The past week has seen two FDA clearances awarded to Israeli companies building artificial intelligence analysis tools for medical imaging.
The first, announced last Tuesday, was Aidoc’s tool for triaging cervical spine fractures. Much like the company’s prior offerings, the new product analyzes images to prioritize high-risk cases for review by a radiologist, and does so within the specialist’s existing clinical workflow.
“Our C-spine solution tested the limits of what we could do because of the difficulty in detecting spine fractures due to the variability by which they present,” Michael Braginsky, Aidoc cofounder and CTO, said in a statement. “We are proud that we were able to release such an accurate solution that is already showing value in clinical practice.”
The other, cleared on Thursday and announced today, is Zebra Medical Vision’s AI triage tool for intracranial hemorrhage. Called HealthICH, the software product reviews non-contrast head CT scans to highlight potential bleeds for radiologist review. According to the company, the tailor-made neural network is capable of generating these alerts in spite of smaller bleed sizes and more common artifacts, such as metal, that are observed within a patient’s brain.
“We are thrilled to get a 3rd nod from the FDA in such a short time and to be able to add such an important new finding to our growing AI1 ‘all-in-one’ package,” Zebra CEO Eyal Gura said in a statement.
WHY IT MATTERS
Radiology and imaging analysis is often cited as among the strongest medical use cases for AI. With specialists often tasked with reviewing dozens of scans at a time, these kinds of tools aim to reduce their burden, direct their attention to the cases in greatest need and help patients receive their care more promptly.
The number of such tools being cleared and entering the market has greatly accelerated over the past year. Aidoc and Zebra are at the front of this arms race, and both companies have adopted a suite-based strategy in which a variety of software scans built around a similar core technology can be purchased within a single suite of workflow-friendly products. This approach is not just more convenient for the hospital, both companies have said, but is also responsible for an increasing pace of clearances that is unlikely to slow down any time soon.
“One thing that’s important in this is that we understood that a hospital can’t do 12 different point solutions. What we need is a comprehensive set of tools that would cover them in the background, constantly, across different anatomical regions,” Aidoc CEO Elad Walach told MobiHealthNews around the time of the company’s last FDA clearance. “If you look two or three years back, you’ll see that a lot of the AI companies are focused on creating algorithms. If you look now it’s much more about solutions, … and what can be achieved and what that means for patients. What that means is that in the last year, radiologists have seen more and more evidence about the value of AI and the fact that it can change the standard of care.”
THE LARGER TREND
Zebra Medical Vision raised $30 million just over a year ago, and announced its first clearance not long after that. Aidoc’s latest round came more recently, when the company announced a $27 million Series B in April. Both companies’ clearances were preceded by Viz.ai and IDx, which each earned de novo clearances early last year for their AI-based detection software products.
