 |
| Elad Benjamin, Zebra CEO |
The latest in a long lineup of efforts to introduce machine learning into medical imaging is a deal between Zebra Medical Vision and Intermountain Healthcare. The not-for-profit health system based in Salt Lake City includes 22 hospitals and more than 185 clinics; it expects that the deal will assist its radiologists with automated diagnostic algorithms.
As part of the deal, Intermountain is leading a $12 million financing in Zebra Medical with participation from existing investors. Its investors include Khosla Ventures and Marc Benioff, founder and CEO of Salesforce ($CRM). The Israeli company was co-founded in 2014 by Elad Benjamin and Eyal Gura.
Zebra already has algorithms for use in bone health, cardiovascular analysis, and liver and lung indications. The company is partnered with Dell Services, which is a cloud-based imaging storage providers with more than 1,100 clients, including Intermountain. The pair announced a deal to provide machine learning algorithms for medical images earlier this year.
"We are excited by the opportunities that machine learning and computer vision algorithms can provide. These tools will help us improve patient care, by analyzing imaging data at a large scale for the first time, in addition to textual data," Intermountain CFO Bert Zimmerli said in a statement. "When we researched this field and the various technologies available, the Zebra platform approach stood out in its proven ability to digest millions of imaging files and create new algorithms rapidly."
Zebra's platform has been validated using hundreds of thousands of cases. In bone health, its algorithm uses existing CT scans and offers results that it says are equivalent to the Bone Density T-Score generated by dual-energy x-ray absorptiometry (DEXA) scans--but without the additional tests or radiation.
In emphysema, Zebra's algorithm analyzes chest CTs to detect emphysematous regions in the lungs and quantify the volume of emphysema in comparison to the overall lung volume. And its fatty liver algorithm analyzes CT data to automatically segment the liver and calculate its density.
"We are privileged that one of the top healthcare systems in the U.S. has placed such confidence in our team and our platform," said Zebra's Benjamin. "In an environment where computing power and machine learning frameworks are becoming a commodity, the ability to quickly and efficiently curate large quantities of data from a world class integrated healthcare provider can make the difference between simplistic tools and insights that can truly add clinical value and positively impact patient care."
The giant conglomerates that dominate medical imaging--Siemens ($SIE), Philips ($PHG) and GE ($GE)--are all working to sort out how to maintain their scale, and more importantly grow, in the rapidly changing segment. They are starting to work to offer customers cloud-based image management, as well as the next step of the application of machine learning to images that's expected to make it easier for radiologists and physicians to interpret and contextualize imaging data--eventually, alongside the ongoing influx of massive amounts of genomic data. In addition, this effort is attracting big data players IBM ($IBM) and Alphabet ($GOOG), as well as a slew of startups.
- here is the release