Digital Surgery BTK (1)
===

Speaker: [00:00:00] So before we get started, I wanna be open about my financial disclosures. My name is Jake Greenberg, and I consult and receive research support from both Medtronic and J&J MedTech.

Speaker 2: My name is James Jung. I consult for Intuitive Surgery. Today on Behind the Knife, we're talking about cameras, sensors, and software that are turning the operating room into digital data.

Not just hit record and store the video somewhere, but getting the structure and actionable information that could tell you how long you spent dissecting, how often your camera came out of the abdomen, and whether your team actually debriefed after the case. I'm Dr. James Jung, a minimally invasive surgeon at Duke University, and I'm joined today by my colleague, Dr.

Jake Greenberg, our vice chair of education, our fellow Dr. Zachary Weisner, and one of our research residents, Dr. George Lev, to discuss what's new and what's next in digital surgery

Speaker 3: Thanks, Dr. Jiang. This is an exciting topic, [00:01:00] but it can also be a bit of an expansive one. How should we think about digital surgery, Dr.

Greenberg? Can you help us describe the emerging field?

Speaker: Yeah, I mean, I think it's a, it's a new concept to many people, and it's a, as you mentioned, a pretty expansive one. And I think the definition that I most align with is one that was put forth in a SAGES white paper, which was published in Surgical Endoscopy in 2024.

And they basically define digital surgery as anything that inserts a computer interface in between the surgeon and the patient. And when you think about it that way, you realize it involves multiple different branches, right? We're talking about enhanced instrumentation, advanced visualization, robotic surgical platforms, of which there are now an increasing number, tools for both data capture and data analytics, and then connectivity between all of these things and from surgeon to surgeon, potentially in different operating rooms or maybe even different hospitals.

So it's a really exciting time in surgery because all of these instruments and all of these new tools are continuing to advance at a very rapid rate. [00:02:00]

Speaker 3: Awesome. What's really interesting about these platforms is that they're starting to accrue some impressive results in the field of surgical education and OR efficiency, specifically in terms of facilitating self-assessment, expert assessment, efficiency, et cetera.

High-performing athletes spend a lot of time reviewing pre- and post-game footage. But in surgery, we're still somewhat struggling to have accessible, HIPAA-compliant footage to review and have not yet optimized how to get the most out of it the way that people have in other fields.

Speaker 4: Yeah, and what's interesting too is, you know, we're not seeing a ton of literature on any individual tool just yet, but we are seeing individual use cases baked into studies on other subjects.

So for example, this paper came out, validity of video-based general and procedure-specific self-assessment tools for surgical trainees in laparoscopic cholecystectomy, which is a bit of a mouthful. But essentially, they used THETA to securely catalog resident operations and segment cases into easily reviewed steps so that we can ideally train ourselves to evaluate ourselves [00:03:00] using a guided framework.

So even though they were assessing specific frameworks, et cetera- It's interesting to now see these use cases embedded in studies on other subjects.

Speaker 2: Yeah, certainly in one area where digital surgery platforms such as Theator, for instance, in this study has been kind of useful or there has been some emerging evidence to show a use case for in, in surgical education.

You know, we've always been fascinated by how do we utilize the video data to train the next generations of surgeons. And there's a lot of efforts from various different digital surgery platforms to be able to use the video data to help us train our next generation surgeon better, faster, and with better data points.

So in this particular study, it really hasn't shown any impressive kind of results. For example, the associations between the residents' performance and the attendance performance were quite poor, [00:04:00] and whether that's because of the validity of the measurement tools themselves or the use of the video is still yet to be explored.

But, you know, there are emerging evidence to support that, and this is feasible to use the video derived from what's happening in the OR and use them for surgical education.

Speaker 3: Yeah. And on their website, they tout some specific hospital utilization benefits, too. In one interesting study, they analyzed out-of-body time or idle operative time in 649 gynecologic procedures performed at a major medical center in Israel and basically found that outdated insufflators meant more time cleaning the camera, which then in turn meant lost OR dollars, justifying purchase of new insufflators in order to decrease OR time and save the hospital money while improving operative efficiency.

I think this is really interesting because we often don't think about how OR technology can improve surgical quality in this broad of a sense. This study really highlights how broad the opportunities are to use digital surgery to improve the [00:05:00] efficiency of what we do.

Speaker: Yeah, I, I think it can prove a variety of different domains.

