The University of California at San Francisco's Center for Digital Health Innovation and GE Healthcare are developing tools to help clinicians make faster and more accurate diagnoses of specific conditions.
GE, which makes radiological imaging equipment and electronic health records systems, is working with UCSF to develop an entire library of computerized calculations, or algorithms, that will be combined with EHR data in an effort to more quickly read images and help develop patient care plans.
An algorithm already under development can help spot pneumothorax, or collapsed lung. But the long-term goal is to create a clinical computing system to predict a patient's trajectory, automate their triage and otherwise enhance workflows to be more efficient, according to GE.
The algorithms will be stored and made accessible worldwide via GE's Health Cloud database and work with so-called “smart” GE imaging machines, according to a UCSF news release.
“Together, we will develop tools and algorithms that will allow clinicians and researchers to identify problems and ask questions that are only achievable with vast computing power and datasets,” said Dr. Michael Blum, a veteran physician informaticist. Blum is director of the innovation center at UCSF and associate vice chancellor for informatics.
Financial terms of the agreement were undisclosed.
Since the 1960s, clinicians and computer scientists have been trying to harness the power of computers to help with the complex problem of differential diagnosis, yielding diagnostic support tools such as Dr. Lawrence Weed's Problem-Knowledge Coupler and Dr. G. Octo Barnett's DXplain in the 1980s. They've been augmented by a number of more recent tools such as Isabel, developed initially for pediatrics, and VisualDx, initially for dermatology, but more recently adapted to diseases of the body such as Ebola.
The GE and UCSF collaboration will face stiff competition.
Last year, IBM acquired Chicago-based Merge Healthcare, a developer and provider of imaging exchange software and services for $1 billion. The goal was to hook up Merge clients with IBM's Watson database and clinical computing and artificial intelligence platform.
IBM has been working for more than a decade on image recognition for a variety of industries, but launched the IBM Watson Health Group a year ago to put Watson's brain power to work in the healthcare industry.
In June, Merge and IBM announced a “medical imaging collaborative” of 15 medial imaging technology companies and healthcare providers to create AI support for a host of medical conditions, including cancer and brain, heart lung and eye disease.
The first commercial offering from the Merge/Watson initiative – focused on diagnosing aortic stenosis, a narrowing of the heart's aortic valve, could be released in the first quarter this year, said Michael Klozotsky, marketing leader for Watson Health Imaging. It and other AI projects will be on display at the upcoming annual meeting of the Radiological Society of North America Nov. 27-Dec. 1 in Chicago, he said.
