When it comes to video-based data, advances in computer vision have given the world of research a tremendous boost, making the process of analyzing and extracting insights from moving images something scalable beyond the confines of a small team of humans is.
A startup called theatre has applied this concept to the world of healthcare: it uses AI to “read” videos recorded during surgeries to look for best practices, but also to identify key moments where a surgery may have taken the wrong turn. Today it announces $24 million in funding – a sign of how both the medical world is adapting and using advances in AI to improve their own work; and how investors are increasingly betting on the opportunity ahead.
The funding is a significant addition to Theator’s $15.5 million Series A beginning February 2021, bringing the total for the round to $39.5 million and $42.5 million overall elevated.
As with the previous tranche, Insight Partners led this latest investment. Former contributors Blumberg Capital, Mayo Clinic, NFX, StageOne Ventures, iAngels and former Netflix Chief Product Officer Neil Hunt are also attending, as are new contributors iCON and Ariel Cohen, CEO and co-founder of TripActions.
The valuation isn’t disclosed, but the Series A is notable for another reason: it brings in a major strategic investor in the form of the Mayo Clinic, which partners with Theator — based in Palo Alto and also operating in Israel — to use its Video Analysis Tools. Other partners include the Canadian Association of General Surgeons and others that it is yet to announce. In total, Theator’s library has now amassed 30,000 hours of anonymized video with nearly 1 billion frames analyzed.
The market opportunity that Theator is tackling is this: A vast repository of video is already being created in the surgical world, particularly by the camera probes used in non-invasive procedures.
Of course, the main purpose of this video is to allow surgeons to see what they are doing in real time. But Theator’s premise is that this video – used in an effective way – will be an invaluable resource for those doctors, the care facilities they work in, and potentially the fields in which they work (i.e. the broader network of other doctors working in work in the same areas as they do) if it could be examined and compared to similar procedures carried out elsewhere and then cross-checked with the results.
On a human level, this may sound like an impossible task. There are too many videos and the concept of analyzing even part of them sounds too time consuming to carry out. But all of this means something else: the best results have actually stayed with those who are already doing the best work.
Or, like Dr. Tamir Wolf, the CEO and co-founder of Theator, observed (to paraphrase an age-old saying): “Too often, where you live determines where you live.”
“Today there is no real understanding of the basic truths,” he continued, despite the fact that there are tens of millions of hours of video created through visual guidance for various procedures. “None of these videos are captured, stored or analyzed. You lose understanding of what is going on in the operating room and best practices. Being able to see what best practices look like and then sharing them is our goal.”
And this is where AI comes in.
Wolf describes Theator’s platform as “surgical intelligence”. It takes many hours of footage and can identify key moments in each procedure in real time.
So, given a six-hour pancreas surgery, the system uses machine learning and computer vision to structure the raw footage, compare this video to other videos of the same procedures, and then match what’s happening in the videos to the results of previous procedures to refine the key features of the “good.” Results’ and note where things diverged.
The data is then shared with individual physicians, teams, their facilities, etc. to create a better understanding of existing patients (to better manage follow-up) and future procedures.
Many people tend to focus on the aftercare and the complications that can arise after what is otherwise considered a “successful” procedure, but Dr didn’t have enough data and insight into the surgery itself.
Wolf notes that some hospitals are getting worse results than others for surgeries also labeled “successful” because there were no real-time complications during the actual procedures.
Why is that? “We don’t know,” he said simply.
Theator actually came about from this question, which Wolf asked himself as a doctor, but also as a friend and relative of patients.
In particular, he recalled that both his wife and a friend/colleague happened to have the same surgery at the same time, but in different hospitals. Both went technically fine, but one had a much greater long-term impact than the other. Trying to get to the bottom of why this happened has in part motivated what his startup is pursuing.
“Theator’s technology has proven to be a critical next step in surgical advancement,” said Brad Fiedler, VP at Insight Partners, in a statement. “The integration of AI and computer vision into the operating room improves surgical care and changes surgery for the better. We are excited to double our investment as Theator’s AI and computer vision expertise is now improving patient outcomes with an ever-growing number of commercial partners.”
To date, Theator has negotiated its contracts with healthcare providers – i.e. hospitals and clinics where procedures are performed – although you can imagine a scenario where insurance companies, individual doctors and maybe even patients would want to access this type of data to learn more about it understand what’s going on, and perhaps more importantly – a bit like dash cams – to record what’s going on in case something goes wrong.
It’s something Theator isn’t currently pursuing, but it’s an obvious opportunity.
Likewise, there’s a whole world of procedures that the startup isn’t currently tackling. Wolf described minimally invasive procedures as “low-hanging fruit” in this regard, as these surgeries already use cameras and record video. Over time, there are a number of other, even more complicated procedures that you can imagine could benefit from similar treatment.
At the same time, the market is evolving. Not everyone wants this type of testing or believes that it can give an accurate picture of the total circumstances that go into a single operation or treatment of one person over another. It sort of just focuses on the aspects that the camera can capture.
And one could argue that once the parameters for what ” correct” are specified. It’s basically the same problem you get with other uses of AI when it essentially paints itself into a logical corner that clearly no longer makes sense to the human brain and our actual reasoning, and is in fact no longer “intelligent”. , but the opposite .
Don’t let that distract you from what Theator’s technology has the potential to be. It is just to be remembered that, as with any AI, there is undoubtedly much more that needs to be codified as to how this intelligence can be used in context.
Meanwhile, “We’re slowly seeing a shift in the minds of surgeons and others in this ecosystem that there needs to be more transparency,” Wolf said. “Moving to competency-based insights is part of that.” This potentially leverages this technology not only for operations and best practices for everyone, but also for training. “Video will be at the core of how surgeons are evaluated to see if they can get out of residency and into full practice.”