I really appreciate the opportunity to speak here. It's been a great meeting so far, a very lively discussion this morning. And I hope that sort of thing is going to happen during this talk, because really what I'm going to do for the most part is going to present a bunch of cases, and I'm going to show some pictures, and we're just going to see what things look like, and I tried to pick a few things that might be of interest. If I could only say one last comment, I don't want to delay the lunch break, and I just had one final comment about the topic we were speaking about just before the break, and it's not because I'm trying to have the final word by any stretch, I promise you that, because it's not about that. It's about the issue of how difficult it is sometimes to produce or to perform randomized studies, especially in the context of surgery, where you might be talking about two vastly different physical experiences for a patient. But I would just want to remind folks, or let them know if they never knew it, that the way that that was actually performed in one of the most important sorts of studies like that, which was the NACBP B06 study of lumpectomy versus mastectomy, was to have something called pre-randomization, because the idea of going into the room and saying, Mrs. Smith, you qualify for a study, I'm going to flip a coin, you either get part of your breast out or your whole breast off, was thought to be just ridiculous, and no way anyone would enroll in that. So what happened was the investigators would go and flip the coin behind her back, or calling for the randomization, and they'd walk into the room and say, Mrs. Smith, we have a study, you would be randomized to this thing, go ahead with it. So it's a concept that is also going to be explored in a study that Bob Timmerman is trying to push one more time to compare surgery versus SBRT for early-stage lung cancer. So that's my only comment on that. We'll get back to the main topic at hand. But it is always a challenge. It's not a fault of the field of neurosurgery not to have done level one studies for every single thing, because it's extraordinarily difficult to compare such a different physical experience. Anyway, on to this particular topic, no disclosures. So we'll just talk in this situation about neuroimaging for AD surgery. Really the majority of it is going to be that first bullet, some standard imaging and some selected case examples. I hope maybe a pearl or two from each of these cases. The sources of artifacts and the special topics are really a minor part of the discussion today. And the first quiz question for any resident is that who is this? That would be King Tut, as a matter of fact. That's right. Okay. Just keeping you paying attention. Also, this is the second Homer Simpson picture, just to kind of keep it going and kind of wake it up. Okay. So I'm going to go through a couple of common cases. You know, they're going to be organized more or less along the spectrum of starting out with the simple and very common things towards things that are perhaps somewhat less common and have maybe occasionally some unusual features in the imaging that you might want to exploit to be able to target them properly and treat them properly. So brain metastasis is pretty straightforward. As a matter of fact, this is not even my case. I'm borrowing it from Jason. It's a case he submitted in a textbook chapter, a very straightforward lady with metastatic non-small cell lung cancer. So you can see brain metastasis, by and large, rather easily on T1 contrast images, usually a thin slice sort of study, axial corona, if you want to get two good views. Now, I will mention, however, it was brought up earlier, something about the chiasm dose. The thing about this particular case, because again, we're always trying to point out little safety pieces here and there, is that this is a situation where the chiasm is certainly in play. The prescription dose in this case, according to Jason, was 16 gray, and the maximum point dose to the optic pathway was 12 gray. Now, I want to point out that it was mentioned that that's obviously one of those sensitive structures we're worried about. We're always focused on that. The Quantec paper on optic nerve and chiasm, perhaps neurosurgeons don't know this particular series of articles. There was a big project put together by APM and Astro. It's called the Quantitative Analysis of Normal Tissue Effects in a Clinic. It was a large group, crowd-written experience of a number of papers that looked at all the published information about radiation dose to toxicity for various sites in the body, including some sites in the brain, and some sub-sites in the brain. So the one that speaks specifically to optic nerves and chiasm graphed the available data in a manner that you would see such as this in the paper, with these two dots being the ones that are the most, I would say, relevant for radiosurgery, because these are the single fraction toxicity doses. And in each of these two dots, there was a low incidence of radiation-induced optic neuropathy. But I do want to point out that this individual dot that speaks to the 12 gray being, perhaps a point where you want to be cautious, actually was a mean dose. If you drill down to the article that's referenced, they were referring to a mean dose of the optic chiasm structure, not a maximum point dose. And that's why in this particular analysis, anyway, the upshot of that is that the QuantTech recommendation would be to keep the maximum point dose certainly less than 12 or so, if at all possible. And that's, I think, was complied in that particular case. Okay, so moving right along, another of the common cases we see is, of course, meningioma. This