Ferrum, a startup developing an AI patient safety platform to prevent medical errors, today announced that it secured $9 million in seed funding. According to CEO Pelu Tran, who’s also a technology pioneer at the World Economic Forum, the plan is to use the capital infusion to refine the company’s product as Ferrum makes engineering, sales, and marketing hires.

That could be good news for hospitals and health care systems, where medical mistakes sometimes occur. It’s estimated that preventable errors harm more than 1 in 10 patients, with at least 12% causing permanent disability or death, according to a report published in the journal BMJ.

“Ferrum is pioneering the use of AI in a unique way by augmenting patient safety, starting in radiology but looking to expand to other areas of medicine too,” Tran told VentureBeat via email. “In the U.S., only 2% of radiology scans are peer-reviewed to check for missed diagnosis or diagnostic errors, and of that subset, less than 1% are found to have errors, meaning that most medical errors are missed because we’re choosing studies at random. The problem is that it just isn’t feasible to have two radiologists review every single scan, although that would be safer. What we do is enable hospitals to review 100% of scans for potential errors with the help of AI, which then enables human radiologists to spend their time peer-reviewing the studies where there’s actually likely to be a missed significant finding.”

Ferrum provides a HIPAA-compliant evaluation system that runs computer vision and natural language processing algorithms on patient data to spot gaps in health coverage. By scanning radiology images and associated reports for discrepancies, correlating the findings with physician notes, and flagging those discrepancies for secondary review, its tools identify high-risk patients for screening and intervention, and track — or escalate, if necessary — unscheduled or missed follow-up appointments.

Ferrum’s platform can be deployed to an on-premises server or cloud environment and integrated with existing systems from GE, Carestream, Vital Images, Agfa, and others. Health systems using its platform have spotted thousands of patients with “significant findings” that might have otherwise gone unnoticed, the company says. For example, CGMH, one of the largest hospitals in the world, captured 35 times more missed findings using Ferrum’s suite than it did with expert peer review.

Beyond CGHM, Ferrum’s platform is live in the U.S., Singapore, Taiwan, New Zealand, and Australia at dozens of hospital systems that collectively serve 12 million patients. Perhaps encouraged by this momentum, last month, nonprofit health care system Sutter Health said it would pilot the product across its Sacramento, California facilities. (In the first 90 days of the Sutter pilot, Ferrum said it reviewed 10,000 CT scans containing lung tissue and found 83 lung nodules in the initial batch that warranted follow-up appointments.)

“We’re essentially a ‘second set of eyes’ for hospitals, which creates a better environment for patient safety,” continued Tran. “This kind of technologically augmented safety is extremely widespread in nearly every industry, such as cybersecurity, manufacturing, financial services, aerospace, etc. Health care shouldn’t be any different, and this kind of change is inevitable … As an industry, we want to be better in whatever way we can for the sake of improving patient care and quality, and the use of technologies like AI will help us do that.”

Blumberg Capital led this latest round in San Francisco-based Ferrum, which saw participation from GSR, Vulcan Capital, Valley Capital Partners, and an undisclosed strategic family office. As a part of it, Blumberg Capital CIO advisory chair Patrick Steel will join Ferrum Health’s board of directors.

The global AI in medicine market was valued at $719 million in 2017, according to Allied Markets Research, with startups like Ferrum and rival MedAware leading the charge. Health Data Analytics Institute‘s (HDAI) AI-powered platform that analyzes over a billion patient encounters to improve health care outcomes recently emerged from stealth. Spring Health raised $22 million to expand its AI service that matches users with a mental health treatment plan. Clew Medical launched a predictive analytics platform to help prevent life-threatening complications for patients. Charlotte, North Carolina- and Santa Clara-based LeanTaaS snagged $40 million for AI-imbued solutions that ostensibly optimize health clinic operations. And KenSci nabbed $22 million for its prediction product that helps practitioners cut costs intelligently by identifying contributing clinical and financial factors.

Sign up for Funding Weekly to start your week with VB's top funding stories.