With clock ticking, Israeli AI start-up slashes stroke treatment time
"A patient not treated in time and left disabled, remaining in a hospital bed, is both a tragedy and a financial burden."
By EYTAN HALONThe Viz.ai stroke triage and notification platform (photo credit: VIZ.AI)
When a stroke occurs, every minute counts. Immediate medical attention is critical to prevent long-term disability or death.In the case of strokes caused by a blood clot in the brain – known as ischemic strokes, or large vessel occlusions – a new standard of care treatment called mechanical thrombectomy has proved to be highly efficient, safe and cost-effective.Despite the emergence of the breakthrough procedure, which involves removing a blood clot from the brain using a stent, as few as one-in-ten eligible patients in the United States actually receive the treatment.The problem is twofold. Even in the world’s leading health systems, few physicians are currently trained to perform the lifesaving stroke intervention treatment, and availability is especially problematic outside of major hospitals.To make matters worse, it can take hours for crucial brain scans of potential stroke victims to be evaluated by the necessary hospital decision-maker, potentially damaging the patient’s chances of recovery.Aiming to ensure the delivery of critical stroke treatment on time is Viz.ai, a Tel Aviv-based start-up using FDA-approved artificial intelligence and deep learning technology to automatically identify suspected blood clots on CT angiogram scans and alert on-call stroke physicians within minutes.“It’s about getting the right doctor to the right patient at the right time, just in time to make a difference,” Viz.ai co-founder and CTO David Golan told The Jerusalem Post.Viz.ai co-founders Dr. Chris Mansi and David Golan (Credit: Viz.ai)“For a stroke, time is of critical importance. The patient arrives at the hospital, and we treat them as fast as possible.”Golan co-founded Viz.ai with neurosurgeon Dr. Chris Mansi in 2016 after meeting during his machine learning post-doctoral fellowship at Stanford Business School, with both eager to help patients with time-sensitive, acute conditions receive treatment faster.
“As we started diving into this world, strokes became the prime candidate for such an endeavor. It’s the most time-critical condition in the whole of medicine,” said Golan.Once a CT angiogram scan is complete, the image is uploaded to Viz.ai’s cloud software, which analyzes whether there has been a stroke within approximately five minutes and, if confirmed, sends an alert to both the treating physician and operation team.The company’s platform also enables clinical teams to coordinate complex care and treatment decisions between hospitals in large hub and spoke networks, which previously led to significant delays.In a recent 300-patient performance study, Viz.ai’s automatic notifications resulted in faster notification of the specialist in 95% of cases, saving an average time of 52 minutes, and even as much as 206 minutes.“Every minute saved can give the patient an additional week of healthy life, so by cutting down the period of time by 52 minutes, you are essentially saving one year of healthy life,” said Golan. “Stroke patient outcomes can be dramatically changed with a system like ours.”The Ziv.ai platform received FDA clearance in February 2018, a groundbreaking first approval for an artificial intelligence-based computer-aided triage and notification system. Today, the technology is commercially available and already installed in approximately 100 American hospitals.The company is backed financially by investors including Google’s GV fund, Kleiner Perkins, former Google CEO Eric Schmidt’s Innovation Endeavors, Yahoo! co-founder Jerry Yang’s AME Cloud Ventures and by Danhua Capital.“The best sales force we have are the key opinion leaders. We have people speaking on podiums talking about how Viz.ai improved outcomes at their hospitals, and they’re really promoting the device,” said Golan.“There is also a huge financial incentive for hospital systems and administrators to treat stroke patients quickly. A patient not treated in time and left disabled, remaining in a hospital bed, is both a tragedy and a financial burden.”Viz.ai is now seeking to replicate its stroke treatment success by accelerating accessibility to healthcare in other acute and longer term scenarios.“For example, in cases of trauma, patients are often taken to the nearest hospital, but that place might not have the necessary physician or surgeon. You do a scan, triage it to the right specialist, who can then take on the patient and ensure he receives fast treatment,” said Golan.“We are also looking beyond acute care, not only when minutes matter, but when days, weeks and months also matter. We’re talking about early detection of cancer and suspected pulmonary embolism, and connecting patients to the right doctor.”