just happens to be a typical case we have treated in the not-too-distant past. A 40-year-old gal came in with headaches, and she'd had some nausea and vomiting. CT and MRI revealed a mass. She had a couple of other neurologic findings. Fairly straightforward scanning technique was used, an MRI of the brain with and without contrast, a thin slice. And of course, T1 predominantly is going to be used. So I think you can see the lesion in a pretty straightforward way that it lights up. It's right in here and over here. And what I was trying, actually, to illustrate on this one, to just try to bring into play an additional sort of aspect of safety is something you may want to point out. It doesn't really project as well as I would hope. We do have a stretch of the basilar artery that is at least borderline in play. And so although we were going to give this the full treatment, I was trying to find a case that would just be an example of something where maybe a large vessel is at least looming or is at least in the neighborhood. I will say that for this particular case, the chiasm is not really a dose-limiting thing. If you look, this is a sequence of axial images, and this is the coronal slice, sup-inf level of the chiasm. So we're every bit of 6 millimeters away from that for individual slices. So that turns out not to be a player. The only reason I mention the idea that at least it's good to have an awareness when you have the occasional large vessel coursing through a tumor that you can treat it, that you're going to be treating, there are a few cases that have been reported of large artery stenosis. I mean, internal carotid artery stenosis. One example is something reported in 2005 from the Mayo Group, right? This is a couple of photos, the pretreatment lesion and a post-treatment image showing a fairly abrupt cutoff and lack of flow through internal carotid. Now the quantic didn't speak to an actual large vessel dose limit such as this. If you look at the cases that have been reported, they are typically going to have doses on the order of 30 or so gray to the vessel and generally to the full circumference of the vessel. So I think that's maybe something where it would be time to get a little bit nervous. This is a case we treated about five years ago where it was fairly obvious that you had a stretch of the internal carotid going through this particular cavernous sinus meningioma. We treated it with a prescription dose of about 16 gray. And whereas in virtually every situation you might want to try to blast a meningioma and not really care about hotspots inside it, we happened to try to intentionally avoid putting the worst of the hotspots and actually did not try to put so much of a hotspot. We actually kept it relatively homogeneous, unlike many of the cases we treat. And so this is just the 19 gray isodose volume, just we didn't want to traumatize that particular vessel too much. This fellow did pretty well. We also did actually a pretty good, even more conservative, kind of max point dose estimate. And he fortunately did okay. Okay, so we'll switch on to acoustic neuromas. Already a very good talk this morning showing you all the outcomes and the rationale for doing it. So let's just look at some of the pictures. And by and large, you're going to see these things quite well on a T1 study with contrast image, the thin slice, of course. Occasionally you do get a little bit of extra information from a T2, not only because you're probably going to want to be mindful in particular of the dose to the cochlea and related structures, but especially when you already see that you're kind of concerned about that, and you think it's going all the way down the canal. If you get a T2, sometimes you'll see a little bit of a trailing edge of venous congestion at the end. And when we see this, at least we feel a little bit better about pulling back the target somewhat. It just maybe eases our conscience. The point was brought up earlier about how if you're going to treat the tumor, you're going to treat the tumor. If the cochlea is there, you're not going to physically move the cochlea away from the tumor. It's true. But in any case, we do feel a little bit better in pulling back the target zone if we happen to see that maybe some of this was a little bit of venous congestion showing up only on a T2 scan. So this was the targeted volume for that particular person. Now we mentioned this whole idea of the cochlear dose. This is one of numerous papers that speaks to this idea of the concern that the cochlear dose is relevant in terms of adding to your risk of hearing loss. This particular paper just happens to have a couple of nice images of seeing it both on a CT scan with bony anatomy shown right there and a couple of other imaging modalities. Our own particular limit is, I think, as consistent with what Dr. Friedman had mentioned, to try to keep it, of course, as low as possible. And ideally, we do everything we can to get a four-gray mean dose or less. There is also, for the record, a Quantag paper that speaks to radiation therapy and hearing loss. It is largely dominated by an analysis of some of the fractionated data. It only says a few minimalist things about specifically the idea of risk in the single fraction setting. As a matter of fact, it does not actually, in that particular analysis, give a specific comment on a cochlear dose limit. It speaks to the observation in some studies that the prescription dose in this setting to the tumor itself should probably be limited to the 12 to 14 gray range, which I think is now where everyone could consider 11 to 13 or 14 or so would be a pretty reasonable standard of dose. And it should go along with a decent chance of hearing preservation. This