Over 150 Total Lots Up For Auction at One Location - CA 05/31

PHDA and Amazon Web Services team up in machine learning healthcare sponsorship

por John R. Fischer, Senior Reporter | August 19, 2019
Artificial Intelligence Health IT
AWS and the PHDA plan to integrate
machine learning into various healthcare
segments and solutions
The Pittsburgh Health Data Alliance (PHDA) is partnering with Amazon Web Services (AWS) to apply and integrate machine learning capabilities expected to enhance various areas of healthcare.

Consisting of University of Pittsburgh, University of Pittsburgh Medical Center and Carnegie Mellon University, the PHDA will join the Amazon.com company in a machine learning research sponsorship that is expected to advance care in healthcare segments such as cancer diagnostics, precision medicine, voice-enabled technologies and medical imaging.

“The goal is to develop innovative, data-driven solutions that will improve healthcare. The PHDA provides funding for compelling research projects at the universities and additional support to develop these technologies to power commercial products and services,” Zariel Johnson, program manager at UPMC Enterprises, the innovation and commercialization arm of UPMC, told HCB News. “Many of the most exciting and innovative projects require that research teams build, train, and test machine learning models and otherwise manipulate large, complex data sets. Through this sponsorship with AWS, teams will have access to flexible cloud computing resources and machine learning tools that they can use to accelerate their investigations and scale to meet future needs.”

The PHDA relies on big data, such as EHR information, diagnostic imaging, prescriptions, genomic profiles and insurance records to determine best approaches for treatment and prevention, and to help patients better understand the care they are receiving.

Like many organizations, the PHDA is seeking to integrate machine learning and advanced computing power into healthcare offerings and practices to gain and apply insights that can improve treatments and services for patients. Amazon offers a variety of products for this task, such as Amazon SageMaker, a cloud machine learning platform and Amazon Elastic Compute Cloud (Amazon EC2), a part of Amazon.com’s cloud-computing platform.

New machine learning technologies and advances in computing power, like those offered by Amazon SageMaker and Amazon EC2, are making it possible to rapidly translate insights discovered in the lab into treatments and services that could dramatically improve human health.

Through the AWS Machine Learning Research sponsorship, PHDA researchers from Pitt and CMU plan to advance research and product commercialization for eight projects. One specific example is Pitt’s use of AWS resources to form an objective, predictive tool for guiding surgical interventions of abdominal aortic aneurysms, the 13th-leading cause of death in western countries, before symptoms appear. Another is the development by CMU of algorithms and software tools to better understand the origin of tumor cells and predict how they will likely change and grow in the future.

Other areas of interest include medical imaging, mental health, the creation of individual risk scores for cancer patients, and reducing medical diagnostic errors with data mining.

“Among eight projects that AWS will support, three are developing new technologies to enhance the prognostic and diagnostic capabilities of medical imaging. In all three cases the research teams need to manipulate multimodal data sets, rigorously test their algorithms, and make iterative improvements as new data are incorporated. The AWS tools and environments offer a means to do this quickly and efficiently. Other projects that will benefit from the support provided by AWS are developing tools to advance precision medicine for cancer, identify new biomarkers for mental health assessment, and improve the process of diagnostic coding.”

Data used in the projects will be secure, anonymized and remain with PHDA institutions.

You Must Be Logged In To Post A Comment