Transcript

- Colleagues and friends, thank you for joining us for another session of "The Virtual Operating Room." My name is Aaron Cohen. Our guest today is Dr. Walter Jean from Fleming Neuroscience Institute. He's the chief of neurosurgery there. He's a tremendous mind in neurosurgery. He has done an incredible work with video lessons at the Journal of Operative Neurosurgery. He has written one of the most advanced and recent books on skull-based surgery named "Skull-Based Strategies" by Thieme. I've always admired Walter's curiosity, innovative creativity and modernity, and really his impression on neurosurgery in terms of virtual reality, augmented reality our second to none. So today he's going to talk to us about the application of AR and VR in neurosurgery. Walter been a pleasure working with you on a few projects in the past, and I'm truly excited today to learn from you. Please go ahead.

- Well, and first of all, thank you so much for this opportunity. This is a very esteemed series, as you know, and very popular, and I'm honored to be a part of this and hopefully share some interesting ideas with you all today. So again, thank you and let's get started. Title "Citius-Exiguus-Titius: Mixed Reality in Complex Intracranial Surgery." I'll explain the Latin in a bit, in the middle of it. This is who I am and this is where I'm from. This is the Lehigh Valley Fleming Neuroscience Institute in Allentown, Pennsylvania. These are my disclosures, no big deal there. And we're talking about virtual reality and augmented reality. I think most people know what all this means at this point. Virtual reality is computer generated environment that you can interact with, and in our case, we're turning am ordinary CT scan and MRI scans into three dimensional renderings that we can interact with. It becomes very, very convenient to examine three dimensional anatomy and then drill and implant markers and things like that for surgical planning, which is what's something that we'll talk about in detail as we move on here. This is what we do. We put on goggles and use controllers, and then we can examine all this in three dimensions, whichever angle you want, and also interact with it, as I said, drill, eliminate, even resect the tumor if you wanted to in a mock simulation. Augmented reality is a cousin technology. It's not the same thing. It is a projection of what you see in virtual reality onto the visual world, the real visual world. Every given Sunday with NFL games that yellow line is augmented reality, and you're using the television to project that so that the viewers can see it. It doesn't really exist on the field, but you can see it and you use it for enjoying the game. So all this spells out some new ideas, how to use this. Old school, when you have a approach problem, you say, well, how should we do this case? And then you get an answer from an old guy with white hair that says, you do this. And then you start questioning, well why are we doing that? Because I say so. This is not to denigrate training in the old style, but this is a lot of how we were trained years and years ago. Experience from your mentors, you take it, you drink it up, you soak it up, and you say, okay, well that's, I'm gonna do that as well. And so a bicoronal, bifrontal subfrontal approach for something like that. Okay and whatnot. New school, you don't need to rely on experience per se. You can interact with all this anatomy in virtual reality and test it for yourself. So you examine the 3D anatomy. You drill the bone openings, and you can drill it until the cows come home. You can, you know, use 10 approaches and take a look at what the exposure looks like, and arrive at your own conclusion as to which exposure is best. So something like this, a relatively complex tumor that's hard to get to, you can drill all the approaches, and again, make your own determination of what to use instead of relying on the white hair guy. Again, the old school now about using measuring things and drilling intraoperatively. How do we get to the IAC in an expanded middle fossa or middle fossa approach for vestibular schwannoma? Well, the old guy says, well, you draw a line for here, you draw a line there, and then you bring your protractor in the operating room and you bisect the angle, and there how you find the IAC. So that's the, you know, not so reliable way. I don't mentally have a protractor in my head. And so now I rely on augmented reality to tell me where to drill. So in the microscope, I can see markers that I implant, and if I also put in the cochlear and the semicircular canals and whatnot, I'm not relying on a mental protractor to do all that. I can drill by seeing the labyrinth, cochlear, internal carotid, and so on and so forth, and know that I can drill here safely to get to wherever I need to be. So old school, new school in a very brief summary, that's kinda how we're using this technology/ Show you in a real life case as to how to combine these two things. Here's a unusual location for ventricular tumor. It's not in Monro, it's on the side of Monro. So Monro designed approaches are not particularly appropriate. So let's set up the VR for it. You can see that in the VR the skull was rendered, the ventricles are rendered, and the tumor is rendered. You can dive into the ventricle, take a look at things, and now you can implant markers. You implant a trajectory. So you now you know if you just line up that trajectory, head and tail, you can get to the tumor with a very, very small opening. And there you have it. Our opening is, the craniotomy is tiny, and the access to the ventricle is even tinier. And you have a zero doubt that you're gonna get there. So in this setup, what you have is you a rendering of the virtual reality model. You implant your approaches, you pick your approaches, and you implant your markers, and then you execute it with AR. Now, the AR is dependent on navigation, of course. So in the operating room, three things have to talk to each other to make that AR accurate: your navigation, your microscope, and of course the augmented reality system, the computer system. So in a nutshell, I'm gonna give you, this is what we're gonna talk about today. I'm going to give you some examples of really interesting cases to use this in various different ways to make surgery more efficient. So we're driving now with GPS, with heads up display. Essentially that's what it is. You