is just another picture of it. I mean, there is the fifth nerve is in the neighborhood. We try to pay attention to where the fifth nerve is in general when we treat some of these acoustics. We can't always see it too well if the tumor is a bit bulky and it's not there. But it's also in the neighborhood. And in the context of trying to limit the risk of other cranial neuropathies, there is a stated, suggested upper limit of 12.5 gray, once again coming from the Quantec analysis, to the brainstem, broadly speaking, because that's how it's generally been reported as a descriptor of dose to that structure and at risk of toxicity to cranial nerves. And that's there for the record. So Trigeminal neuralgia, you know, for some reason, I just flat out forgot to put a case in like this. We had a great talk about this. So since I just flat out forgot to put images, this is an artist's rendition, OK, of sort of what you're going to want to do. You know, you've got your brainstem, you've got your nerve coming out, and it was well described earlier that you're probably targeting the proximal section of the fifth nerve. And there can be some debates about exactly how to do it. We typically put it so that we, yeah, you know, this is the other thing about when you actually say what you really do at your institution, you know, you're sort of showing a little dirty laundry and stuff like this, and you're revealing some superstitions and all this kind of stuff. And you can say, where the heck did you get 87? What does that mean, 87? Why not 90? That's a nice rounder number, because 90 was existing around. Well, I can do the, do I have to explain this? Yes, I do. OK, 87 was when they, you know, a number of years ago, before I was sort of involved in the center of that, where we treat on this gamma knife, there was the recognition that there was, in the NRC, the recalculation of the cobalt calibration, and everyone thought that there was off by a couple of percent, and so they adjusted downwards their doses. I can't explain why. But in any case, we generally do it so that the 50th, the 50 percent line is effectively touching the brainstem. So somewhere in the mid-40 gray or so dose line does touch a voxel or two of brainstem, and knock on wood, you usually can get away with doing pretty well with that, so that's my bad version of that. So AVMs, we had another great talk on that, so I'm certainly not going to review all the clinical data. We also saw some differing opinions and different, you know, approaches, fractionation approaches and that sort of thing. I'll just show you a recent case we had and invite feedback or comments from anybody who would have done it differently or prefers to do it another way. This is a guy who had actually had a longstanding history of this particular lesion. It was pretty bulky. He actually managed sort of conservatively. Nothing was going on. It was so big, no one had to touch it. Along the way, he did get one or two embolizations. He did actually sort of get a little bit more and more bothered by seizures, and so radiosurgery was offered. Of course, there was an initial diagnostic angiogram. Now, we did for a time use the sort of system that would actually use orthogonal images of the nidus on CT angiography. We did that for a long period of time and just got away from it. We were lazy. But we're usually pretty comfortable just doing it straight off the T1 MRI, although we have an awareness, of course, of what the angiogram looks like, and in particular, in a sort of bulky case, unlike it was interesting that you mentioned that there's some notions coming out of the backwards thinking, we've usually approached it in this kind of forward thinking sense. And I'm not doing this justice here. I tried a couple times to get a movie of an angiogram. It just was not working out, and it's not happening here particularly well either because there was a little bit of loss of fidelity here. But what I was going to try to show you here was that we use the angiogram and try to figure out where the feeders come in, and if we were going to break it down into two or three stages, we typically treat the feeding vessels first and then progress onwards. And you really don't appreciate that particularly well here, but that was what I was meant to illustrate. But just to point out what that is, in this particular fellow's case, there was a somewhat more inferior and anterior feeding source of it that you'll see. This is our first isocenter right here. Here is the one that we treated about four months later, and here's the one we treated actually about four to five months after that. We're walking sort of upwards, superiorly, and backwards. We think in the direction of flow, if you will. When we do this, tend to moderate the dose just a little bit. We gave a prescription dose of 17 gray to each of these sectors in each stage, and it was prescribed in such a way that it admittedly involves a little bit of art, because if you really wanted to know, it would be a really, really hard thing to know exactly where those doses were. We try not to overlap the 17 gray dose volumes to the best we can, but it's actually a moving target. There's a lot of deformation going on. For instance, what I'm trying to represent here is that this is the image at the time of the first treatment, and this is the image at the time of the actual third stage, if you will. It's pleasing and gratifying that we're already seeing some measure of ablation of the abnormalities in stage one, if you will, but there is a consequent distortion, deformation, remodeling, if you will, of the brain parenchyma to some extent. And in my opinion, this just becomes then a little bit of the art sort of thing as opposed to