Efficiency is certainly one of them. Cost is another. I mean, I, I bet if you use technology like this to look at how many instruments I actually use in a major laparotomy tray or from my laparoscopic trays, I probably use, like, four or five, and yet we're cleaning, like, 100, right? So SPD is working harder.

The nurses have to count everything that we're never ever gonna use. And honestly, if we pared down our kits to bare bones minimum, we'd probably be saving cost and efficiency, too. And I think digital tools could identify what actually gets used, how frequently it's used in every case, and potentially find other efficiencies outside of just operating room time.

Speaker 2: Yeah, absolutely. Obviously, digital surgery is beyond just video, but really the video gives us the ground truth on what's happening. You know, we could really see it. You can't argue with it, right? You can't argue against it. So it, it really gives us an opportunity to review things that we just did, you [00:06:00] know, just because we were told that this is the way to do it.

So in Dr. Greenberg's perspective, that video footage could demonstrate that we're opening up a lot more instruments than we actually use, and we could actually prove that by using these digital tools. So I think from the efficiency perspective, there has been a lot of use cases for these digital surgery tools.

And really just to briefly mention what's out there in the market, the big players who are attached to the robotic platforms and their current robotic platforms with Intuitive and their hub system, uh, and then their emerging robotic platforms, including Medtronics, J&J, and Moon Surgical and others that have their own kind of digital surgery platforms package, such as touch surgery and Medtronics and so forth.

And then there are the digital surgery platforms that are not associated with the surgical towers or the robotic platforms such as Theodore, the study that was referenced, and the OR Black Box. [00:07:00] There are some differences, uh, such that, you know, in the OR Black Box it captures not only the in light view, but also the room views as well.

So if you're thinking about efficiency, you know, that could be useful to understand the whole room state. But just to say that they're emerging and heterogeneous set of digital surgery ecosystems that are becoming available for us to use for efficiency gains. So Jake, at Duke we've implemented the OR Black Box system, and as vice chair of education, we've been utilizing it for education purposes.

Uh, I wonder if you could talk a bit about your perspective on utilizing the Black Box platform for surgical education.

Speaker: Yeah, I mean, I, I think we've used it in a variety of different ways. The platform records not only the surgical video feed, but it also records a variety of other data points, such as the vital signs from the anesthesia machine, as well as all of the audio and video in a de-identified way from the OR itself.

So you [00:08:00] get to see team dynamics, team interactions, interpersonal communication. Um, and we've used it for a multidisciplinary teaching conference, uh, which we entitled Better Together. I know UT Southwestern was, I think, the first to do something like this too, and I can't recall what they titled their conference, but they were really kind of the innovators around this space.

And we've been using it to get together with all of periop, so nursing, anesthesia, our scrub techs, the surgeons, the students, the residents, and we look at intraoperative videos. We, you know, what happens when a patient bleeds? What is everyone on the team doing? If there's a code in the OR, what do you end up doing?

How well do we treat our learners in the operating room? And we're using the data that we get from that to try and think of ways to collaboratively improve patient care and improve team dynamics as a group, which I think has been a great use of that tool and technology

Speaker 4: And there is a little bit of supporting literature already out there too for the type of use case that you're describing, Dr.

Greenberg. There's a paper from Campbell et al in 2024 using [00:09:00] OR black box technology to determine quality improvement outcomes for in situ timeout and debrief simulation, which really combines a couple of those features really nicely, similar to what you were just addressing. So in this study, teams practice debriefs in the room in which they were about to operate at 6:30 AM for 15 minutes before a 7:00 AM case.

They did it 30 times and found that those who did the sim weren't better at the timeout element but were significantly better at the debrief. And then they were able to analyze those post-sim discussions to show gaps in policy understanding as well as a lot of interdisciplinary team insights in the culture of safety and how to improve it, which, you know, as a resident who's gotten to attend these Better Together rounds, feels similar to how we grab everyone's opinion to comment on what we've heard.

Speaker: Yeah, I mean, I think one of the main underlying goals of that conference has been to improve the psychological safety in our operating rooms. And when you read some of that data, the data on psychological safety shows that teams that have high levels of psychological safety actually [00:10:00] tend to report more errors and more problems because everyone feels empowered to do so, right?

They're probably actually doing the same amount of errors. They're just more open with one another in recognizing them and then leaning in to try and improve them. And, and that platform and getting all of those key stakeholder groups together and to talk around video that we can see and evaluate with our own eyes and ears has really helped to try and move the needle on that culture in the operating room so that people who I think historically have not been empowered to speak up now feel able to do so.

