Viz.ai raises $21M to better detect the early signs of a stroke

The company will be broadening out its AI platform to detect other medical conditions as well

When it comes to certain life threatening diseases, time is absolutely crucial. That's especially true when it comes to someone having an episode such as a heart attack or a stroke; the time it takes to get a hospital, and then wait for a procedure, can literally mean life or death. 

What if software could detect these problems earlier, potentially allowing patients to get the treatment they need in a shorter amount of time? That's the idea behind Viz.ai, an artificial intelligence healthcare company focused on detecting time sensitive conditions. 

On Wednesday, the company announced that raised $21 million in Series A funding led by Kleiner Perkins, with participation from GV, formerly Google Ventures, bringing its total funding to $28.5 million.

Founded in 2016 by Dr. Chris Mansi and Dr. David Golan, Viz.ai's flagship product is called Viz LVO. It's an AI cloud-based platform that analyzes brain scans for large vessel occlusions (LVOs); those scans are then sent to a neurovascular specialist to facilitate early therapeutic intervention.

"What we developed is a system that connects a CT scanner in any hospital in a hospital network with the phone of the neurosurgical and neurological specialist who’s on call for strokes. I’ve got it on my phone and I can be interrupted at any point with an alarm, and it will say that, in this hospital, there is a patient who is having scan right now and needs my attention," Mansi said in an interview. 

The initial idea for Viz came from when Mansi was a neurosurgeon at Stanford in 2014. He operated on a woman who had been hit by a car and though the surgery had been a success she died hours later. It was determined that the fact that it had taken four hours to set her to the hospital and into the operating room. That incident spurred him to create the technology to reduce that waiting time. 

"Prior to this, we didn’t really have great treatment for these worst kind of strokes. The vast majority of stoke patients will end up in a nursing home, paralyzed on one side of their body, often not able to speak. In fact, that happened to my own grandmother. It’s an incredibly sad way to end your days. The difference with this treatment is patients who are treated fast enough will walk out of hospital and they’ll be able to function," Mansi told me in an interview.

"The big difference from the situation previous is that in order for that patient to come to one of colleagues who’s on call to care, it would take a hugely inefficient process, hugely inefficient workflow. There would be lots of different doctors involved, the scan would need to be read, often by a non-neurological expert. The emergency physician would have to get through to another hospital, and it culminated in many hours passing, typically about three to five hours, before that patient was treated, which, when you contrast to the fact that every one minute two million brain cells die, is too long."

In February of this year, the U.S. Food and Drug Administration (FDA) granted a De Novo clearance for Viz LVO, the first-ever computer-aided triage and notification platform. Viz.ai also recently announced its second FDA clearance for Viz CTP through the 510(k) pathway.

"We have about 500 hospitals in the United States that can treat this kind of stroke, and you have about 5,000 hospitals, so that’s one in every 10 can treat it. Hospitals are normally arranged into a hub and spoke network, so if you’re in the hub, you’re a patient and you happen to go to a local hospital in the East Bay, that would not be a specialist hospital. It would not be a hub. You would have to be transferred to Stanford or UCSF, so the on call doctors in the Stanford or UCSF would get a notification on their phone, they’d be able to view the images on their phone, they’d be able to communicate with the doctors in your local hospital, and then they’d get workflow tools to get the ambulance there much faster," Mansi explained. 

In its FDA trial, which involved 300 patients, Viz compared the scan to specialist notification time, which is the time that the neurological specialists was made aware of the case, and in 95.5 percent of cases Viz was able to save time, on average 52 minutes, though the amount could swing wildly, from six minutes saved to 206 minutes saved, depending on the day of the week and time of day. 

The company will use the funding for to expand to new markets. Right now, Viz is in around 10 hospitals across the US, mostly in the southeast part of the country, in states like Tennessee, Georgia and Alabama, where there is a disproportionate number of strokes, leading the area to be known as the "Stroke Belt." It is looking to eventually expand to parts of Europe and Asia as well.

"The issue in parts of Europe, including the UKNHS, is that this is only useful if you have the doctors who can actually do the eventual treatment. We don’t remove the clot; in fact, we don’t make a diagnoses. We assist the doctor. We’re very much an assistant tool for the doctors to make their lives easier and move patients around faster. When you look at the NHS, the number of doctors who are able to do mechanical thrombectomies needs to improve. It's a very similar situation in Asia as well," said Mansi.

The money will also be used to extend their product portfolio to other medical conditions beyond only strokes, though he could not tell me what those might be at this time.

"Generally it’s going to be things, we will only do things that have real clinical impact, so anything that’s time sensitive and that we actually make a difference to, we’re very interested in. Anything that the doctor’s are doing to improve things right now, they’re doing a good job and don’t need the technology. Things that are time sensitive typically involve either the brain or blood vessels or cancers and there are few things that we’re working on we’ll release in the next six or so months."

In addition to the funding, it was also revealed that Mamoon Hamid, General Partner at Kleiner Perkins, will join the Viz.ai board of directors.

"Mamoon is considered, and was describes to me, as the best SaaS investor out there, and when you think about our business it’s a SaaS business in healthcare," Mansi told me. 

"It’s not just that; I did a lot of diligence if you look at the companies he’s invested in like Slack, Box, a lot of enterprise software businesses that have become big unicorns, all the CEOs have expressed to me something I’ve seen, which is he’s a phenomenal listener and is able to really support you in the tough times, as well as the good times, to steer you along the right track."

The ultimate goal of Viz is to take healthcare that is "reactive" and turn it into something "proactive."

"If a patient is having a scan in the area that I cover, I want to know about that patient as soon as possible, no matter what the disease. The way that we are using deep learning, reinforcement learning, all the technologies that the wonderful companies like Google and Facebook are really pushing forward, is to, as quickly as possible, identify a specific disease and use that to trigger a much faster, more efficient workflow. That proactive healthcare is one that is much better for patent and much better for doctors because they have much better outcomes," said Mansi.

"We want to offer this for as many life threatening, time sensitive conditions as there are out there because it’s only by doing that that you really have a transformative effect on healthcare. If you look at this space it’s a bunch of other companies in it, but I think Kleiner and GV are really the top tier investors in this space, and I think the thing that convinced them was this holistic approach on a relatively simple idea. You’re able to curate data about patients so the right doctor knows about the right patient at the right time."

(Photo credit: Viz.ai)