set up your map, you set up your destination and where you're gonna go. And then the augmented reality over the dashboard's gonna tell you, turn left, turn right, make a stop here, and you get to your destination. So very recently, we had published this in ONS, the first 100 case series of our endeavors, and we're well into about 120, 130 cases now. And we find that these are the five things that are very, very, this is how the system is the most useful. So as you can see, guidance to target, the precision of the opening and allowing you to not get into trouble and avoid critical structures by navigating away from anti-targets. All these are how those 100 case kind of splits up as to how often each use was highlighted in that 100 case series. So let's talk about nitty gritties now. Let's pull apart some individual cases and talk about them in great detail. I did this tumor 20 some years ago, and look at this big opening, right? We used to do FTOZs in one piece. Take out the zygoma, take out the orbit, take out everything, take out all the bone in the world, and that was the style that I trained in about 20 years ago. Now everything's about keyhole. And this mixed reality is very, very good in keyhole surgery, and I'll tell you why in the little bit. Everybody wants their keyholes to look beautiful. If you place the keyhole in the perfect place and place keyhole in the perfect size, you're gonna have a beautiful view. Problem is that sometimes your keyhole shows you a train wreck because you put in the wrong place. So this is what I call Goldilocks principle. You want to place a keyhole exactly where you want it, exactly the size you want it, so that you make a surgery safe, effective, and at the end, efficient for the patient as well. So what mixed reality does for you for the keyhole surgery is that you can put the template of that keyhole, you can put that keyhole in an anatomically correct location so that you know this is the perfect size and the perfect location for that Goldilocks keyhole and you don't need to make a giant opening FTOZ to get to a very similar tumor in this case, yeah, because you know, yeah, that your keyhole is in the right place. You don't need to, you know, account for error and missteps by using a big opening. So this is something you turn on the eyepiece. The eyepiece then shows you that template. We had drilled this opening in the office apriori, and then you know, your keyholes gonna get you to the tumor very, very precisely. And this tumor came out very nicely through a mini-pterional craniotomy, no longer needing the FTOZ that we used 20 years ago. So here's an example of another keyhole surgery. We're using the same system to do trans frontal sinus approaches for planum sphenoidale meningiomas. You can see that the keyhole is only one centimeter by two centimeters large. We go into trans frontally. And by the way, the incision is over the glabella. We call it open size seagull incision. So it's not a big incision, no big bone drilling. Again, mixed reality gets you the target very, very precisely. Here is a mini, mini transcollosal transchoroidal, where the opening is about one centimeter by 2 1/2 centimeters or so, getting down to this target, again, utilizing the mixed reality to avoid the draining veins and to cut the corpus in the right place to get to the target, and having your angles all correct so you don't have to account for error by making a big opening in the skull. So transorbital surgery, this is great use for mixed reality. Those are our templates that we did in the virtual reality. Now we're drilling the orbit with this transorbital approach. And you can see that I know exactly where my tumor is behind there and therefore where I need to start my major drilling and so on. And that marker that you just saw was on the anterior clinoid. We're doing anterior clinoidectomy from the transorbital approach. doing the Hakuba peeling from the transorbital approach, all of it guided by markers and opening templates that we had implanted in VR so that then we can use in AR as we track the anatomy intraoperatively. And we know that we're gonna get to the target with absolute very little fuss, with not a lot of wiggle room and doubt and so on, and makes for very efficient guidance to a transorbital surgery. And by the way, no one has done 5,000 transorbital approaches. So novel approaches, this really, really gives you a sense of security because you've done it, you've rehearsed it, you know that you're gonna get there without a lot of trouble. Here's a case that I know Aaron would be very interested in, because it's a, what we call, what he calls node to node work, what I call dot to dot work. It's almost like just connecting the dots in like a children's drawing. We implanted several trajectories to this transchoroidal and transventricular approach, and we all simply had to follow each trajectory. So this is the first trajectory from the sulcus into the temporal horn of the lateral ventricle. Then it goes transchoroidal. There's a second trajectory. That transchoroidal approach through the horn is gonna get you to the interpeduncular cistern. And then you can see that on the tops of the screen there, the trajectory is guiding me exactly to the tumor. So what we did here, so basically one dot to the next dot and the next dot to the next dot, all these right hand turns get you to this tumor simply by following the dot to dot to dot again, just like the children's game where you connect the dots and you get the drawing. So we successfully did this by connecting the dots into the temporal horn, temporal horn transchoroidal fissure and then into the ambient cistern and posterior clinoid, which is where this tumor resided. This is a very interesting case as well. It's insular glioma in a patient that we could not do awake. I didn't not think that this particular patient had the personality and maybe the maturity in being awake during this operation. So we're really in a bind here. So what we then we decided to do was simply put the fMRI representation of Broca's onto the AR, well onto VR, and then with AR projected onto the actual tumor. So on the right there, what you see in purple is what the fMRI believes to be her Broca's Area, her speech production area. The green is simply the tumor, and then the red, of course, the vessels. Now you can correlate the real time vessels that you dissect in the fissure with the AR