Kibbutz Shefayim, Israel-based deep learning startup Zebra Medical Vision announced yesterday that it has received an 510(k) clearance for an artificial intelligence tool that can read medical imaging to identify potential cases of pneumothorax — a build-up of gas between a lung and the chest wall.
The HealthPNX product scans a chest X-ray or digital radiography scan for signs of the condition, and then flags the image for review by the clinical team.
The new tool will be a part of the Zebra Medical Vision’s AI1 (All in One) integrated offering, which eases the radiologist’s workload by highlighting potential brain, lung, liver, cardiovascular and bone diseases from imaging. According to the company, its newest tool was part of a deep learning research project that used millions of medical images to train a chest X-ray AI network to identify more than 40 common clinical findings.
WHAT’S THE IMPACT
As of a 2014 study, incorrect or late diagnosis of pneumothorax impact roughly 74,000 American patients annually. Much like other AI-powered image scanning tools, deployment of HealthPNX could help focus a radiologist’s attention on high-risk cases and catch some of those that are falling through the cracks.
WHAT’S THE TREND
It’s been less than a year since the Israeli startup raised $30 million and earned its first FDA clearance for an algorithm that uses CT scans to quantify patients’ coronary artery calcification. Zebra’s products join the ever-growing market of FDA-cleared algorithms, of which many such as Viz.ai’s Contact and IDx’s IDx-DR are also focused on flagging potential cases for radiologists.
ON THE RECORD
“We are happy to add this important capability to our All-in-One (AI1) package and add more value to busy radiology departments,” Eyal Gura, Zebra-Med’s CEO and cofounder, said in a statement. “Health providers across the US that already use the many Zebra-integrated [picture archiving and communications systems] and worklist systems, will be able to easily deploy our solution and quickly realize ROI and improved outcomes.”