trying to really say that we have been so millimeter by millimeter precise with lining up or abutting isotope volumes of this level or that level. It changes just a little bit. So it's an unavoidable challenge right there. OK, I'm going to throw in this cavernous malformation one. Now, cavernous malformation treatment has been, to some folks' minds, a little bit controversial anyway. In this particular case, we had a 55-year-old guy who had actually been pretty symptomatic from this particular thing. And well, one time when he was originally diagnosed, he actually had seizures from the angiography. But I don't think that was the lesion itself necessarily. In any case, he had a progressive kind of multi-component neurological decline. He was actually a truck driver up until fairly recently. But eventually, he was getting a sense of disorientation. And it was really not allowing him to continue to work. And so we ultimately offered radius surgery. This particular lesion was actually initially seen just on an initial diagnostic non-contrast CT scan. Later, there was a little bit of funny business on an MRI that suggested something odd about it. We did an angiogram to rule out any kind of shunting things. But it didn't really tell us anything that was less evident from the CT scan originally. And for planning, we just used, once again, a pre- and post-contrast T1 MRI. This is the one. Let's see, so this is actually his initial non-contrast CT scan, which was showing some characteristic calcifications right here. And it's a location that's not especially suitable or particularly easy, in my colleague neurosurgeon's opinion. Bob Brees wasn't too crazy about diving into this particular area. This is, oh yeah, this is my attempt at a movie, right? And this is even the normal one. This is really bad, right? I mean, I just had a sequence of automated slides and showing you the flows. If you missed it, you didn't miss much, right? And that was a dog walking across to show you the different whatever. And this one doesn't even have an AVM in it, so it was really stupid that I tried to do that. But the point is that we just did that as a backup. But we wound up planning it on the MRI. I mentioned that it's quote, unquote, controversial. The controversy, such as it is, would have been, I think, triggered maybe by this paper from the mid-'90s from the Pittsburgh Group. They analyzed their data in this particular way that is shown on the right-hand column. They had a bunch of patients who were treated. And they looked at how long it had been in between the previous episodes of bleeding and treated them. And they followed them. And they tried to make the case that following treatment, there was a longer interval to the next episode of bleeding. And some may feel that that's a quite reasonable way to look at things and a plausible argument. There were some letters to the editor back and forth. And I would be curious to see what neurosurgeons in the room currently do, where they do or don't stand on this particular topic. Because the editorials that spoke to this were critical for a couple of reasons of this particular thing. But do any of you guys, do you have any strong opinion about cavernous malformations? Do you treat them? I guess I'm curious. Do any of the neurosurgeons treat with radiosurgery or anything? Yeah. Anything that would be castigated for treating cavernous malformations, that's right. Yeah. That's why it's always putting out your dirty laundry. I knew it was going to be. This is the first one I've seen treated. And it's the first one we treated in about eight years. And we thought it was a special case because of the location. And what I didn't show you was that there was a, well, OK. So this is just the lesion as it was when it was treated. And a couple of other slices down, what I didn't show you was that what we didn't consider target. But there was a value to getting the T1 pre-contrast, because you could see a little bit of recent blood in the ventricles that we thought was from this. And so we thought there was a component of active bleeding. And we were triggered to do it because of what appeared to be a rather progressive neurological decline. And given all the caveats that maybe it does, maybe it doesn't do the felony good. But we took a plunge and did that. There are still a number of institutions reporting outcomes in this particular condition. But I understand the controversy. And it's not a commonly treated place thing in our institution. I knew I would get an eyebrow raise or two if we even put something up there. So OK, pituitary tumors, well, we had some nice discussions about that also. And this is just a recent case of someone, not too recent, a couple years ago actually, a case of a gal who actually had a functional tumor, had had a couple of surgical resections, and had recurrence. That's actually the typical situation where we're finding things. Of course, surgery is, generally speaking, the initial therapy when possible. But they do sometimes occur in awkward positions. We can generally see them pretty well on standard sequencing. And that's what we did in this particular case. You can see it, I think, enhancing long about right in there. And on the actual setting, it's in there. I was glad to see, actually, this particular paper from Jason's group. Because we do struggle sometimes to actually see and be convinced of what is or isn't the actual residual thing. And there is an experience out of the UVA team where they actually wound up saying they were a little bit, they threw up their hands. They had some metabolic, biochemical evidence of recurrence. And went ahead and treated, essentially, the entire cell without being necessarily convinced of a