representation of the vessels, so you know your guidance exactly right on target. And if you trust your fMRI, which, you know, this is the only way you can do this with a non-awake patient, hopefully you know, you stay away from Broca's. And for this particular case, we were very, very lucky. In fact, the fMRI was right on. We avoided those areas. We got into the tumor and we got an intended near total resection of this thing without damaging her speech. And in this particular case, that AR recitation of Broca was the only safeguard to protecting her speech because we, again, we couldn't do it awake for various reasons. So this is hot off the press here. This is our final, the video that we just sent to ONS and got published. This was a very multi-step, multi-stage approach to a chondrosarcoma that we did in the CP angle and it went down all the way to hypoglossal canal. And what we did was simply do this, repeat the same thing. In VR, drill it, drill it, see all the angles and say, okay, we don't think that we can do this through the nose. We are best to do this through a combined petrosectomy approach. And by the way, we had to get down to the hypoglossal canal. How do we do that? Okay, we drill it in the office, the right hand corner, bottom corner, shows you that green line, basically, if you followed that green line, you're gonna get to the hypoglossal canal to get that last piece of tumor. It did that flawlessly in this case. And we were very, very lucky to have all this technology to guide us there without a lot of fanfare and a lot of sweating knowing that we can get there. Cerebral vascular applications is even more powerful with this system, and I'll show you why. Let's start with aneurysms. Now you don't need AR, VR to guide you to a aneurysm for most of the cases. But for anything that's unusual, for anything that the approach is a little bit odd, it actually becomes very, very helpful. So this is a two for one kind of a deal, where we wanna do two aneurysm clippings in one approach or a mini-pterional approach. And this gets a little tricky because you don't often cross to the midline every day for aneurysm clipping. So an unusual trajectory makes this technology very, very powerful. So it's a reach over approach. Your positioning is already a little bit odd. You're kind of putting it in the middle of the two approaches and you're viewing the second aneurysm backwards. You're not supposed to be looking at it that way, but you know, so it's anatomically challenging. So we practice it and practice and practice in VR. Here's a clipping of the first aneurysm, which of course is much more straightforward, not a big deal there. But you know, even for this one because we had placed the head in the position to kinda sorta help with the second operation, second aneurysm, it was a little bit off. Now we're reaching over to the other side and you say, okay, now where the heck is this thing? And you can see the blue dot there with the label three. Well that's the base of the aneurysm on the other side. So if you believe your navigation, in the case we do, because we have enough trust in the system now, we simply say, well it looks to be a very odd place for that dot to be, for that base of the aneurysm to be, but we're gonna trust the navigation and trust the AR, and lo and behold, it gets you exactly to the base of that contralateral side of that ICA terminus. So that again, you don't have to fuss and sweat bullets in trying to find this thing hours and hours and hours because of the unusual trajectory. So again, the reach over approach, both aneurysms got eliminated. Now, not a lotta people clipping basilars these days, so, you know, this is not something that we do every day and it wouldn't hurt to get some help now, would it? So here's the situation of a Hunt-Hess 3, 58-year-old basilar apex aneurysm, that's very odd shaped. And my neuro interventional radiologist didn't wanna put a WEB in this and said, why don't you go take care of that? And I said, fine, let's use our tools to the greatest benefit and see whether we can get there safely and whatnot. So we practiced it, we'd rehearsed it in the office and make sure that we knew what we were doing. And AR with navigation tracking that microscope was very, very helpful in two regards, and I'll show you what it is. So again, this is a challenging anatomy of this basilar aneurysm. The first place where AR is very useful was in getting rid of the PCP, the posterior clinoid process. So as you saw, and I'll pause it for one second here, that P, I hope it's a P, shows me where the posterior clinoid process is. Again, no doubts, look yeah, you can use your anatomy, but wouldn't it be a little bit of help in this very rare operation to get some little assistance. So there it is, I need to drill right there to get rid of the PCP. Get rid of it so the third nerve has a little bit of room, I have more room for the temporary clip. Again, I don't have to sweat bullets, whether this is the right place or wrong place. Navigation can help you, right? You can put the probe in there, but how many times have we had a situation where the probe is blocked by the microscope and you're fussing with raising the microscope, to put the probe in or you know, having to put the probe in and having to look at a monitor to see whether that's your PCP. It's much easier just to project that PCP right there and when you're done with it, get rid of the AR for the PCP and move on to other things that you're interested in to pay attention to. And as we move on with this video, I'll show you what we pay attention to next. So we did the posterior clinoidectomy with the Sonopet, and then now we're in the midst of trying to find the basilar trunk. And you can see the AR signals tells me that the basilar trunk is over here and I'm working over here because I'm just completely stuck in this massive clot. Can't find anything with this clot in my way. So I know I have to get to here, but I'm working here knowing that okay, I need to get to that trunk to put the temporary clip in there. Let's see whether it's accurate and I wouldn't be showing to you if it weren't. So it got me there by leading me there and then get rid of all the blood so that I can see it. And then the test clip goes in, our first clip goes on, not the best clip, it's only a pilot clip and then the second clip goes on much better. I try to avoid all the perforators of