 Welcome everybody. I'd like to thank and congratulate you, Kobo, for organizing this day. And I think it's important because, as cardiac anesthetists, we often don't have a chance to A, interact with each other, and B, get some expertise and some experts together who can help you better understand both the workflow, the nobology, and how to improve your practice. And really, that's what today is all about. It should be relatively stress-less for you and more stressful for us. So, I was tasked with talking about the mitral valve, and what we're going to do over the next 20 minutes or so is look at the 3D mitral valve data sets. We aren't going to spend a lot of time on acquisition because I think acquisition is basically a hands-on exercise that you will have the opportunity to indulge in in the next little while. We are going to talk briefly about optimization and manipulation. But the reality is, I'd like to sort of take you through some of the options in what you can use your 3D data sets for and, in particular, perhaps some of the measurements. The only disclosure I have is that I do receive book royalties, and as you can see, Wayne Gretzky is not going to get me very far. In terms of 3D mitral valve analysis, we're going to look at different ways to manipulate your 3D data set. We're going to look at things like angled views, 3D live views, layout views. A bit about cropping, multi-planar reformatting, ice slice, and mitral valve models. So what's it good for? What do people do with their 3D data sets? For some of my colleagues, it's just the ability to obtain a 3D data set, and that's sort of the pinnacle of the day. For others, it's what can we use it for, and we can use it for a few different things. We can look at the 3D data sets in terms of understanding mechanisms, so why the patient has a particular pathology. It's designed specifically to help diagnose various pathologies, and those includes prolapse, clefs, the location of mitral regurgitation, and the severity of mitral regurgitation. A bit about quantification, so we can use it to obtain more accurate linear measurements. Mitral valve area, for instance, in mitral stenosis, mitral regurgitation being a contracted area, and models. And what do models give us? Models give us numbers, numbers, numbers, and are those numbers relevant? Well, it really depends on your practice. And importantly, it helps guide procedural, provides procedural guidance as well. There's several good papers out there that look at the benefit of 3D, and there's no question that it does provide some benefit. There's literature to support that, mostly in the location of pathology and the diagnosis of pathology. So what do we need to manipulate with our 3D data set? Well, there's two groups that you can divide them into. One is using no-special software, and the other is using specialized software. Amongst the no-special software group, you can look at your 3D data set in terms of how you acquire your 3D data set, and how you optimize your 3D data set. The on-fast view, which is sort of the described view to display your 3D data set, you can look at something called angle views and cropping. With specialized software, you can look at multi-planar reformatting or reconstruction, ice slice, and mitral valve models. And we're going to spend some time looking at all of these. There are several papers out there that also look at ways that you can use your 3D data set. And this is one older paper from 2009 that looks at a framework for describing how to manipulate your mitral valve models. So, as we've already heard, the acquisition of 3D data sets is really a trade-off between spatial, temporal, and temporal resolution and size. So these are three mitral valve data sets, and I think by now most people here will recognize them as being 3D data sets of the mitral valve. They're required from the same patient. How do you think they differ? Or do they differ? Or did I just clone all the data sets? Anybody? Any thoughts? Okay, so no one likes this one. How about the other two? Any thoughts about the other two? So if you have the choice, would you take data set one, two, or three? Two, okay. So what's the difference? Well, the difference is this. This was a live acquisition of a complete valve, okay? And I did this with just the 3D live button, and it gave me a 10 hertz frame rate. This is a zoom, okay? 13 hertz, and the last one is a full volume, okay? So again, you can have a very different display depending on what your acquisition was, okay? If you want to see the actual data set, I sort of cropped away a few things here, but these are the real numbers that you will see. So what can you change on your data set? So we've already heard a little bit about gain, brightness, smoothing, compression, vision, and chroma. And these are the things and the buttons that you need to learn about on your machine. Buttons are made to be pushed, okay? Seriously. When you get to your machine, understand your numbology, alright? And use it in a stressless setting. Take a 3D data set of the mitral valve and someone who has a normal mitral valve and play with that data set. See what you can do with that data set. Because no one's going to ask you, where's the prolapse? The patient doesn't have prolapse, okay? And that's how you learn. All of us learned, the experts in the room at least, by exploration and it's a voyage of discovery. And it really is a journey of discovery. You have to push all the buttons and say, that's what that does, that's what that does, that's what that does. The clinical experts in the room are important today because they will help you with that voyage of discovery and maybe eliminate a few steps along the way, okay? But essentially you can have different changes in gain, which is the top. You can have brightness, which is the center, or smoothing, alright? And all of those will display things slightly differently. We've talked about vision, so this is the H vision. So brown is closer to you, white is mid, and blue is designed to be farther from you. And it gives you a little bit of depth perception. And you can eliminate all that by just using a different chroma map. And this happens to be chroma green, okay? So there's different maps and different display versions that you can use. And you have to sort of decide in your own practice what's going to be best for you. What about display? Well, Wendy was part of a paper that actually, with Roberta Lange, that actually told us this is how we should be displaying the mitral valve, okay? And we display what we call an on-fast view. And what's remarkable about this picture is that