Speaker 4: As someone who just loves quality improvement, this is a really exciting space to me. But a whole other bucket of utilization for this technology is telepresence. And so far we've been discussing digital surgery platforms that are principally observational or allowing us to analyze, but we know that Intuitive is working on telepresence, Proximie is working on telepresence, and other companies as well.

Speaker 3: Yeah, and for listeners that don't know what Proximie is, it's a teleproctoring [00:11:00] software that can integrate into your existing hardware in the operating room and allows for an augmented reality experience. So imagine if you're doing a case, a surgeon who's sitting in another room in the hospital or even a different state or in a different country can be plugged into your feed- Mm-hmm

for any of your existing cameras, and that includes laparoscopic cameras, robotic cameras, or even open cameras that are mounted in the operating room. And it allows them to draw on the screen, directly communicate with you, and teleconference about interesting, difficult, or new cases that are being performed And they've done this in interventional neuroradiology here in the United States, and spine surgeons have used it with good results in the Middle East.

And Dr. Greenberg, I haven't actually seen this one in person, but we've discussed some prior medical legal concerns regarding telepresence. And I'm curious to hear what hurdles there are right now to broadly implementing such a call-a-friend, phone-a-friend system in the operating room.

Speaker: Yeah, so I, I, uh, even before getting to that, would love to share, like, my favorite day of COVID, which was, using Proximie to do a SAGES hands-on course [00:12:00] completely virtually.

In the midst of COVID, when the SAGES meeting, like every other national meeting, got canceled, SAGES wanted to try to use a technological platform to remote teach a hands-on course, and we ended up using Proximie to basically teach surgeons in their own homes using an animal model that was made for the procedure itself to do transversus abdominis releases.

So I was sitting in Madison, Wisconsin, on my computer, helping a surgeon in Arizona whose pool in his backyard looked very nice in the middle of the Wisconsin winter. And I watched him and coached him through how to do a transversus abdominis release completely remotely, and it was my first experience using this type of technology, and it really kind of for me showed the potential for this in the future.

I think this is going to be a way that surgeons can help other surgeons and patients in the very near future. But to your point, Zach, there are pretty significant medical legal hurdles to this, right? If I'm sitting in an operating room in my hospital and a surgeon [00:13:00] requests a telepresence or a telesurgery consult, uh, at a hospital where I don't have privileges, it's probably the best thing for the patient for us to be able to help them, but there's a whole lot of medical legal stuff that goes into that, right?

You have to be licensed across state lines to practice in every state, and you have to be credentialed in different hospitals. And so I think those things all need to be worked out. But if you think about this in a purely patient safety and patient outcomes perspective, like, if you can get help from an expert surgeon or from a colleague or from a friend, i-i-it's only gonna be better for patients.

So I, I do think that there is clearly a place for this in the future. It's just working out those, those pretty significant medical, legal, a-and other concerns first.

Speaker 2: Yeah, I, I totally agree, Jake. We certainly see, you know, especially surgeons in general have tendency to want to show up for our colleagues and help out our colleagues in need.

So for us, I think a tool like this, [00:14:00] if we are the end users of this technology, its intended purpose is perfect, because we would love to do that. But, you know, how our healthcare systems, legal perspectives are structured, you know, that probably needs to be addressed before this could be used for real time telepresence or teleproctoring or telecoaching.

But certainly in an in situ situation, in sim labs or in situations that you described, Jake, with your colleague in the backyard for education purposes, this is really incredible use case. And I would love to avoid the long plane rides to do surgical education conferences. It's much better to do it at the convenience of home.

Speaker: And I mean, I think obviously we practice right now in a great group of colleagues and partners who, when I am in trouble in the OR, I can call and they come, right? And they help. And having that extra set of hands and eyes is incredibly invaluable. But if you're a remote surgeon or if you're out in a rural community and don't have support, using this [00:15:00] is the right thing to do.

I think the, the clear need for it is there, and the fact that we can do it is going to be, I think, better for patients in the future.

Speaker 2: Agree.

Speaker 4: It does make me wonder a little bit about trust. You know, you know who you're gonna call for every specific problem you encounter, or who's the expert in this or who's the expert in that.

So when it's a stranger who pops on your screen, h- how do you think about that trust?

Speaker 2: Yeah. So I think, Joey, that's a really interesting point, right? Is this gonna be kind of like a crowdsourcing for expertise? And how do we credential these so-called experts to teleproctor you or to be there in telepresence?

