Seeing is believing: iMRI and the future of precision image-guided medicine

Episode Transcript

Recorded live from the Cook booth at SIR and featuring leading experts in the field of interventional radiology discussing a wide range of IR-related topics. This is the Cook@ SIR Podcast Series.

Good morning. We’re live here at SIR, excited to be chatting together today. Today we’re going to be talking about interventional MRI as a follow-up to some of the conversation we had last year. And if you’re interested in some of the things that might be out in the future of image-guided therapy, I’d love you to listen in today. So my name’s Joshua Krieger. I’m a part of the iMRI Leadership Group. I look after our research, clinical insights, and market intelligence.

My name is Riad Salem. I’m Chief of Interventional Radiology at Northwestern and Vice Chair for Image Guided Therapy. My areas of interest are in interventional oncology.

And my name is Kush Desai. I’m Associate Professor of Radiology, Surgery, and Medicine at Northwestern in Riad’s section. I’m the Director of Deep Venous Interventions, and I’m the system Medical Director for Supply Chain and Value Analysis at Northwestern Memorial Healthcare.

Thanks so much. Just a pleasure to be here with both of you today. Just a couple quick housekeeping things. Disclaimer, as we are discussing things today, we’re not talking about any commercial products. And if you hear us speak about procedures that we’ve done within the magnet, we’re talking about non-clinical use there. As far as disclosures go, mine’s pretty straightforward. I’m a full-time employee of Cook Medical.

I guess for me, relevant is I am a consultant for Cook Medical and other types of device companies.

And same here. I’m a paid consultant by Cook Medical and other medical device companies as well.

Thank you so much.

So I want to look back. So we were able to chat last year, and that was kind of a turning point in the program. That was the first time we had spoken about this publicly. And so I want to look back to—just really quick—where we were at that time. We were just really focusing on “Is this viable? Is this something where we can really start to push forward knowing that there’s clinical value here as we continue to build the program?” And I think we’d reached a good point and were beginning to talk about it, but maybe if you could walk us through. I know that our first procedure had some challenges, do either of you want to speak to where we started just about a year and a half ago?

Yeah, it’s interesting. It’s sometimes difficult to remember how challenging it was, but now that you bring the memories back, let’s see, we couldn’t find a wire, meaning that access was done. We have a wire, but how do we get from the right internal jugular vein into the infradiaphragmatic inferior vena cava? And what in clinical practice takes usually seconds I think took us over— hours really. And now we’re here not that much later where that is now the afterthought. So we’ve come a long way.

For me, the excitement is that there were some apparently sort of insurmountable hurdles of visualization that we had to get through. And now those have become an afterthought. So it just sort of speaks to the tremendous advancements you have made in developing this kind of technology, but really move things forward big time.

So 18 months ago we did that first non-clinical procedure. It took us about four hours to find a wire guide. I think the second one we came back and we had a vascular implant inside of about an hour, which was a wild improvement. And then that third one we stepped into and did our first crossing over to the portal vein. Didn’t set a stent, but that was kind of where we were a year ago. And maybe want to take a little bit of time and just talk about the last year, what have we really been working on? Where are we now, a year later? And maybe start with something that’s been just amazing to watch, both incredibly skilled with image-guided therapy using CT, using fluoro, using ultrasound. But those all have kind of established ways that you’re viewing the body. And I wonder if you could speak to, maybe Kush on the vascular side, could you speak to what MRI has opened up as we’ve really thought about what are the imaging planes that you need, what’s the speed and the quality that you need at each procedure stage?

Yeah, so I think from the very beginning we realized for this to be successful, we have to be able to make the translation of the technology from what we’re doing with fluoroscopy currently in the vascular side to MRI as seamless as possible. Certainly we want to harness and elevate the benefits of MRI and all it affords, but it has to be almost that somebody could walk in and within a relatively short period of time, it’s like, “Oh yeah, I got this.” And that, as we kind of peeled back the layers, it’s not easy to do. But what it does offer us, now that we’ve not only refined device and imaging and all the other things you don’t think of, communication suite, monitors, all that kind of stuff, is that now we’re thinking about, really focusing on, okay, now we know we can see wire guides, we know we can deliver stents and devices and whatnot with MR-guidance, but what can MR-guidance offer us that flouro simply can’t?

