says Johnson, "That's exactly what we're interested in doing with the Pittsburgh Health Data Alliance, so there are a lot of shared interest and goals."

The AWS award "fits into that bigger story as the flywheel is spinning faster and faster, and these pieces are coming together to drive healthcare technology innovation in Pittsburgh," says Hartman.

Mandy Roth is the innovations editor at HealthLeaders.


Get the latest on healthcare leadership in your inbox.