CEO: Fabien Beckers
Based: San Francisco
Arterys is working on a cloud platform to apply deep learning to the analysis of medical images. Its ultimate goal? To make image analysis an automated, consistent and data-driven process, and to do it on a global scale. Using its cloud-based software, the company seeks to bring together information and expertise that has historically been siloed off at individual hospitals, allowing each physician to benefit from the work of thousands of other physicians.
“There is user variability in medical image assessment,” said CEO Fabien Beckers. “Each individual has a different way. What we’re trying to do is aggregate all those methods to normalize and quantify the result.”
When imaging is done, for example an MRI or CT scan, it is sent to the cloud, where the physician may log in and view the images and analysis. The Arterys software runs on a web browser, so individual hospitals need not install any special equipment.
The FDA cleared the Arterys software last fall. It works with MRI scanners to visualize and quantify images of the heart. Its 4D Flow component allows physicians to noninvasively assess blood flow to and around the heart and can help them make treatment decisions. And just this month, the FDA cleared its Cardio DL software, which automates ventricular segmentations from MRI scans. This dramatically reduces the time taken to obtain an image of the heart as radiologists have traditionally used software to manually segment the ventricle.
What makes Arterys fierce
Armed with the cloud, artificial intelligence and big data, Arterys is on a mission to transform healthcare. The system becomes smarter over time—as more and more physicians use it, it builds up a repository of images and interpretations that can shed light on patterns over different kinds of patients, Beckers said. While this aids the automation of image analysis—which expedites physician decision-making and response in time-sensitive situations—it can also improve our understanding of patients, their disease and how different patient groups respond to treatments.
One reason that hospitals have been reluctant to connect to the cloud is the fear that patient data could be hacked or breached, Beckers said. Arterys deals with this by extracting a patient’s private health information from a scan before it is sent to the cloud. The company never sees the patient’s identify or information, which remains at the hospital level.
What to look for
In December 2015, the company raised $7 million in a Series A round—which it later bumped up to $12 million—and announced a partnership with GE Healthcare that will include the Arterys software in new GE MRI scanners. The solution, dubbed ViosWorks, will deliver three-dimensional cardiac anatomy, function and blood flow in a 10-minute scan along with the cloud-based, real-time processing of high-resolution images. They plan to roll out the system this year.
While Beckers said Arterys is open to partnerships, it’s currently focusing on its tie-up with GE. The company is eyeing regulatory approval in the U.S., Europe and Asia and considering other areas where its software may be useful.
“We want to apply [our technology] to many other things, like oncology and other disease areas where we think the combination of cloud AI and data can transform healthcare,” Beckers said. — Amirah Al Idrus, @FierceBiotech