And in this case, I’ve said this before, it’s objective metrics following before and after an intervention, whether it’s opening a vessel or indeed closing a vessel. And those things we kind of take as surrogates in fluoroscopy and angiography and we impart sort of subjective learnings from our years in practice. And here I very much think we’re going to come out with a number saying we improved flow X-fold. And that’s a powerful thing, not only in the daily care of patients, but looking forward in progressing how we take care of patients in the next generation of interventions.

I think Kush brings up a very important point, which is our task, our mandate in the next year, two years, three years, is going to be to ensure that comfort level that he was talking about, that somebody that is not comfortable with this can walk into that suite that you’re developing and can be well-versed. Use the skills that he or she already has as an interventional radiologist and be comfortable in five minutes, 10 minutes, 15 minutes. Because the basic building blocks of what you’re trying to do are embedded in the suite, but then their skill set with their ability to look at coronal and axial and sagittal imaging is immediately implementable. So we want to really work on that and that’s really an exciting component.

Yeah, absolutely. I want to touch on the suite in a minute, but you talked about a number of different views that you can use. I wonder if Riad, maybe you could speak to maybe first just with the precision needle guidance work, some of the work that would do the biopsy or trocar placement prior to ablation, the differences that you see with that kind of orthogonal biplane that we are working with in the MRI as compared to some of the other modes and then that imaging that you’re getting out of the MRI as you’re tracking your needle down live.

Yeah, so one of the things that I talked about yesterday is that the concept of lifelong learning, and I think this is emblematic of what I was referring to, which is the ability now to visualize in multiple planes, those that we are not usually used to visualizing in axial coronal in some cases and now sagittal, but now sort of the off-axis angulation that we use, which is optimized for the person and for the lesion size and for the lesion location.

So what this allows us to do now is to really identify the optimal angulation and angle of attack. And then in the world now of precision medicine and biopsy and specific biopsies and repeat biopsies, this is going to become more and more important, more and more of a request by oncology to biopsy a specific component of a specific lesion because of resistance to chemotherapy or immunotherapy, et cetera. And so the fact that we were able to really align that guidance pen down that orthogonal plane unique to that lesion, target a very small lesion, five millimeters, four millimeters, with live advancement and confirmation that we are there in less than 60 seconds is a game changer for that patient population. No doubt about that.

And maybe either of you could speak to something there that I don’t see often in other modes: the ability to, when we’re running live, we’re running 3, 4, 5, we’ve pushed the machine to 15 frames a second, in some cases—real time, no reconstruction lag. But then once you’re there, once you’ve got a tray across the lesion, being able to take 20 seconds and have a very high-quality confirmation image.

Maybe I’ll start. So the reality is what we now have with these tools are high-resolution, high-detail imaging that we can get in less than 30 seconds. So one of the things that’s going to be important, at least in the biopsy immunotherapy realm, is confirmation of location, confirmation of tissue, et cetera, and the fact that within 20 seconds you can get the highest-contrast tissue resolution that we don’t have with ultrasound, CT, and obviously not fluoro is again something that’s going to really impact our ability to take care of these patients.

Let me extend that to therapeutics. So we’re talking ostensibly about needle placement, right? And needle placement is of course important for biopsy. And we talked about this before, that small lesions, we’re being asked to do these difficult things, but now consider for example, the fact that we want to treat that lesion, we want to ablate that lesion, and there’s inherent uncertainty with the current modalities of what are the margins of that lesion. It kind of gets fuzzy in the corners, and MR, as we’ve seen, it’s sharp, that’s the margin. And what does that mean in terms of treating a lesion? Yeah, lower risk of recurrent disease of course, but in lesions that are in parallel spaces, as Dr. Woodrum has shown, you’re talking about I can get away with just skirting the edge of the aorta, for example, not hurting it, the edge of the ureter and I can completely ablate the lesion, margin to margin, without damaging things I don’t want to damage.

I think the work that Dave Woodrum is doing, the outstanding work that he’s doing and the confirmation of the ice zone is really a game-changer. And I think underappreciated. I think some of those images tell a story that has never been told, the ice ball, the ureter that’s two millimeters away with clean plane and no inflammation is not something we’ve seen with CT because of lack of contrast. And it is a story that needs to be told, and I think it’s going to be fascinating when more and more of these images get released and people see that it will significantly change how people at least address that patient population.

And I’ll throw out there, does that change how guidelines are then promulgated moving forward? That you move from surgical resection, for example, being first line to an ablative therapy that has very distinct margins, that you are not waiting necessarily for path to confirm that your margins are clear in a frozen section for example. This has the potential to change how specifically cancer therapy is delivered.