course. And then, so for this relatively rare operation these days, this mixed reality guidance system really paid off and helping us get there very safely and get the job done, again through not a giant opening. So there's that one. So here's another one that we recently published on a 66-year-old with this very unusual ACOM setting, okay. So it's a seven millimeter that needs to be clipped because she has family history and it's only fed on one side, the hypoplastic on the right. Now we say, well what's the big deal? The big deal is that he has a left forehead skin graft that took like 16 operations to do when he was a teenager. Basically he necrosed his nose from nosebleeds and they took forehead skin to recreate a nose for him. So that left forehead is absolutely no touch at all. So what are you gonna do with this? Well, okay, you can coil it. Patient didn't want to hear about coils and wanted to clip. Okay, so we said, okay, let's figure out a way to get there without going through the forehead or even touching the forehead. All right, well transorbital. We're doing a lot of transorbital work now. If we can take the sidewall of the orbit to get to tumors or interesting places in the middle fossa, why can't we take the roof to get to the anterior fossa? By the way, the transorbital approach was initially started for anterior fossa work. Now they abandoned that because of leak issues. And you know, we do have leak issues still, but approaching the anterior fossa with transorbital work for transorbital approaches is doable and we show you something about this one. So because we couldn't go the forehead, we said, okay, let's go through the orbital roof. That shows you enough of the aneurysm. We had very little doubt trans palpebral, all the AR markers for our transorbital approach here. The burr hole, so to speak, is actually on the roof of the orbit, okay. So after we drill that burr hole, we took a little bit of the rim and this is more rim than we absolutely need, 'cause this is when our, you know, still want some security, security blanket, took the roof, took more roof, took more roof and then here's the optic nerve. If the optic nerve gives you A1, A1 gives you the ACOM, ACOM gives you the aneurysm. So the AR marker here is for the A1. We got very close to the aneurysm there, put a test pilot clip, there's your aneurysm. So absolutely no question about our trajectory and absolutely no restriction on any space, any surgical freedom issues, because we know that our opening is big enough having rehearsed it in VR. So that clip went on, we didn't touch the forehead whatsoever. We only went through trans palpebral. Now the leak is an issue. I'm not gonna sugarcoat that. But fortunately the leak eventually stops as hopeful, we all think eventually stops. And that was fine. And I'll show you a little bit of another one later. So keyhole surgery, again, back to keyhole surgery for aneurysm, it's a little bit scary, because you just don't know how the aneurysm's gonna behave. What if there's an intraoperative rupture? Do I have enough space to deal with it? What you can do in testing all of this is not only just test it, the opening for the amount of space it gives you, but also test the opening of the Sylvian fissure. So on the right side, what you see is A VRN representation in yellow of the Sylvian fissure. You know that this aneurysm is completely, completely surrounded by CSF in the Sylvian fissure. And if you have enough space to dissect the fissure with the bony opening, you're fine. So look at how we test this out now. This is the key. This is a pretty big keyhole, by the way. Didn't even need that much space. What we're doing now is testing. So here's the keyhole that we did for this mini-pterional for this MC aneurysm. And this is just to show you that we follow the exact template that we had designed for it and then the aneurysm exposure was a no brainer. There's plenty of space with this mini-pterional opening. You don't need a big opening for these MCAs. And I'll tell you why we're so confident, it's because we did this. So in the office, we said, okay, our initial design opening is way too big. We don't need that so-called pterional, full pterional opening. All you really need is an opening that encompasses both these trajectories. One trajectory is for the dissection of the Sylvian fissure and the other trajectory is with a clip applier. That's all you need to do. So this is way too big. Let's reset this model and re-drill and we drill inside out by the way, so that we know our burr holes are always in the right place because you're drilling with all the anatomy right in front of you. So all you need is it big enough to encompass those trajectories. And again, it gives you plenty of space to deal with this aneurysm. So now back to Transorbital and ACOM again. Why am I showing you another case? Well, because some people were not convinced that that first case was actually transorbital because we took off the rim. This one we hardly took off any rim. And so here is that ACOM, okay, right in the center of your screen there. That's the aneurysm that we're looking at. And we said, okay, I think we can get there with plenty of space with the transorbital. Here's what you see, we're drilling off just this little bit of the rim, doesn't even show you any dura whatsoever. So this rim is about 1.5 centimeters by about 1/2 a centimeter deep. Only that overhang, if you will, of that orbit is drilled off. And then the rest of it is just roof work similar to that first case. And again, the step-by-step or the dot to dot in this particular situation is simply finding that deep orbital, deep optic nerve, way deep towards the foramen. There it is right there. And you follow that distally, that gives you A1. So optic nerve gives you A1. A1 gives you ACOM. ACOM gives you the aneurysm. And you can, I'll show you step to step, but here's the A1. A1 then leads you towards the communicator. And once you have the communicator, of course, you wanna show the contralateral A1 in this case, since we have one in this case, not the last one. And then you dissect the other, the A2s. And, you know there's a perforator, let's try to avoid that. You have to have minimally invasive instruments, single shafted bipolars and single shaft scissors, et cetera, et cetera, et cetera for this. And one side of the clip, one blade goes there and the other blade goes the other side. And