Intuitive reveals insights into its AI strategy for surgery

Intuitive revealed insights into its AI strategy

During this years Innovation Week in Turin - Jason Hart (Vice President, Global & US Marketing, and Corporate Communication at Intuitive) took the stage along side Dr Filippo Filicori (Northwell System Chief, Surgical Innovation. Program Director) - gave an impressive talk about how AI and smart surgical robots are coming to operating rooms sooner than you think.

But let me start with the rampage that is the adoption of surgical robotics in the USA. Read this graph folks...

Let me be clear - this is da Vinci adoption. And this trend is undeniable. Da Vinci RAS has now surpoassed both open and laparoscopic surgery in the US market for applicable procedures - and this is what I ranted on about at SRS in 2024. It's not the % of "all surgery" that counts - but the "Applicable" surgery. And boom !

Let me also dig into this graph that Jason showed (From January 2025) but a key reminder of how we go from today's numbers of surgeries to tomorrow's.

In the absolute immeditate "Line of sight" Intutive can go after 8 million procdures today - it is just really slowed by how fast surgeons can uptake and switch. No other barrier. (Hang on could AI help that accelerate? Could AI give them a reason to move to RAS from their current surger? Hold my beer.)

But the 22 million super big pie is where this can all go as technology gets better - pricing per case adjusts - and different offerings of systems become available - we get into Cardiac - endo cardiac etc etc (for another post). But I guess AI will play a vital role in the jump from today's 2 million to 8 million, and then technology - new platforms and AI - to make that leap to the 22 million. AI in da Vinci, AI in GI Ion (my name), AI in Ion cardiac (also my name). etc etc

Let's dig into what they showed on AI

I need to start with a disclaimer or I would not be bein fair to Intuitive - John and Filippo. What you are about to see was shown at the Italian Tech Week - and they were clear that this is NOT in the product today. It is a mock up of what could be done.

But for me - this is a pretty advanced mock up - and absolutely is feasible that they could implement this. I think we will see something like this deployed piece by piece in the next 12-24 months.

I love this mock up video they did - and it shows some key things that I think they will implement. First - top right - you can toggle into the Electronic Medical Records - which will be essential for the AI to get a full understanding of the patient. I think this connection will be one of the things that allows the AI to better understand the surgical scene and surgical outcome in relation to all the data of the patient. And allow da Vinci to write information and store images, video and notes to the EMR.

You can see a case progress bar at the bottom - where it could have prediction of progress for the different surgical phases - and (like Strava) allow you to see your progress against your benchmark or other benchmarks.

But the big thing that should jump out on this screen should be the "Outcome risk" boxes. In real time AI will be assessing the surgeon interactions and matching that with a 17 million patient database - and helping the team understand possible risks. But it will not just advise.

There is now a big reveal about "FORCE FEEDBACK" and I understand now better why they didn't call it "Haptic feedback." - They want force and haptic.

Part of the intervention of the AI will be to "Force limit" movements of instruments. The AI will - in real time - adjust the force an instrument can apply - give resistive sensations back to the user - and minimise the risk of tissue damage as assessed by the AI. That's smart!

Support with structure navigation has been thought about for a long time. With work coming out of ORSI and other institutions. But here we see a mock up of how Intutive might highlight the different tissues and label them. And they have 17 million cases under their belt to use for training data! So I think their AI will be very good at this. As it is the biggest training set out there. That is an advantage over everyone else.

And here you can see as you stack up the AI insights - "This is the tissue" - "We know that the outcome risk is excessive force" - it then hands over agency to the grasper and turns on "Force assist" - and sets an upper limit of what force can be applied to the tissue. And as was said - you can turn this off is you want - but imagine how good this is for the learning curve or a resident. No matter how much force the resident tries to apply the AI limits it based on risks.

And now we get into the meat of the AI. We are heading to an anastamosis using sutures. The AI sees we are in this phase - and up pops a small assistance "Auto suture." And as we have all seen from Johns Hopkins work - this is real.

Auto-suturing is engaged and not only will the system do the needle pass - knot tying - it will indicate entry and exit points. We will see AI make perfect needle passes every time. But allow the surgeon to intervene with their knowhow and at any point make corrective adjustments. This is augmentation.

What they showed in their mock up demonstration was the possible next steps in surgical automation. Which is different than "Autonomy" - as it was clear they see the surgeon still at the heart of the procedure being augmented in decision making and assistance by AI. Is this part of the stepping stones to autonomy? Probably yes - but that is a big leap.

Let's dive a bit deeper on what the surgeon - Dr Filicori discussed.

I found this insight fascinating. The concept was that the AI could use data insights to look for things like stapling positioning for optimal staple line placement. I discussed this in a blog some time ago - but now we are seeing the seeds of this innovation coming to da Vinci. I believe AI will start by sharing insights - but rapidly will move to agentic actions for placing the stapler - asking "to confirm" and then running a smart firing cycle.

Here's where I will speculate a little - so forgive me. They discussed a lot about telesurgery - tele coordination and collaboration. I think AI jumps into the middle of this as well. I felt it was hinted at but not implicitly said. I think that AI will be running at both ends of the telesurgical link - and assisting the surgeon performing the case and the team supporting the case.

I suspect that the issues of distance that are still in some people's minds about "the key surgeon not really being there" can be diminished by applying a bi directional AI supported layer. Think of the AI force limits at distance. Think about the suturing assistance when training a new surgeon at distance.

Then think about getting patient tailored risks at distance because the supporting surgeon will not have physically visited and seen the patient and their history. Perioperatively the AI will do all that risk assessment and feed back the essential insights to the assisting remote surgeon.

And coming back to the bubble chart at the beggining. How do we go from 2 million procedures a year to 8 million to 22 million?

AI across the continuum of the patient journey. Pre to peri to post and all run through assistive AI and agentic AI.

How could anyone justify doing manual laparoscopy or open surgery (except for clinical indication) when the BEST surgery could be guaranteed by having an AI assisted robot in the middle of the care pathway.

I predict just by the errors being stopped - outcomes will improve and thus the health economics will pay for the technology. We are about to see a major shift.

There's a reason we are at 42% of procedures being done RAS in many places - but the deployment of AI will tip that even further in the USA - but will then push it out globally.

With AI assistance and better outcomes it is going to be impossible to ignore the robot and not use it. Clinically and economically it will be a no brainer. AI in combination with robotics will be the edge.

Intuitive's vision (for now) seems to be that surgeons will not be replaced but augmented. (well that was how I watched the video.)

It should be clear to all of you now why da Vinci 5 is at the heart of their future with its powerful compute 10,000X and collaboration with NVIDIA.

It should be clear why force feedback is an important part of this and why "Force Gauge" was added.

It should be clear now why their Intuitive Insights is about "insights" and not data.

It should now make sense about their slow - but thought about - role out of tele collaboration and tele surgery.

It should make sense now why they want smart stapling, smart instruments and smart Vision. AI needs input data - smart peripherals give it.

It should help you understand that rushing to deploy AI will not beat a 360' strategic approach to it: where the robot is built to utilise the AI agentically - the instruments can give precise force information to the AI to act on. The Insights of millions of collected procedures with kinematic data can train the AI. Better vision (on it's way) seeing beyond the human eye - will allow the AI to do more than we can imagine today.

While some have been playing checkers with robotics and AI - Intuitive has been playing chess.

Watch the full video - starts at 17 minutes

These are only opinions of the author for educational purposes only based on what he sees publically and interprets. They are his own opinions. Images and video owned by Italian Tech week.

Doing a medtech startup - involved as an investor ? Get my book my new book - https://mybook.to/medtechsurvivalguide