And having the pleasure of seeing some of those cases and discussing with the two of you and Dr. Woodrum some of those images, it’s remarkable to me when there is that confidence in the imaging in the treatment, the confidence that gives him and the confidence that then gives the patient that they’re receiving the treatment that they really genuinely believe they are.

It also inserts a level, and Kush alluded to this before, a level of objectivity that is clean, and you can get objective metrics of here’s the ice zone compared to the normal tissue. That doesn’t exist, again with the normal tools. That’s huge. Objectivity and reproducibility is something that we all need to strive for, especially in the oncology space.

So reproducibility or maybe bring that back to a subject that we touched on. Reproducibility of experience in this suite is something that Cook and Siemens—this is a joint partnership between Cook and Siemens MRI, we’re working with Siemens ultrasound—this is really a team effort and with physicians like yourselves to ensure that this is clinically valuable. But one of the things we’re looking to do is really standardize that suite to bring it to a place where it can be rolled out and expectations for that suite can be consistently met, that it is a good solid functional experience from day one. And I wonder if you could speak to— you’ve really seen, I think more than anyone, the evolution of that suite. You were with us on the first procedure day, and you’ve seen it in its current form. How have you seen that change, and how has your comfort level walking into that suite changed because of that standardization in the changes towards what is really a state-of-the-art CT or cath lab with an MRI at its center?

Yeah. Well, I would say that there’s no doubt that at the beginning—a year, year and a half ago—there was a level of discomfort, of unfamiliarity, which is natural, because we currently as interventional radiologists have setups that are pretty standardized: the patient position, the monitor position, where I have to put my hands to intervene. And the reality is this is going to be successful the more we can really replicate that feel, that environment, with some necessary changes, I mean, some extra monitors because of the way the procedures are done.

But I think what I’ve grown is to become extremely comfortable with the new sort of paradigm of where the monitors are, which is relatively similar, how I will use my hands to intervene, where I will look, how I’ll communicate, how I don’t say “fluoro” anymore, I say “scan.” But it’s quick, and we’re built for that. We are able to adjust. That’s one of the really great things of interventional radiology in this kind of collaboration is we are built and hardwired to really adjust to these types of live changes. It’s been less than a year, and it’s moved— it’s advanced significantly. I’m quite comfortable now, and I think teaching that, again, is going to be extremely quick to trained interventionalists.

So as we walk into that suite, I want to give people an idea of maybe a couple of the procedures and where they’re at today. So maybe we’ll talk quick about some of the time-to-target needle work we did here recently. Maybe step into where we’re at on timing and that with a vascular stent placement. And then I think maybe talk for a minute about TIPS and then some of the future physiologic kind of information we can draw out of that. So we’ve got this suite, and I wonder if you could speak to some of the recent time-to-target work that we’ve been doing. As you would convey to someone who’s not been able to participate, you’re walking into this suite, you’re picking up tools that look familiar, and what is that workflow, what is that process, and how is that different than a typical CT workflow?

So with CT, you are bound largely by plane. You’re bound by the axial plane. And I suppose you can tilt the gantry and get a little bit of curve, but let’s say we have a liver dome lesion. I think that’s probably the most relatable for most folks. If we have a liver dome lesion, and we want to try and do something with it in CT, you are taking really a leap of faith to pass it from a lower spot, non– maybe transpleural reflection, up to that higher spot. And there is inherent error that is introduced with that. The difference here in MR is we were able to take a very small lesion and find the, as we had just mentioned, the plane that’s going to take us there, which is an unnamed plane, an oblique plane, and put the pen down and ride a needle up to it in well, frankly, seconds. So it changes the workflow. And I would submit changes the safety profile.

I mean, I pulled some of the timing off of this last one, and we weren’t operating full sterile, but from the time that we did a planning scan to the time we walked out confirmation, I think our average was somewhere in the six or seven minute range.

That’s right. I remember that. We were able to improve on that very, very quickly as our comfort level grew, as sort of the in-room flow became more familiar to us, those numbers dropped significantly. And it dawns on me as Kush talks about sort of the oblique plane. One of the things that I teach my fellows because we’re limited by plane, is I try to teach them that they have to triangulate in their head, which many times is true. But with this, the triangulation is done for you with this oblique plane. This is the oblique plane I’ve been referencing indirectly, is find that right plane, but you have to do it in your head. And so it’s really interesting. But no, those time-to-targets were dropping and I think it’s an important metric for us to continue to work on and report because people with this kind of technology will be worried about room efficiency. There’ll be a skepticism over that, but if we’re talking in the minutes, again, this will assuage everybody’s concerns from that perspective.