focus. And reported a pretty good outcome in this particular situation. Of course, some hypopituitarism from whatever function had been in the remaining gland. But reasonably good efficacy in terms of improving the actual abnormal hormone secretion. I'll just show this particular picture. Because we, up to this point, hadn't even mentioned that other modality that has occasionally been used for radiosurgery. There'll be a paper coming out. It's in press now on a proton radiosurgery experience. And I think it's interesting because, well, first of all, protons were, if I'm not mistaken, maybe actually the first particle that was used for radiosurgery at all. The thing about it is that in those old, old, old experiences in the late 50s or 60s, what they actually did there, my understanding of it, is that everyone knows the whole notion of a proton having a Bragg peak and a spread out Bragg peak. You adjust the energy that goes in. My understanding is that that first wave of effort with it from Lexell himself, I believe, they didn't have necessarily enough imaging to guide the treatment. So they were actually blasting effectively the Bragg peak through and through. So it didn't work out too well. It was abandoned for a long time. Came back again in the 80s and 90s. So it's still interesting because when you look at that dose distribution, well, you can choose to think it's interesting or not. Certainly, they came at it with a couple of beams. It certainly got a good stopping point of the beams, as expected, when they go in. But I would say that the generation of proton therapy that's being employed here is not exactly giving you a particularly sophisticated dose distribution. The other thing you'll see, again, coming out in press, not this particular paper, but a couple of others around it, will be a similar experience actually using particles and characterizing some dose-volume relationships with a little bit of temporal lobe necrosis. So you'll see a little of that coming out in the literature. And I don't think it means that protons are bad or not usable in this particular setting. But I think they are not immune from all the same dose-volume considerations that we have to be careful with in using photons. So look for that one to come out there. I'm just going to put this one up there. It's just a nice study that came out last year. Always impressive to have these very large studies. This is a nice study by the cooperative group, the Gamma F group. And we're certainly hoping to get numbers like this in the forthcoming registry. I hope we can. So this is a nice, large, big data kind of experience. So glomus tumors, I don't think it was on the agenda. We don't treat a lot of them. But we get occasionally a handful of glomus tumors coming in. We have an ENT actually who sees most of them. And they kind of come through his clinic primarily. But there is an experience of treating them with radiosurgery, of course. And so I'll show you a couple of pictures of people that we treated. This particular gal, 57 years old, she'd actually had initial surgery. And this is a recurrence. We've treated some de novos. But this one happened to be a recurrence. She initially was unclear. But it was a pretty unequivocal growth on MRI. And some symptoms go along with that. And the only reason I'm picking this one out is because there was a brief phase of interest from a neuroradiologist who was particularly keen in doing MRN geography for these lesions. And I just happen to have a couple of cases with pictures like that. So I'm just going to show you there. So these are just the basic sequences there. You can see a little bit of a signal on T2. And I realize it's not showing up too, too well projecting across the room. But it is visible on T1 with some contrast enhancement right there and there. But these are the kind of funky images, like I say, we were playing around with for a while. This is just a coronal image from one of the coronal images that was put together to make a maximum intensity projection or a MIP version of this. And I did not try to do this in a spinning fashion. But of course, this would spin around if we had it. So the lesion is right there. And if you were to spin this thing around, you would also see it spinning in space. There is a faint contrast enhancement. So as you spin it around, you get a better appreciation of it putting a little bit of pressure on the vessel right around there. We got into the habit of treating these folks based on not an original idea. We borrowed the dose of 15 gray from really the Mayo experience and did that. And I'll just show you this one other picture because this is just a little preliminary series we put out a while ago. But we actually did include this particular patient, who as luck would have it, is the one patient who actually had 16 gray instead of 15 gray. And she's the one person who had toxicity. I don't think that that was an abrupt dose cutoff threshold. I think there was maybe just some bad luck in her particular case. She happened to have an acute worsening of some of her symptoms. She was having some 12th nerve symptoms, as I recall. And she was actually hospitalized. And because we were still a bit new at the game, we actually imaged her in addition to giving her steroids and all those usual things. The only interesting thing there is that what you can see here on two days afterwards, I don't know if you can appreciate that there's a pretty good, you know, I'm lying because this is another patient we treated with that thing. Oh, right. No, that's the patient we treated that I showed you before. I apologize. This is not this patient's images because her tumor's right here. And what you can see is actually some necrosis inside the