we decided to use a left hand despite being right hand because of the approach here. And then the clip goes on again, plenty of space, not a lot of fuss. Now you're gonna say, well you just generated a blister aneurysm for yourself. That is true. So the ICG's gonna show you that, that there is a little bit of residual there. So we, you know, use the doppler and then ICG and there's your blister, right? You got a little bit of a residual aneurysm there. All we did was we put this a curve clip there just to laminate that blister and that was it. So you can see that's it's totally doable. See, so there's your eye, there's your opening, tiny lip of the orbit taken away that we replace obviously. And then there are your clips. And yeah, very minimally invasive. There's hardly any incision. Patient went home day one or two and recovery was very, very smooth, in this case, trans palpebral incision. There's your post-op angiogram and 3D spin. And you can see on the top side here, this is as little, this is the craniotomy, so to speak, only that little rim of the orbit. And then of course this is proving that it is in fact a transorbital approach as we use most of the exposures through removal of the roof of the orbit. There is not another realm where AR and VR planning and execution guidance is as powerful as AVMs. AVMs is the single place where this is most applicable and incredibly useful. I'm gonna show you small cases and go to bigger ones to give you the sense of that. What it gives you is a very, very clear understanding of the angio architecture, because you go into this skull with the brain taken away and you look at the AVM, and you basically use this 3D painting device and paint all the inflows and outflows. And then in your mind, having done that, you have a very, very clear understanding, okay, I got two feeders, I got one drainer. These are the areas where they are, all I need is this opening and I can get to this whole thing. And again, going to dot to dot, all you're going is focusing all your attention to the feeders, dot one and dot two. You don't care about the drainer, you don't care about the nidus, you don't care about any of that stuff until those dots are eliminated. So dot one, let's get rid of that and then dot two, let's get rid of that. After that it becomes a bloody tumor, that's it. It becomes a bloody hemangioblastoma, something like that. The drainage is a complete non-sequitur because at the end of the case it's already dead. You take out the bloody tumor. After that, those deep feeders are done and you're done. So the intimidation factor of AVMs is completely eliminated and it also makes you faster. All right, so let's get a little bit more deeper into this topic because it's a very, very important topic. This is a multi-embolized grade three with onyx to spare, okay. And very, very confusing and intimidating when you just look at it, the angio, okay. And even intimidating when you look at the 3D rendering. You go in there and say, I have no idea where the drainers are, where the feeders are, which one is vein, which one is artery. So, but you can decipher that. And once you decipher that and you paint it, if you paint all the inflow with one color and outflow with another color, it becomes no brainer, no pun intended, when you're operating on this one. So let's look at it piece by piece. The first thing you would notice on the angio is this squiggly MCA feeder. All right, find it. Okay, here's a squiggly MCA feeder. I'm gonna paint it green, I'm gonna paint it green in 3D space. And I'm gonna put a dot where I think I'm gonna disconnect it. Now next one is a little bit more scary. It's a PCA feeder and it's very, very deep. Okay, again, I'm painting it in blue in this case, and I'm gonna in my mind say, okay, this is the major, major feeder that remains after all the onyx and this is where I'm gonna disconnect it. Now here comes another intricate part. This dot where I want to disconnect this PCA feeder is directly behind, let's go back. It's directly behind the venous varix. Now what to do about that? All right, so there's the feeder where I wanna disconnect it. The light blue is that venous varix. So if I have to get to this pink dot and go through that venous varix, and in between the venous, the drainer and the SSS, I could get myself in trouble. So what I did was I decided to go transfalcine, for this particular feeder. Now that's kind of crazy when you think if you go transfalcine, you may not find this thing. Well, I have my tools and I'm confident I can do it. So let's set up that transfalcine approach, okay, so that I can avoid that large drainer at the beginning of surgery and eliminate this major feeder without touching that vein, transfalcine approach. There it is. And then that dot is where I'm gonna cut the falx. Okay, so let's look at the entire plan now. There are three major feeders. Okay, here's the transfalcine approach, for the deepest and biggest feeder. And then the other PCA feeder is labeled VR 2 in pink, orange dot to show you where to disconnect. And then finally that MCA feeder is much, much smaller. This is now completely deciphered in my mind. I equate this like, you know how in medical school when we had the brachial plexus and you had to draw the C5 and C6 and C7 all in different colors and finally understand brachial plexus. This is how I understand AVMs now. Without this, I'm not sure I'm at all even competent AVM surgeon if I didn't have this tool. Now in real life, here are the dots and here are the markers. So first move is a transfalcine move. I know where to cut it. Boom, there it is, there's the feeder, no doubt about it. Let's do ICG to make sure. There you go. Eliminate that, and the single biggest feeder, that deep PCA feeder is gone, boom. Next is the other more lateral PCA feeder concentrated on dot number two. Don't worry about the nidus, don't worry about any of the other stuff that's in your field. Concentrate on number two, done, number three, done. With those three big ones done, again, you simply have a blood tumor. So you circumferentially dissect, circumferentially dissect, and the rest is really not a lot of excitement. And then of course at the end, the veins are completely non-sequiturs because once you've disconnected this thing, the vein is just for the taking. And very fortunately she did great and in fact her vision, believe it or not, improved after this resection. So