GE Healthcare and the University of California, San Francisco, have announced a partnership to create a library of machine and deep learning algorithms to help guide treatment for patients with a range of complex symptoms and conditions. The algorithms will be deployed globally through the GE Health Cloud and smart GE imaging machines.

"This enables physicians to spend less time reviewing the normal, so they can devote more time and deliver more rapid treatment to patients with potentially life-threatening conditions," Charles Koontz, chief digital officer at GE, told HCB News.

The partnership will initially focus on high-volume, high-impact imaging to develop algorithms that distinguish between a normal result and one that requires follow up or acute intervention.
One of the algorithms under development is for pneumothorax, which is a collapsed lung. It will rapidly sort through images and help prioritize those that potentially indicate abnormalities, said Koontz.

As the library of algorithms grows it will be able to predict patient trajectories, automate the triage of routine care, improve process efficiency and foster the development of more personalized therapies.

In addition to integrating data from imaging technologies, GE and UCSF will also incorporate clinical data sets from EHRs and other sources.

Koontz believes that health care technology like this will eventually become the standard, as computational power and deep learning become more widely recognized and the benefits are more clearly illustrated.

"The opportunity lies in proving that digital health is more than just hype — that it can improve financial and operational outcomes for clinicians and health for patients," he added.

Business Affairs Homepage