Yeah. Kush, there was one other thing that you mentioned the other day that I’ve really been mulling over and that was maybe we don’t appreciate always the value of a confirmed negative. And the level of accuracy we’re able to get and the level of soft tissue contrast we’re able to get in these non-visible liver lesions and other locations. I wonder if you could maybe expand on that.

Yeah, let’s frame the problem. So the current problem is that let’s say we have a really small or really difficult lesion to get to, all right? And let’s say that using current technology, whether it’s ultrasound or CT or both, that you do your best. You get there, you get samples, and it’s negative. Then the question is, do you believe the negative? And if you don’t believe the negative, a dispassionate way of looking at it is why’d you do the biopsy in the first place? If the specificity of what you’re doing is low, then you shouldn’t maybe do it or you should consider an alternative pathway. I think this is primed to turn that whole discussion on its head because now you can get to really small lesions, you can show that you’re across the lesion because you can see the margins of the lesion better and get there rapidly. Risk of bleeding, et cetera goes down. And if it’s negative, you can document that I am indeed across it and sampling error is as close to zero as possible.

Yeah, it does turn this whole process on its head because in fact, there are many lesions that aren’t even attempted because they’re small. So there’s a huge opportunity right there. And as Kush is saying, if it’s a small lesion and your result is negative, our default is to not believe that, is to be suspicious of that. And so you address the unmet need of small specific lesions, you then satisfy the unmet need of disbelieving a negative sample in a small lesion, and you’ve just addressed a huge unmet need here because this happens and lesions are getting smaller and smaller and accuracy and specificity of what you are actually targeting, that affect therapy, is critical.

And I would just say we’re still forming some of this, but we do know that as we move this towards commercial availability, that needle portfolio is really at the forefront of that, both for biopsy and then access towards future transcutaneous therapy, stepping into another layer of complexity, something that we’re working on very diligently. Kush, I wonder if you could speak to some of our work in the vasculature. And a year ago we were able to place a stent. I think we were starting to get comfortable with that. As far as vascular navigation and the delivery of therapy, what are we playing with today?

Yeah, I mean it’s no longer about the “can we get there?” and “can a wire cross the lesion?” It’s about we’re visualizing it, optimizing the device delivery, making the device basically one that I would not be able to separate from the devices I use today. There’s no compromises being made. It’s a non-linear growth in this area. I mean, there’s work to be done, there’s no question. But it’s non-linear growth in this area. And I think even the most hardened vascular practitioners would have to admit that the promise behind using a truly physiologic modality that has the ability to do, especially with modern computing techniques, analysis of flow, is one that can’t be ignored.

And I’m excited. I mean the research on that physiology side, it’s certainly not all being done by Cook, but we know of researchers around the world who are looking at how do we do this both before and after procedures? How do we start to type thrombus so that when you’re going in to clear something out, what you’re going into? How do we pick up that flow through the stent so that we’re starting to gather what is an effective threshold? Maybe we’re not looking at— 50% is the number we throw around for things anymore.

For convenience, for simplicity.

Well, because we’ve not had, that’s right. We’ve not had the data behind better vendors. Yeah.

It’s sort of like— take claudication for example, arterial claudication. We just assumed that if somebody was a claudicant, that you’ve got to open the vessel. Well, that’s been turned on its head, right? There’s very clear data that exercise programs actually do better than revascularization in a lot of people. And so the implicit assumptions of a 50% stenosis being something that is hemodynamically valid, we have the ability to turn that on its head with actual flow data. Yeah, I’m very much looking forward to that.

And then, Riad, I wonder if you could speak, we had just gotten across into the portal vein a year ago. I think this last TIPS case to me was just a ton of fun to watch. I wonder if you could speak to where we’re at with that today, not because it’s the next thing we’re going to have out in the world, but just as an example of what the technologies are capable of and how the familiarity with them and the visualization is changing your interaction there.

Yeah. I mean, I think my first love is, of course interventional oncology, but a very close second, and maybe a tie for first, is portal hypertension and TIPS. And the reality is when you really look at how you do TIPS, these are all surrogates. These are all surrogates of where the vessels are located. You don’t really know. You advance wires and catheters because the you know vein should be there and the cava should be there. But now when you have live imaging of the vein in real time, constantly while you are imaging simplifies the procedure. And that’s one of the fears. TIPS is sometimes touted as the most difficult procedure. I don’t necessarily agree with that, but that’s the reputation. And with that comes the idea, well, how are we going to teach people to do TIPS and how can everybody do that? And I only do two a year, how can I really gain that expertise?