that tells me that maybe there was some vascular steel phenomenon that was affecting her vision when this thing was rip roaringly filling and draining and filling, draining all the time. Another powerful use similar to AVMs, right, fistulas. Fistulas are challenging surgical entities because, you know, in 2D you know where the fistula is. You look at all these 2D panels and say, oh there's the fistula, but how do you translate that into surgery when you have to do it in 3D? So this is a Cognard like grade bazillion fistula affecting the patient. And in 3D it's impossible. So we onyxed and onyxed and onyxed and put more onyx, this man has more onyx than you can shake a stick at. and here's the dermatological problem that ensued after a while. So we got this down to a size where there's only one tiny little thing left. So here's the fistula's point. If I, hopefully you can see my arrow. But my endovascular colleague said, well, you know, getting there is gonna be impossible 'cause all these loops, you know, I can't get there. All right, well surgically it's not a problem getting there, but surgically it could be a problem finding this thing, right? So how do we do this? Okay, again, you have five little loops here. All right, let's find those loops in VR and mark all this and paint it in 3D painting so we know exactly. So here's that drain, that exit to the fistula, that's all those loops. We're gonna cover that in blue. We're gonna open the craniotomy right there. And in intraoperatively we're gonna do the same thing. So there's that tiny little incision that we need. We're opening away from the skin necrosis or the skin regrowth, I should say. Open the arachnoid, find that draining vein. Now we have to trace it, right? We have to trace the fistula's point. What I do now? Well let's do the endoscope. Let's follow the yellow brick road in VR, AR, and we simply put the endoscope in and trace it, trace it, trace it, trace it. There you have it. There's that fistula point, naughty, naughty, naughty, hiding, hiding. And we use our single shafted, minimally invasive bipolars, buzz, buzz, and that got disconnected very nicely, and not a lot of fanfare getting this fistula down that way. But I think without all this technology, I think finding it in 3D could be a significant challenge. Not gonna talk a lot about bypass. I'm not a big bypasser, but this particular case was of interest that I think it might be worthwhile showing you the power of the technology. Here is a vertebral bypass, okay. So we got a vertebral stenosis very near the origin of the vertebral artery. And you can see from this is the Truca operation when he takes the occipital artery and puts it into V3. And this will work wonders for this man because of, you know, here's the occipital artery, plug it into V3 between C1 and C2, transverse foramen. Suture in the bypass right there. That's the again, the Truca bypass. And for this man it will work because of his vertebral insufficient. So you can see that the one vert is huge and it's reef retrograde filling the other vert. And he has some symptoms that are very suspicious for VBI. So we decided that this may be a beneficial operation for him. So now two difficult things. One in the isolating the occipital artery is not that difficult, but isolating the occipital artery in the area where it can plug into V3 and finding V3 to dig it out of the spine may be a little challenging. Not so challenging now with the mixed reality. So first step is finding the purple, which is with dot number three, which is the occipital. That is not very challenging. But now dot number three is on the transverse foramen C1. So we're getting the vert prepared now. It guides you right to that foramen. Here's the vert dissected below C1. Now we're gonna next drill C1 transverse process and dig that vertebral artery out of there so that it can be a good recipient once it's out of the spine. So there you go, there's the recipient dissected now. Now you can swing that pedicle of occipital artery into the recipient field and the suturing is relatively straightforward thereafter. So again, this augmented reality guidance of using all the VR implanted markers becomes very, very useful in this particular case. And we had a successful patent bypass. This is the final case I'm gonna show you in the vascular realm. And this is important because of all the anatomical readjustments that you can do with AR. You're almost using AR for guidance and giving up your navigation and I'll show you what that means. So this operation is for taking out the styloid process, styloidectomy for Eagle's. And you can see that the internal jugular is completely pancaked and she has symptoms of idiopathic intracranial hypertension from this. So now taking out the styloid, not a big deal. But the scan was done with the head in neutral position, but the operation was done with the head turned. So we had some significant navigation errors because of that positioning mismatch. And I'll show you in a second. There's your initial calibration of the system. We opened and we dissected the neck and then we found that there was a mismatch because of the position. What we then did in intraoperatively we used what we can see as anatomy and correlated what we have in AR. So the occipital artery and the internal jugular vein forms an X. If we can find that X, okay, that crossing point between the occipital artery and the internal jugular vein in real life anatomy and also segment that in VR in our machine, all we then need to do, even though the position is mismatched, is to match up that X and then our AR guidance will be back in place. So what you're seeing here is the correction of that navigational error, removing the AR signal to match what we see in real life as the crossing point between the jugular and the occipital artery. And once we do that, then we know that the styloid is gonna be exactly where the styloid shows in AR. So boom, it comes out. So this gave us the idea that if you segment in virtual reality, any sort of surrounding anatomy that you can identify, be it a blood vessel, be it a crossing point between nerves and whatnot, you can actually correct your navigation mismatch intraoperatively. And wouldn't this be interesting if this becomes some way to correct for brain shift? So imagine if you will a giant glioma, but you can identify surface markers around this glioma such as crossing blood