it doesn't just include the mitral valve, but it includes other structures. And those other structures are important for helping orientate a 3D volume in space, alright? So you've got the left atrial appenditure, the aortic valve here. This is a Phillips version, this is a GE version, okay? Different patients. But bottom line is you get the anterior leaflet, the posterior leaflet segmented. And this is the seductive nature of 3D echo. This is why we do 3D echo because we don't have much valves. And contrary to Dr. Ender, I do believe it is about making pretty pictures, okay? You could also reverse this, and we've heard a little bit about this. So this is the ability to display the mitral valve both in the on-fast view here and from the left ventricle. And this is a single button technology on a lot of the machines now. And it's important though to recognize that when you start rotating things, you have to know which place you're rotating. Are you doing a horizontal rotation? Are you doing a vertical rotation? Are you free rotating? So understanding how that volume is moving by the trackball is important. And this actually is a horizontal rotation, and you can see P1, P2, P3. This is P1, P2, P3 now, okay? So we've rotated around the horizontal axis and not necessarily a vertical axis. So what about angled views? And this is a paper from one of the cardiology fellows that we were involved with here at Toronto General. And what he basically said is, you know, we have the on-fast view, but what else can we do with that on-fast view? Well, if you rotate it around the Z axis, okay, which is like the face of the clock, and we give fancy names to all of them, we can put a paper out that people actually will read and look at different pathologies, okay? So what's the pathology in this picture? So the red arrow gives you a hint, okay? It's pointing directly at a segment there, and that segment happens to be a P2. And if I play this, you'll see this P2 appear much more dramatically, right? So suddenly you've got P2, and you can see this is that famous cobra heads that they described for the P2 in a 2D image, right? So this looks like a cobra head. And this is looking at it. What do you notice about the cobra head? What's that little flippy thing? It's a cord, right? So this is flail, right? So this is a flail P2. And just by Z rotating around the axis, you can actually display the pathology and understand the pathology a little bit more precisely. So the other thing you can do is just look at 3D live views. I don't want to press any buttons, but 3D live on my machine, fine? Okay, you can get away with that. So this is the full volume, not a full volume, but this is the entire mental valve in 3D live. And what you can do is you can use your elevational width, as happens on the epic machine, and G will have a similar product where you can just scroll back and scroll forward. So now I've scrolled back, so I'm at A3P3. And this is this portion here, A3P3. Now I'm going to scroll forward, A2P2, and finally, the most forward, A1P1. And all I've done is just slice through the mental valve by using 3D live. And if I'm looking for prolapse, I can actually tilt this so that I'm looking at it not from the left atrial point of view, but through true sagittal section. So very simply, you can look at a mental valve without using any specialized software. So we talked a little bit about the layout views. And what a voice. I'm a ventriloquist now. Okay, so this is Max's discussion about the epics. Max has been writing with the entire furniture now that he's seen one. Yeah. We're now cutting through the two scallops. And now we're going to continue cutting through here. Oh, because that's the video you've done with all the furniture. Okay, so again, you can use no specialized software and still manage to scroll through the mental valve. And if you're fortunate enough, you can hear Max's voice describing everything as you move along here. GE also has similar software. Again, you can look at sort of two orthogonal planes here through the mental valve. And you can also, on the website, be shown how to look through the mental valve by using these planes. So cropping. So once you have a full data set, fortunately for the mental valve, we usually just take the mental valve. But it's important to understand that you do have a full data set that if you press auto crop, you get 50% of that data set, more applicable perhaps to the LV volumes than the art to the mental volumes. You can use box crop. And if you haven't done this before, it's important to learn the concept on a machine that it is a cube and you can move the faces of the cube independently and trim out the excess that you don't want. And if you're unhappy with the cube being aligned with your data set and you want a free crop, you can use something called plane crop. And that just removes a different portion of the data set. You can use iCrop, which is an adjustable box that happens to be on the Phillips machine. And again, you can adjust its boxes, rotate the box and basically display it in a 3D rendering of whatever volume you want. And again, you should play with these tools so that you can understand how to use them when they become important in describing pathology. The same thing exists for GE, as we mentioned. Their descriptions are view crop, parallel crop, two-click crop, and crop tools. And if you use GE machines again, you should be very familiar with how these work. So what about specialized software? There's a few specialized, a bit of specialized software. You can either use 3DQ or Mitchell Valve on the Phillips machines. And what you do is get deconstruction of your 3D volume into multi-plane or reformatting or reconstruction. So here, it's an end systolic frame that we're using. And we're trying to, again, scroll through that valve to look for leaflet motion in relationship to the annulus. And again, you'll be shown how to do that if you so choose by one of the experts in the room. The blue plane represents a short axis through the valve. The red plane is often the commissure of view and the green plane represents the sagittal cut through the Mitchell Valve. So again, you're looking for pathology here and these tools help you understand pathology and the location of pathology. You can use eye slice, and this is eye slice. So much like you can eye slice through the left ventricle and show short axis of the left ventricle, you can also set the machine up to eye slice through the Mitchell Valve. And with this, you get cuts through the Mitchell Valve, displayed all on one screen now, okay? And you can look at the