So now what this does is neutralize and counters that narrative and deals with that learning curve necessity. If you always know where the hepatic vein is and you can place a catheter under visualization and where the portal vein is because you’re seeing it live, full time, that’s a game changer. So yes, I mean, there’s still— we’ll need other components that we’ll need to study to really have TIPS as an implementable program in an interventional MR. But we’re going to get there. But it just shows you what the potential of actually seeing inside the body, the organs— Again, right now when we do things under fluoro, it’s all by surrogates. I know that this wire’s going this way and that way, hence it should be in this structure. We don’t see the actual structure. This is different.

And just to paint a picture for the people, this is an audio medium. When you were doing this last, what I was seeing on the screen is a coronal image of the IVC and the hepatic vein coming off where your needle was about to exit. Next to it, a full coronal of the portal vein, and then next to that, an angled sagittal that was joining the exit point of the hepatic with the entry point to the portal and just walking down it.

That’s right. We were simultaneously looking at coronal and sagittal images—live—while the needle was advanced. So take ICE techniques with ultrasound, intravascular ultrasound techniques. This is a thousand times better because you’re actually seeing the true needle and the true structures in all of the planes simultaneously. So again, it sort of flips everything on its head when you think about live visualization inside the body at all times.

So you talked about removing some of that fear. You talked about giving people—instead of relying on surrogates that comfort is built over time—giving them direct imaging. I wonder if you could speak about—as you look at this, maybe first in that transcutaneous needle space—how you think about training the fellow. If we put one of these in your facility, training your fellows on some of these complex needle procedures on training younger attendings on some of these procedures: how does that change from the current paradigm?

Yeah, I think it’s nothing new for us, right? Because I’m training fellows on things that I was not trained to do. So the spirit, the soul of interventional radiology is venture into the unknown and innovate and sort of make it up as you go. So I think that the spirit is very much alive. It’s really not any different than it was when Dotter invented the specialty. And all this is, is how do we think about doing it maybe a different way, and maybe the things that you just assume, as Riad just said, I assume that my wire’s to the right of the spine, so it’s in the inferior vena cava because I know I accessed the vein. Well, that assumption no longer needs to be actually had. You are in the inferior cava. I’m oversimplifying it, but you just have to rewire how you think.

Yeah.

Yeah. I mean, I think the fact that as interventional radiologists, part of our training is the diagnostic component, where we know the body from head to toe in three dimension and what it looks like on imaging, make us uniquely qualified to take fellows that— again, as Kush was saying, we’re training things that weren’t available, and we have an angio CT unit now. We have intravascular ultrasound. We didn’t have that, and now we have yet another tool that they’re going to learn very quickly because now they’re going to add another dimension, the sagittal dimension, the live dimension. So we are hardwired for that. And I think embedding this as part of our training program is going to be no different as to how we’ve incorporated other things. And our trainees and our specialty is very well-equipped to do that. This is what we do.

Yeah, absolutely. The hope there is, if you’re talking about that lesion up in the dome where previously you had a mental unknown, unnamed oblique plane that you formed over 10 years of practice, 20 years of practice, that instead of taking 10 or 20 years of practice, it’s simply there.

Yeah. I won’t be saying, “Triangulate in your head” anymore.

No.

Because Kush trained with me. I’m sure you heard that 50 times. You don’t have to do it in your head anymore. You can just triangulate. You have the image, follow down that oblique unknown plane, and you will get to that target.

So maybe as we look ahead and begin to wrap up, I wonder if as you think about the spaces that you’re in, as you think about the practice at Northwestern, if we dropped one of these complete suites with the needle portfolio in the future with the vascular portfolio, what are some of the things that are interesting day one? What are the things that would become clinical practice?

I mean, for me, there would be an immediate sort of restructure of how we perform the biopsies in the immunotherapy patients. So that’s the easy win. It’s a quick win. The cancer center will love it. Patients will appreciate it, and our ability to treat specific disease is going to be important.