vessels, you know, certain anatomical things that you can identify. As you resect that tumor, as long as you can match your AR signal of those peripheral markers as the tumor collapses, you keep matching, keep matching, keep matching, you can theoretically account for brain shift for something like for a giant glioma. That's a hypothetical thing, but this case taught us that we can do this intraoperative recalibration with identifying real time anatomy and correlating with AR anatomy and then the rest of what you contoured in AR shows up in AR, will be anatomically accurate thereafter. Quote, this thing "is a play thing for neurosurgeons charmed by the magic of high tech." I found this sentence in a paper that was written in 2003 and believe it or not, it was written about navigation. So we are now obviously 20 years beyond 2003. How many of us do not use navigation? Is it really a play thing for neurosurgeons that are charmed by the magic of technology? I would say it's not. So hopefully this mixed reality is also not a play thing for us. And in fact, 20 years from now became much, much more prevalent and useful technology. The critical question is this, or are these, I should say. Number one, does this is make you a better surgeon? And number two, what are the benefits of the patients? Okay, now I would try to attempt to answer that in the remaining time. I don't think it makes me a better surgeon. It makes me a faster surgeon and it makes a lot of minimally invasive operations possible. And if minimally invasive surgery is theoretically beneficial for the patient, then that's the benefit of the patient, that the recovery is faster, the pain is less, and so on and so forth. So Citius-Exiguus-Titius, what does that mean? I stole that obviously from Citius-Altius-Fortius, which is the Olympic model of faster, higher, stronger. So the citius is the same. But citius, faster; exiguus, smaller; titius, safer. So that's what I think this technology helps me with. It makes the AVMs go faster. It makes my opening smaller and I know the location are safe because I've done the practice in rehearsal in the office. Now, in an attempt to prove this, we did a, within that 100 case series, we did a 17 case matching comparison. We took 17 cases from the AR, VR cohort and then we tried to find 17 matching cases identical in diagnosis, in complexity and whatnot that we did without mixed reality and compared three things, the EBL, hospital stay, and duration of surgery. Now none of this became, none of this was statistically significant, but it all trended towards mixed AR, mixed reality being beneficial for EBL, length of stay, and duration of surgery. So I think that this is simply a matter of numbers. Hopefully when we get bigger numbers, those p-values will get to be 0.05. But these are matching cases. You can see that MVD, we match an MVD with mixed AR and matched one without, an ACOM aneurysm done with and done without, and all that trended towards beneficial use of it for EBL, length of stay and hospital duration, which then of course translates to a better recovery for the patient. I don't have time to discuss the clinical anatomical research realm using this technology. We have been very busy at this most recently with pioneer surgery and well obviously the transorbital work. The transorbital work, I think I'll say a little bit something, something in that the eyeball, okay, the globe is not something that you can study in the cadaver because the globe from fixation to the cadaver becomes a raisin. That globe, how much does the globe obstruct your transorbital approach? You can't tell in cadavers. In VR we can render the globe as the globe, as the globe, as the globe. It doesn't change. It's the globe, the size of the globe in the live human being. And we can move that globe in VR exactly however much you want by clicking, you know, one millimeter, two millimeter, three millimeter, five millimeter. We can measure the translation of the globe, and that came with it with this particular paper or the torpedo approach, you measuring the globe, how much you need to move it to get certain exposures and so on and so forth. I put to you that that cannot be done in cadavers. The other interesting one that I had to point out was in "Focus" where we used disease specimens, meningioma specimens to look at exposure. Again, you can't do that in cadavers. Cadavers don't have meningiomas. And this particular study studies how much the meningioma affects the approach exposure in those particular patients. Future directions, and this is very important to discuss. Well we talked a little bit about recalibration for brain shift. We can recalibrate the anatomy intraoperatively with AR, matching all the AR landmarks. Maybe this can eventually be used for doing this. The navigated drill thing is something that I've asked for a long time and hopefully it will get delivered to me at some point. You navigate the drill and as you drill, can you imagine if you drill in the middle of fossa floor and in your augmented reality you first see none of the cochlear and then you see a little bit more of the cochlear and you say uh oh, I'm getting very close to the cochlear because it's getting more and more opacity on your AR with that drill. Uh oh, let's stay away now, let's stop. Same thing with MC, with that, with the carotid. If you're doing a Kuwasai, you're drilling close to the carotid, oh oh, now it's bright red, better stop because you're getting very close there, better stop. So a navigator drill in that regard would be very, very useful with augment reality if it ever becomes doable. I'm gonna show you a final case and tell you how exciting the future might hold with this technology. This is a very straightforward case, osteoma. Why am I showing you osteoma at the end? Very unglamorous, but it shows you the power of the technology. Here's an osteoma that you would have to cut out. Okay, you're gonna have to scan the patient with the tumor out to generate that peak implant for your implantation. Well, not so much if you're using this technology, right? You cut it on VR, okay. Now you send this model to your peak maker and say, make the peak to fit this. They then send you the peak in an SDL file and you put it in in VR and say, okay, it fits. Okay, go ahead and manufacture this. Now you're gonna do the cutout intraoperatively exactly the same way you did the cutout in VR because you have