How Proximie’s computer vision and gen AI capabilities are helping to shape smarter, more connected, and data-driven surgical workflows.
Operating Rooms (OR) are some of the busiest, most complex environments in healthcare. Every moment, from wheels in to wheels out, carries valuable operational insights. But much of that data is either missed, manually entered, or inaccurate. In one pilot study from Proximie, manually collected data from the EHR was inaccurate by 59%. That’s not just a documentation issue, it’s a lost opportunity to improve surgical care.
This is where Proximie comes in. Using ambient computer vision, we capture accurate, high-fidelity data from the OR without disrupting the surgical flow. But we don’t stop there. Proximie connects people, data, and devices across the surgical ecosystem integrating with systems like EHR to create a unified, intelligent view of the OR. AI then transforms that data into real-time insights helping healthcare teams unlock extra capacity, streamline workflows, and ultimately treat more patients. AI makes the data intelligent and actionable.
And it’s already delivering impact. In our latest whitepaper, which analysed three major healthcare systems in the US and Europe, we found a consistent, global, universal opportunity for efficiency in the OR of19-25% throughout the three health systems, the equivalent of adding one additional surgery per day, per OR, without extra resources. That’s 9,000 more procedures a year, and $90 million in additional revenue.
Here are five things you need to know about how AI is already shaping the modern OR, and how Proximie is making it work for surgical teams today, not tomorrow.
1. Proximie captures the complete picture of your OR, because we’ve seen it all.
At the heart of Proximie’s AI is computer vision: the ability to capture and interpret video footage from the OR and understand what’s happening. A real-time view of whether a patient is present, when a procedure has started or ended, or when a trolley is wheeled in for the start of a surgical procedure. But this doesn’t happen magically. Proximie’s AI models learn by seeing. The more examples, the better the model becomes. Ultimately, the better the modelling, the more we can help teams become more productive and more cohesive. And with over 100,000 hours of individual surgical data on Proximie’s archive, captured from ORs in over 50 countries, we have seen it all.
Proximie’s AI capabilities have learned from a wide range of real-world scenarios, across every single surgical specialty and via usage from over 40 of the world’s leading medical device companies, so it can tell the difference between what’s relevant in the OR and what’s just coincidence. For our customers, this tapestry of layered AI insight, expertise and intelligence is revealed via a series of real-time contextual dashboards within our Intelligence Suite. This gives teams a complete, system-wide picture of how their surgical days are unfolding, across multiple teams and multiple ORs.
2. It’s not just about learning surgery, it’s about understanding your OR
No two operating rooms are exactly alike. That’s why Proximie’s unique technology stack and AI functionalities have not only been built on thousands of hours of surgical footage from other ORs, but also on footage from your OR.
Once Proximie has been installed in your OR, we enter a control period. During this time, Proximie’s computer vision and AI observes real activity in your specific room to label what's happening. This helps Proximie and the AI models understand the unique way that your team works within any OR; from door usage and screen positions, to lighting quirks and team behaviours.
It means the activity in any given OR moves beyond descriptive and subjective views of what the sources of inefficiencies are, toa single source of truth that shows where your system’s pain points are, and how they can be acted upon. With Proximie, we can help teams move beyond disconnected, contextless data points to a more holistic, integrated approach that captures a complete and objective picture of what is actually happening.
3. How we maximise AI’s potential, and minimise ‘hallucinations’
Just like humans, AI can sometimes be thrown off by unexpected scenarios - a left-handed surgeon in a room that’s only seen right-handed ones, or by bulky new equipment changing the layout entirely. This can sometimes lead to AI ‘hallucinations’, but in the OR, where the stakes couldn’t be higher, we know we can’t afford a misstep.
Our systems are designed to resist these moments of confusion by fractionally delaying judgement, (and always obscuring critical data at source), and checking for realistic timings, changes to OR setup, and referencing historical data to make better decisions. This enables Proximie to paint an objective view of your OR and a foundation for continuous optimisation.
4. We’re always improving how we apply AI in surgery
AI isn’t a one-and-done process, that’s why Proximie continues to garner unique OR knowledge every single day. Insight and data that is ultimately fed back into the system from which it came to improve accuracy.
This continuous improvement is a key reason we’re confident in the long-term value AI can bring to our partners and their teams. Proximie is not just supporting what happens in the OR today, we are ultimately evolving alongside how surgery is delivered tomorrow.
5. AI works best when we’re honest about its limits
No AI system in the world can claim perfection, and that’s especially true in high-stakes environments like the OR. That's why Proximie is built on more than just algorithms. It’s built on transparency, partnership with clinical and operational teams, and continuous improvement.
We know AI can’t solve everything. But it can do something powerful: surface the right insights, at the right time, to help surgical teams make better decisions. That’s why we work closely with teams during periods of change, continuously reviewing how our models perform and where they can do better.
This commitment to real-world impact is already making a difference. In the UK alone, 135,000 on-the-day cancellations costs the NHS£400 million annually - and 80% of these cancelled cases could have been avoided.
By using AI-enabled data analytics to understand exactly what is going on in an OR - instrument use, room organisation, and other potential bottlenecks - surgeries can be performed more effectively. And in away that creates more scalable and equitable healthcare systems.
AI that evolves with every case, learns from every setup, and adapts to the realities of the OR is how we move beyond the buzzword. It stops being a gimmick and becomes a conduit, an enabler, and an integral part of the surgical team. This isn’t about technology for technology’s sake. It’s about putting actionable insights into the hands of skilled clinicians and teams, helping them reclaim time, streamline workflows, and ultimately deliver better care to more patients.
At Proximie, we believe AI in surgery should feel less like an add-on and more like second nature, seamlessly woven into the fabric of the OR, continuously improving, and always working alongside the people who save lives. That’s how AI moves from hype to habit, and from promise to practice by making a real and tangible impact. Every single day.
To find out more about Proximie’s Intelligence Suite, click here.
