 Okay, then we can move on. So partially what we said this morning is like, okay, today we want to talk a little bit more about the pathophysiology, about some of the basic principles. We started with a talk more about the molecular part and how blood is behaving at some parts. Then we said we look at more like what's happening in cell-cell interactions. And although it's not direct flow, already it was said that you can use hydrodynamic models in order to look at this. And so now we want to go one step further and going again to the pathophysiology, but now a little bit closer to the clinic. And the idea is now to look a little bit more like, okay, when you do clinical measurements of hemodynamics, how can we assess things that are important from the point of view of physiology and clinics when looking at flow phenomena? And it's my pleasure to have Javier Bermejo here, who is the chief of non-invasive cardiology in Gregorio Marion Hospital in Madrid. But also it's like besides being a clinical cardiologist, he's one of the only cardiologists that I know that likes to work with engineers and hires engineers and thinks that engineers are useful or might be useful at some point, at least to spend a little bit of money too. So, and he's really working on how can we get more out of clinical data where we try to get hemodynamics in. So it's my pleasure to have you here, so. Thank you, Bart. I'm definitely convinced we need to put engineers in the clinical field, 100% no doubt. And part of the things I'm gonna show you this morning is related to this approach we've always believed in for the last 10 years in Madrid. So I'm gonna show you sort of kind of a perspective of the kind of research we've been doing, trying to as Bart was summarizing with the bottom line is what can we get out of the imaging and signals you can get in the scanners you buy to the manufacturers? What kind of information can you get out of it that is not directly provided by the scanners? What kind of post-processing can aid the clinical fields and what kind of physiological information can be taken to the clinical perspective? So it's very, very clinically oriented. I'm not going to very much of the basic physics or basic physiology, but I really want to show you the way we organize our research from the technical part to the clinical applications in order to answer common clinical problems. Probably you've all of you here have gone through this in the med or pre-med or I don't know biological schools where the cardiac cycle is shown in this pressure volume loop. Everyone understands this, most of you, yes? Make sure I know who I'm talking to. So here you got the mitral valve closing, the heart, this is the left ventricular chamber, the left ventricles starts to contract, it builds up pressure without ejection volume, the aortic valve opens, then flow starts to come out, pressure starts to be built in the aorta, the volume comes smaller, then you get this part of the curve until ejection ends, you get the aortic valve closing, pressure suddenly drops through isovalometric relaxation and then once the pressure fall reaches atrial pressure, the mitral valve opens and the ventricles starts feeling, right? This is what we call a lumped parameter integrative physiology of what's happening with a one single measure of pressure inside the chamber and volume inside the chamber, right? One single, we assume that the, there's no pressure gradient inside the ventricle, which we later on I'll show you is obviously wrong and we assume that there is only one measurement of volume for the full chamber. So that's the way we've learned physiology, we learned to assess ventricle effect function both in systole and diastole. Now, obviously this is very difficult, very, very difficult to assess in clinical practice. It's easy to measure anesthetic and diastolic volume, you can do this using a bunch of techniques, mostly echocardiography and MRI and cardiovascular medicine, but if you want to get, you know, high temporal resolution to get a lot of measurements of what's going on volume, that's not that easy and obviously it's not that easy to place a pressure catheter inside the ventricle, you have to do a left heart catheterization procedure. So, quite a lot of what I'm gonna show you in my talk is trying to understand how we can get this information out of things that we can do non-invasively and we can measure impatience in the clinical field. And most of the information is related to what's going on in echocardioc flows and for the whole summer school and the whole session devoted to flows, I wanted to focus everything I'm gonna show you and for my talk is related to what's going on with flow inside the heart, mostly inside the left and the right ventricles and what are the consequences of that and what is the understanding. Nowadays, and tomorrow you'll have two lectures on how we can measure velocities inside the heart, you can measure this in 1D, 2D, 3D and 3D and time on 4D and you can do this using echo with contrast, without contrast, you can do this with using MRI. Each technique has its advantages, its drawbacks, but I'm not gonna go so much into the details of the techniques themselves, but a little bit more into the physiology and the processing. And the first thing you can see is that flow is very, very complex inside the heart, okay? Nature for the mammalian and the human heart has made the chambers in a very complex geometry the way that flow follows very complex trajectories and it's obviously time variant and things are very unsteady and geometrically complex flow so we have to try to make things as simple as possible. And this was the original paper brought up by a clinical guy in King's College in London and if you look at flow in 2D, the way you can see it here, this is from his original paper, there are two parts of the picture, we can say flow is mostly 1D or kind of a tube. You see when the flow is coming from the left atrium inside the ventricle, you see a column which is quite like a tube and when the blood is being ejected from the LVA picks out of the aorta there, you can see another sort of kind of 1D distribution if you just focus on exactly that particular column of flow. And that's what we tried to bring up and we thought of if we could sort of focus at very specific locations of flow at very specific instance of the cardiac cycle, there could be some relevant information out there that could help us understand clinical problems. From the early 80s, have you ever heard of these Miller high sensitivity pressure catheters? These microtransducers here, these are not regular catheters as you can see in the cath lab in any clinical hospital, the difference here is that the electronic transducer is sitting at the tip of the catheter which provides much more accurate measurements of pressure at a specific location without the issues of damping or low frequency response and so on. So these experiments done in the early 80s before the development of all these fancy imaging modalities, what they showed us is that there were physiologically relevant pressure gradients inside the heart, right? Even inside a single chamber due to the, this unsteadiness of flow, flow is changing through time and related to that inertia as I shared you as you obviously know from