to make the implant fit, okay. So we're doing the exact contouring of that, the drill out of the VR model in this real life. And then lo and behold, the implant actually fit just fine. Okay, now again, not a big deal, not a very glamorous kind of a thing, but why am I showing you this? Imagine if you will, that at some point we can custom make aneurysm clips or spinal implants that you say, okay, I need a particular clip that has this kinda curve and it's double shaped curve and you know, and so on and so forth. And you can do the operation this way and then you have a custom implant of a clip that's bespoke, if you will. So this is all art of warish. You know, you engage the enemy, you set up the win before you engage the enemy. The VR lets you generate the operative template so that you can use it in AR to engage the enemy and win. This is that paper that showed that custom implant in peak. Remove the tumor, model the skull, create the prosthetic, confirm that the it fits in VR, and then execute the excision of the tumor with AR to make sure that your implant then fits. So bespoke implants, I'm very excited about this idea with bespoke aneurysm clips. If it ever comes to that, be very, very cool. The work is impossible to do alone. These are all my helpers. I'm afraid I don't have pictures of all the other ones, but my fellows and helpers from all realms are to be acknowledged for their work in all this that we have done in the last five years or so. So I hope this translates through Zoom. I am very excited about this tech as you can see. At the end, does it make me a better surgeon? I think it probably does. I don't think that I'm, you know, I'm not God's gift to neurosurgery by any stretch. And I think that with this technology, I'm just that much more confident in making small, small openings to enhance patient recovery and hopefully that obviously translates to better recovery for the patient. We have yet to prove that small openings and less blood loss is actually beneficial to patients, but intuitively it makes sense. That's why the minimally invasive crane of surgery is catching on. And I'll stop there where Dr. Cohen has questions.

- Beautiful lecture, Walter. I really enjoyed it. Great pearls, very nice technology. There is no question that this technology is going to transform neurosurgery. Is it gonna be exactly in the same form? Are we gonna improvise as we go along? I would say we're probably gonna improvise. There're gonna be injection of AI. There's going to be other, you know, virtual twin models, AI and language models and knowledge that come in to really make this technology much more user-friendly.

- No doubt.

- So I do believe the path forward is consistent with what you're saying. I think there will, as I said, there will be a little improvisation going on. My question for you is, with this technology, and this is a fair criticism and in no way means that the technology is not good, it just means that we have to do better to employ it, is the amount of preparation that goes on for segmenting these manually and all the scans, everything else, the help unit in the OR to help you, it can be significant. So for example, for that petroclival meningioma or let's say for a mid-level difficulty tumor, how many hours of preparation do you put in to get ready for that case using this technology in the OR?

- So, so look, there's no doubt that that is true. You have to segment the cochlear, segment the labyrinth. You have to segment the seven, eight complex, et cetera, et cetera. You have to segment the tumor. Sometimes we have to segment the whole ventricle, you know, for whatever we are doing, call assist or whatever. It is very time consuming. I would say that that the complex case that we are gonna rely on this technology, it takes on the average about an hour of prep time, right? 'Cause you you discover that, oh wait a minute, I need, oh for AVMs even more, right? 'Cause you have to, the surgeon has to go in there and say, okay, now I really mentally need to untangle this Medusa head. But in doing so, you actually learn the anatomy and digest the anatomy. But I would say that for non-AVM cases, where you're just contouring things around the tumor, an average of an hour, implantation of the markers and the trajectories and that, and this and that. Intraoperatively, you also need the helper because just using pedals to turn on and off, all these markers are no good. You need that person in there that you work with day in and day out and say, I want cochlear on, I want cochlear off. I want seven, eight on, I want seven, eight off. Show me the proximal ICA. No, no, that's distal, show me the proximal, you know. And so and so, back and forth, back and forth. And not infrequently we actually discover intraoperatively that, oh wait a minute, we didn't actually segment a piece of anatomy that is critical, right? So can you quickly do that please? An example was that Eagle syndrome where I said, I noticed the occipital artery and the vertebral artery crossing that could be useful. Do it quickly, quick, quick, quick, do do, do, do. And then we do it and then boom, you're doing it. So you have to have a relationship with that person who's with you in the OR. And of course that jacks up the price, right? Because you have to pay that person to be there. So expensive, no doubt. But you know, if this matures, you can, something that we haven't touched on is using this for communication between continents, right? So you can be using this to guide an operation in Burma, you know, because you're both seeing the same thing and you can draw on the AR and they see the line that you draw, oh, just follow this, you know, something like that. So the expense is real and the time consuming part, absolutely.

- No technology in neurosurgery has had a smooth path forward, none. So this is not an exception, but it's true that is the future. So what formula will take in the future, I'm very excited to see it.

- Yeah.

- I wanna congratulate you for being such an incredible trail breaker in terms of getting this technology to the next level. You have been a pioneer in it and I want to congratulate you.

- Thank you.

- So Walter, thank you. Incredible work you have done for academic neurosurgery. It's a role model and I wanna really thank you for being with us and looking forward to having you with us again.

- Thank you so much.

Please login to post a comment.