The next thing I would say, keeping along the oncology space, is ablation. The hepatic ablation work now where we always have to look for confirmation of margins. And as you know, now it’s done with sort of 3D tools and 3D merging tools and fusion tools. Well, I don’t need to merge or fuse anything. I’m going to get the actual margin. So it’s not going to be a mimic or sort of reconstructed, it’s going to be the reality. And then for us—this is where Dr. Woodrum and Mayo Clinic is far ahead, the prostate cancer—we’re doing a lot of research in prostate cancer now. And I think opening up that avenue and showing how with a well-oiled system and a referral base and a patient population that needs it, this is where we would go. So the oncology space, at least from my perspective, will keep that scanner busy full-time.

I would agree with that. I mean, I look forward to the day that vascular is a mainstay in the MRI suite. And then we’ll have to get a second one because the oncology space is a very clear front-runner, especially right now. But it’s going to change the way we treat patients. I mean, imagine a future, and maybe, I don’t think I’m being too dramatic, where radical prostatectomy is, maybe the minority of patients. A very morbid procedure is the minority of patients, and an outpatient treatment procedure is the majority of patients. I mean, certainly as a male, that’s a feature I look forward to.

Absolutely. I mean, as a person, all of us, cancer looms as a part of modern society, that just seems objectively better.

The other thing I might add as I think about how well the prostate is visualized with this tool is, look, I’m a big fan of prostate artery embolization for BPH, but I don’t know that with such specific targeting of a cryo needle sort of in the median lobe where it’s hypertrophied, that one application of a cryo will do the same thing, maybe better, faster, I don’t know. But huge place for investigation because prostate artery embolization is challenging. It’s technically challenging.

Yeah.

It’s easier with a good magnet to place a needle in the median lobe than to do a complex prostate artery embolization. So even some benign applications.

And just as we are looking forward, one of the things I loved is you both came in—I think you’ve been really open about it—a little skeptical of the space. As you think about it today., there are certainly parts of this where we’re still building this out, and there’s a lot of work to be done. As you think about what you’ve seen, are we five, 10 years out on this still? Or is this really something a lot closer?

Yeah, maybe I’ll start. I mean, I would think that that’s going to largely be dependent on the ability to develop devices that are imageable and really optimization of room flow and operations and magnets that can be installed. To me, we’re in the three- to four-year range where this is clearly something that can be inserted, multiple centers can be trained. So I see the light at the end of the tunnel at this point. I wasn’t there a year ago, but I’m clearly there. And the pace, I may change my answer next year at the podcast because things are moving so quickly. But I think that’s a fair representation, and I think it’s a controlled delivery and development of a therapy and a modality, and I think you guys are doing it the right way.

Yeah. I’ll just give an even more qualitative answer, which is that I agree with everything he said. This isn’t the first time that interventional MRI has been attempted.

Absolutely.

And we’re way closer than we’ve ever been before.

Well, I think we’ve mentioned Dr. Woodrum and a number of other sites today. There are folks that are doing this, that have strong-armed this into existence. And our goal, certainly on the Cook side and with our Siemens partners, is to reduce that friction, to be able to provide a full suite from the start of construction to the disposables, and really excited as that continues forward. Maybe I think we probably want to start wrapping up for time, but I just really appreciated the conversation today. I think the perspective that the two of you have coming in for our first procedure, but also seeing what the absolute state of the art is. Even some things that we didn’t talk about today.

Seeing that arc from the clinical perspective, from the perspective of individuals who have really been there through some of the growth of the last 10, 15, 20 years, and the change in clinical practice, I just genuinely appreciate. It is something that as we celebrate 50 years with SIR as Cook is now in its 60th year, and as we have Charles Dotter and Bill Cook on the wall next to us, it is kind of fun to me though that the devices that are behind Bill in that picture are some of the devices we’re talking about today. The last 50, 60 years has built interventional radiology into what it is, and we’re looking at those devices afresh. We’re looking at what we can do with these new imaging modes and building out some of these procedures for the next 50 and 60 years.

Well, it’s been an honor to work with you, Josh, and to work with you Kush. I mean, again, it’s a lifelong learning and lifelong academic and intellectual curiosity, and this is certainly satisfying on all fronts.

Yeah, I mean, this collaboration shows that when you look, my colleague, Dr. Salem, who’s achieved so much in his career, and even now, he’s excited about what’s next, right? That’s the spirit of IR.

Well, thank you so much. Really appreciate it. Everything we’ve done up to today and the conversation now, and for all of you who have tuned in, if you’re looking to find out more, by all means get in touch. We’re hoping to build the future in this space, and we’d love to do it with you. I hope the rest of SIR goes exceptionally well, and thank you so much.

Thank you.

Thank you.