relative pathologies here. Anybody see any pathology? A lot of pathology, right? Everything's pathological in this patient, okay? But you can do that on a single screen now. You can also do it looking just to see where your location is on your 3D dataset there. You can use specialized software with 3D models. And this is an advanced skill, I would have to say. It's sometimes not easy, but you can develop dynamics or static models and use them for quantification. So how many people use 3D Mitchell Valve modeling in their practice? Uh-huh. Okay. So, Ahmed, you didn't raise your hand. Yeah, there's a few experts in the room that do use it, okay? So why bother with this? And I'm hoping that you will have the time and spend the time to learn a bit about whether it's Mitchell Valve N, which is the Philips product, or 4D Mitchell Valve Q, it may be renamed to something else now, which is the GE product, because it does give us important information. So what it has helped us learn is a bit of a Mitchell Angular sort of movements and how the Mitchell Valve actually works. This was all done originally by, you know, micro-photochromometry, you know, putting crystals on pig carts and looking at how the Mitchell Valve moved. But now we can see it quite nicely. This is the dynamic model from the Siemens product, static models, parametric models from the Philips product. But what we're able to show disputably is that the Mitchell Valve is saddle-shaped, that it does have high and low points, that there is a concept of angular height, and that the Mitchell Valve is specifically designed to reduce leaflet stress. Okay. And that's the whole design of the Mitchell Valve. So if we were to break down the models, there's actually four motions to the Mitchell Analyst. We don't see that in QD. We don't see that 3D on FAS. But if you look at a model, there's actually this up-and-down motion, which if you focus here, you can see the valve moving up and down. Sure, it's moving with the ventricle and not independently like that, but it does move up and down. It also has a circumferential motion. It moves in. Okay. And guess what? That's how the ventricle moves, and the ventricle is attached to the ventricle, the left ventricle, so it's going to move counterclockwise. And it has this concept called folding. And folding, again, is important because it reduces the annular stress and the leaflet stress. Okay. And this is all apparent when you start looking at models. So the bottom line is the Mitchell Valve is specifically designed to be wide and short during diastole, because guess what? It wants to be open and it wants to fill the ventricle and tall and narrow because it wants to be smaller during systole. So there's better leaflet acquisition and there's less regurgitation. So what can we use NPR for? Well, we can look at annular and area measurements. So this is Mitchell stenosis, seen from the left atrium, seen from the left ventricle. You can take this to NPR software and you can actually look through the narrows portion of the valve and replace the valve area. Similarly, you can do this with 2DQ. You can take it to software and you can look at specifically annular measurements. Is it wrong to measure the valve area in 3D? The answer is you can do it as an estimate. Bottom line is you have to remember that it's somewhat spatially distorted so that your measurements may not be as accurate as if you took it to NPR software. There's linear measurements that you can make, volume measurements that you can make and annular measurements, and all of these things are regurgitated through valve models. These are not measurements that you can specifically make all the time in 2D, though you can sometimes. So you can look at diameters and these happen to be normalized values. There aren't a lot of normalized values for 3D models out there, but you can look at diameters and leaflet lengths. You can look at things like annular height such as the ordomitrile angle and non-planetaryle angle. You can look at areas and lengths and these are all regurgitated. The question is why bother? Well, you bother because it actually helps you distinguish diagnoses. This is a paper by Lee from 2013 and they looked at pathological mitral valves here. What they were able to show again is that you can determine the difference between say billowing, flail and barlow pretending on ratios of annular height to commissural width and the billowing volumes. And you can see as the valve gets more pathological it actually gets flatter. So again, an understanding, better appreciation of pathology. It helps you predict valve repair success if your surgeons do valve repairs in terms of looking at various quantitative factors like leaflet areas, annular circumference, angles, height and volumes. And this is going to penetrate the literature more in clinical practice more as our surgeons start, as a neurosurgeon start to look at this and say, look, I work at a reference center. This is what I need to know. Same thing with secondary MR. More importantly, it will predict failure. So these are the ones you shouldn't start to repair. And what we're trying to do is create not just the art of mitral valve repair but now the science of mitral valve repair. And your surgeons may start asking for these numbers and they may start using these numbers as opposed to whether things are going to happen or not. And I'm just going to close with 3D color. So you can use it and it's important to use it to look for MR localization. You can take it to 3DQ software or any type of software that has multi-planar reconstruction and actually trace out the EROA. And you can see that EROA is not circular. We calculate a physiological EROA based on a number of different parameters but now you're looking at more of an anatomical one. And it's important to be able to say this is at a 16 hertz frame rate. So the better the frame rate, the more likely you have more images that you can actually use. So in summary, mitral valve 3D data sets are easy to obtain and optimize. You can use different analysis options to look at the data sets. Don't just look at the on-fast view. And if you model, you can get some quantification that can better define pathology and guide treatment. And I'll close by this little vignette here. And I would use this today as building blocks for your practice, okay? Use it to say, today I want to learn about whatever. I already know about the mitral valve, okay? I don't want to learn about the mitral valve. I want to learn about LVE or EORDA, whatever it is. So that you can take something away that's very concrete today. Thank you very much.