And how do we ensure that their technical expertise is qualified, right? So we don't really have a measurement system in which we could understand whether there's an expertise. So all of these need to be worked out. But I can certainly see a situation where this is a colleague that you trust, uh, that you've already [00:16:00] worked with or you were mentored by and you were a fellow with us and you're now practicing in remote part of the country, you're the only specialist there, then I think this could be a really good case.

But certainly the expertise needs to be really well identified and qualified. And this also brings up another question of if this is to be a routine kind of practice, you know, how should physicians who are consulting be compensated? And what kind of aspect is there in terms of employment? Uh, and again, this goes to the medical legal, 'cause now if you're compensated for that practice, then do you have liability?

Do you have stake in the patient's outcomes as well? So all of these are really fascinating questions. The technology i- is really nascent and needs to be worked out. There are other things like network stability and all this stuff that also needs to be worked out. But really cool premise of this technology.

Speaker: Yeah, and I think the things that'll, that'll come into play before some [00:17:00] of this stuff, which is going to be harder to figure out, are some of the decision-making tools and aids that we'll see that coming out of digital surgery, right? Like, as we continue to build large AI databases, there's gonna be machine learning algorithms that are gonna be able to probably very safely and accurately tell me that I have achieved a critical view of safety and that what I think I'm about to clip and cut are the safe things to clip and cut.

Now, right now, my own way of doing that is through the mental representations that I have of doing all the gallbladders I've done in the past, right? But if you feed a computer 100,000 gallbladders, it's gonna have seen a whole lot more than I'll ever be able to do in my entire career, and it's gonna probably know a lot better than I am that it is safe to go forward.

You still have to trust that, to your point, Joey, right? The trust needs to be there, but if I have another tool that's telling me that what I think is safe is safe, that gives me a little bit more comfort and reassurance to continue to progress the case

Speaker 4: Makes a [00:18:00] lot of sense. Thank you.

Speaker 3: Yeah. And I think that in that regard, the system right now that has the most information for that is the integrated intuitive hub for robotic platforms.

It's a very interesting hybrid because you get both the telepresence picture and picture for teleconferencing in the new DV5, but you can also utilize the company's video storage and AI video segmentation for video review later. Dr. Jung, I know that we're already using this at Duke. Do you have any thoughts on the intuitive model?

Speaker 2: Yeah. So certainly it's a well-integrated digital surgery model that can be used for, like you said, telepresence, although, you know, rarely used, if ever, so far. But I think for our trainees, it's important that they have access to the videos that they've done and can segment it for easy viewing. And also, as Jake mentioned, there's a increasing application of artificial intelligence, especially in computer vision models, to have a better understanding of what's happening in the videos.

So in our lab, we've been [00:19:00] utilizing intraoperative video and applying machine learning models to identify significant bleeding and as a way of understanding intraoperative adverse events and as a way of understanding real-time decision-making, but also quantifying risk for the patients at the end of the case.

So there are a lot of these use cases on how we utilize video for safety research. Not only we talked about efficiency and education, but also in terms of increasing patient safety as well.

Speaker: Yeah. And I think, I can't remember who mentioned it earlier, but someone was talking about expert athletes using video for performance improvement, right?

And, and I think we strive and want to do that in surgery, but it, it's pretty clunky still in terms of how you get the video onto a USB stick and then onto a computer hard drive somewhere, and then you have to review it with someone. And I think as these video libraries accrue, you're gonna have much more opportunity to not only see how you performed in the last case, but ideally load your [00:20:00] next patient's CT into some simulator and then be able to practice that same case on a virtual version of the next patient's actual anatomy.

So I think that as all of these things continue to be innovated and moved forward, it's just a really exciting time for trainees because I think you're gonna have a lot more ability to practice and learn both pre- and post-operatively than we ever did in the past.

Speaker 4: I definitely think as a resident, this conversation around getting your reps in and how we're best gonna be able to do that and how many physical versus virtual reps we're gonna need as this technology develops is a fascinating one.

But I also know in another one of our podcasts we talked about the death of laparoscopy and how hard it is to fit in training on all of the things that we already wanna learn how to do. So, you know, I'm curious to how you all see this changing surgical training. Is there a potential leveling of the playing field across differently resourced settings?

Is there more autonomy, or does review potentially become one more thing crammed into an 80-hour week that gets increasingly challenging to accommodate? [00:21:00]

Speaker 2: Yeah, Joey, I, I really like that perspective. And as Jake mentioned, this is really exciting time for a trainee, but also, you know, for the educators this is very important time that we really need to design the curriculum, especially utilizing digital surgery in an effective way so that the trainees could learn surgery in very effective manner.