lesion within two days. And so it's necessary on the grounds of Memorial Sloan Kettering to acknowledge, since Josh is probably in the room here too, that this group, led by Rich Kolesnik and all, have certainly been the chief proponents of the notion that a lot of times with high dose perfraction treatment, we're getting a very abrupt vascular effect. We're getting a lot of epithelial apoptosis. And often you can see a very quick result in terms of a vascular breakdown. So I think this would at least be consistent with that. It's plausible to think that that's what we're seeing. Only 48 hours afterwards, she eventually recovered. But we just happen to have that particular image. And we don't usually image people that quickly afterwards for any other reason. OK, so just in the last few minutes, we'll talk a little bit about just sources of artifacts for this. And I'm not going to overly dwell on this because I'm sure that some of you are going to hear, I'm sure, some nice vendor presentations. And so maybe they'll just speak to you. And of course, there'll be some demo sessions later. So it's always all about seeing how things work in an individual system. And so although there's some general notions that, at least in sort of earlier generation MRIs, there's always a little more concern about some spherical distortion in some of those images, there's always, in any system, a little bit of uncertainty about how you're fusing different modalities of images. And certainly some operational issues. And these were touched on a little bit earlier. And I believe there was something mentioned of some of the QA activities that are recommended within this particular white paper on QA and safety and SRS and SBRT. There are a number of things you're supposed to do. Tim Solberg, who's the lead author on this, is particularly big on end-to-end testing. And so by that, he means to say that you've got all these components. I think it was David's talk this morning. I think he had a nice little sort of a chain link connection between all the different components of a step, some of which involve imaging and imaging artifacts and fusion of imaging. And so the point that's suggested here in this particular white paper is that that's a part of the potential errors and systematic errors you might get. And so to do a lot of testing in QA on that, in particular the beginning-to-end testing and putting a known target in a known location and seeing how that, quote unquote, moves as you go through the process of imaging it and pretending to target it, is a good thing. There is, of course, the thing that is still done on a daily basis for QA of the machines, the Winston-Lutz test. And I have a couple of pictures from that early paper that these folks put out. This is a way of testing the accuracy of your, in a Linux-based system anyway, the way of testing how confidently you are hitting your isotensor when you are aiming at it. And I just had to find, actually, my excuse for doing that is that this is actually Winston. Ken Winston is actually still alive and well. I don't know if anybody knows him. He's actually, this is unfortunate. This is kind of a dour picture of him. He's got a really good sense of humor. And he's semi-retired. He still, he works at both the Children's Hospital a little bit and at something called Denver Health, which is this sort of public hospital across town. That's where he still practices, sort of on a part-time basis. He comes to our CNS tumor boards and has a lot of insight, because he's been there, done that kind of guy, and radiosurgery. And so I do like him, and I respect his opinion about that, because he was there from a long time ago and has been a great contributor. OK, the only other final topic I'm going to say, because, again, we're talking about neuroimaging. And I just figure, what's the jazziest imaging you ever see? Well, I always find tractogory to give you just the sexiest images of all. I don't know. I mean, we don't actually use it for that. But some people have. And I'll just show you a picture of that right there. As many of you know that, like, look at this one. It sort of looks like a sea creature, right, from deep sea. This is maybe some kind of, I don't know, variant octopus, or I don't know, who knows. But as many of you know, this is a form, a way of looking at long tracts in the CNS, from the spinal cord to CNS. It involves diffusion tensor imaging, which, and I'm not a qualified neurorheologist to explain this, but the basic principle, in terms of MRI, is that diffusion of particles will generally go randomly if there's no barrier. But little tiny tracts, little things like the corticospinal tract, will actually impose a barrier to diffusion. And so there are ways of detecting the linear and curvilinear patterns of tracts in that particular way. I will add, however, that although we don't do this, there are folks who apply this sort of imaging to do some radiosurgery. There is particularly a group in Japan who've actually reported a comparison cohort, not randomized, but a before and after cohort, if you will, before using this and after using this form of imaging to complement their radiosurgery for ABMs and claim to have a lower observed incidence of motor neuron toxicity. They suggest a maximum point to the corticospinal tract if you are going to use this sort of imaging and look for it next to your ABM of 20 gray and feel that they have done a better job in treating lesions and avoiding toxicity with incorporation of that. So with that, I'll stop right there and try to not get us too far off schedule, but happy to take any questions. Thank you.

neuroimaging

AD surgery

randomized studies

brain metastasis

AVMs

pituitary tumors

tractography