So for example, using video is very effective, we know that, but there's no way you could go through, sit down and go through two hours of case. You've done the two hours of case, and then now you're gonna watch two hours of case. That's just not possible, right? So I think this is where the technology has to be get better.

The understanding of the video should get better to kind of bridge that two-hour video into really condensed five, 10-minute video, really giving you the highlight reel. But doing that in almost real time processing. You, you really want to be able to review [00:22:00] that within a day, ideally within few hours of the case or in between cases, rather than waiting a month later, right?

So, you know, how do we really process that quickly? I think that's gonna be important.

Speaker 3: So Dr. Jung, I think that what you're saying is really important, and that's really the crux of where some of the different tools in digital surgery are going to be able to be utilized to help with surgical education.

And, and in particular, you're, you're referencing AI video segmentation, which some platforms already can segment cases into important parts like Dr. Geem had mentioned, the clipping of the cystic duct and the cystic artery, or there are big use models with bariatric surgery, looking at the GJ, the JJ, closure of the mesenteric defect.

And, you know, AI software right now already can identify when these parts of the cases are happening, how long those parts of the cases are taking to perform, and what the efficiency of movement is for the surgeons in these specific parts. But right now it's still very new in how we're going to learn to broaden our understanding to apply to a wide variety of [00:23:00] surgical operations.

Speaker: I think that segmentation, though, is particularly appealing as an educator because to me, like, the best way to practice is through deliberate practice, which gives you a defined subset of that procedure that you wanna practice on until you've achieved competence or hopefully mastery of it. And then once you have, you should move on to a different portion of the operation that's a weakness for you so that you can work on that and turn that weakness into a strength, too.

And overall, that's how you globally improve at everything. I think practicing an entire operation from start to finish is very cognitively stressful and challenging. But segmenting it into parts, getting really good at the first and then the second and then the third until you're good at all of them is ideally a, a better way to practice.

Speaker 2: Yeah, Jake, I think that's a m- really wonderful point. And Joe, you mentioned earlier about autonomy, and I think, you know, digital surgery tools and having video and applying machine learning analytics to it could really assess autonomy pretty well, and this [00:24:00] gives you that kind of expedited track, if you will.

Instead of a surgeon that you work with remembering where you were at in the previous procedure, if you have a digital surgery that's personalized to you and could really demonstrate that you've met that autonomy in the previous sessions could really get you on more expedited kind of process, and reach that autonomy throughout the case rather than just the steps that you already had autonomy for.

So I think digital surgery tools could, will eventually evolve to give you that kind of assessments along the way during the course of your training, and really help the educators understand where you are at better than just remembering, you know, six months ago what we did, which is almost impossible for us to remember.

Speaker 4: Which fits really neatly into this shift toward competency-based assessment as well.

Speaker 2: Mm-hmm. Absolutely.

Speaker 3: I mean, as a MIS fellow, I think about, imagine I [00:25:00] had the opportunity to segment my videos where I could just look at every single robotic JJ anastomosis I've done with Dr. Jung and review that data in a condensed fashion as I progress throughout the year.

I think that has a lot more utility than sorting through a couple hours of video to find one specific part of the operation.

Speaker: Yeah, and I think one of the, the biggest problems with feedback for all you guys is that we're not great at giving it, right? There's a bunch of competing tasks. We all go see our next patient.

We forget to tell you what you did well, what you can improve upon. But I think in the very near future, and probably already, there's gonna be digital applications that can review your performance and give you some meaningful feedback as well. And so even if the faculty forgets, which is not ideal, but certainly happens, you get some data that's actionable on what you can do better.

Speaker 2: Yeah. We- we've actually been testing on the Viggo language model to give structural feedback on laparoscopic videos, and [00:26:00] they're actually not too bad. I think it's really still nascent technology, but it's gonna keep on getting better with more training data. So, you know, very interesting times indeed.

Speaker: Yes. I I think this is an incredibly interesting conversation and, and really remarkable technology that is gonna have an increasing footprint in all of our lives as surgeons in the immediate and the foreseeable future. And so I think to those who are training and those who are in practice, learning more about these platforms, their capabilities, and, and thinking about how they can interface with you and your practice for the betterment of your patients is gonna be something that we all have to learn more about in the near future and look forward to where this is heading.

I think it's very exciting. So keep an eye out for emerging digital surgery so that you too can dominate the day