basic physics, there are some small but relevant pressure gradients being built throughout the cardiac cycle inside the heart in the left and in the right ventricles and that is the first very simple information we tried to understand across the mitral valve here between the left atrium and the ventricle or between the LVA picks and the outflow tract during injection, see one part of the cardiac cycle, another part, this is the filling phase or even inside the heart between the apex and the mitral inflow that is also inside the ventricle but throughout diastole instead of systole or even in the right heart across the tricuspid valve. So what is the information you can think we can obtain by trying to measure those pressure gradients? Now the first thing is what is the best way to measure them? How can we deal with getting these pressure gradients out of imaging? So we did, this was an approach that came 20 years ago from Climberton Clinic, they gave us these approximations which is pretty straightforward from a simple 1D flow and study Bernoulli equation and doing some very simple kernels, you could apply maths to the image and obtain the pressure distribution. We did a slightly different approach which was doing some spline fitting and again bringing up the couple of spatial and temporal derivatives and now we get an image instead of depicting flow velocity as we obtain from the ultrasound scanner we obtain amount of pressure difference inside that specific location, right? So this is a 1D flow from the LVA picks to the outflow tract and this is the ejection phase, the blood is coming out from the apex towards the aortic valve and this is the ejection phase so that accelerates flow and produces this pressure gradient here from there to there which eventually is reversed at the end of ejection. So the M mode, this is built by one single ultrasound line scanning with Doppler across this flow from the LVA picks to the order. So this color M mode which is in the ultrasound scanners for the last I would say 30, 40 years that gives us a very interesting map that you can do very simple maths. You only need to decode the velocities and do these derivatives and you obtain the pressure gradient fields and you can obviously do some spatial integration there and provide with the absolute millimeters of mercury between two locations in that specific path, right? So the first thing we did is it was quite some years ago say let's validate this and see what is the use. So we did some micromanometer measurements and we also did this conductance catheter to obtain pressure and volume and obtain these very current measurements of pressure, volume and these pressure gradient curves between this location you see this is systole the ejection being, the blood being ejected and here we've, this is the raw velocity as provided by the scanner and this is the pressure gradient obtained by processing and merging both images. Now we had some modeling data and some theoretical papers saying the more pressure you can build between the apex and your aorta the better your systolic function, your pump function is working. So this ability of accelerate flow of building pressure is a good measurement of overall systolic function in the chamber. So the better you can compress your fluid before being ejected, at the time of being ejected the better your systolic function is and getting this sort of reference method of a very robust measurement the holy grail of systolic function something that really reproduces the way of the left ventricle is contracting is something very, very important in cardiovascular medicine because obviously everything is in vivo everything is much more complicated than an experimental lab and everything is affected by drugs affected by preload, by afterload and trying to get something really robust not dependent on loading conditions is something that in cardiovascular diagnosis we've all been looking for for a long time. And this is what we got this is the gold standard these are the pressure volumes you obtain during acute preload modifications and you get this family of curves as the ventricle load is being lowered and you see these slopes here the steeper your slope the better your systolic function is as you can modulate using drugs this is the baseline this is a reduction of contractility this is increasing contractility by drugs and you get the different slopes here which account for different Emacs which is this reference value for contractility and we got very similar responses for this pressure gradient so we did these experimental animals in the surgical room these were pigs and we did quite a number of experiments to demonstrate this something we could bring out of the conventional echo gave us a very robust measurement of systolic function and that correlated with very very nice correlation whatever kind of interventions we did there so this first we brought up the method we did the software we went to the experimental lab we made sure the gradients were measured accurately we made sure we understood what we were measuring and what did it account for and then the next question was what happens if we apply this into clinical practice what is the information what kind of questions can we answer in patients so we needed to move from the experimental lab to the clinical cath lab and we did a small number of patients in which we did these pressure volumes in the cath lab to make sure the measurements were still accurate and we got a quite heterogeneous population there we had patients with chronic end stage liver disease who manage very complex loaded conditions we had patients with coronary artery disease we had patients with some valvular disease and so on and what we saw there is again obviously we measured these gradients but we also used probably you'll hear later on in the following talks this week all these measurements of strain and strain rate and see what were the best ultrasound derived metrics that correlated best with this reference value of reference index of systolic function and this is the nice correlation we obtained for this ejection integral pressure difference we obtained by the color M mode importantly it was pretty much better than what we got with the strain strain rate obviously or ejection fraction and most importantly we demonstrated here that it was very very weakly correlated with preload or afterload and definitely much better than ejection fraction or stress derived measurements stress did pretty well longitudinal strain here was quite afterload dependent and you'll hear a lot of papers in the last 10 years trying to account what is the relationship between longitudinal strain and afterload and the eodic stenosis and so on but probably most of it is related to not so much to systolic function but changes in afterload this is one of the clinical applications we did we were studying end stage liver disease because as I say there's this hypothesis probably some of you have heard about this condition named cirrhotic cardiomyopathy which end stage liver disease causes some sort of multi-organ syndrome related to cardiovascular dysfunction due to the fact that portal hypertension hypertension leads to increased permeability of the gut barrier and translocation of bacteria from the gut towards the blood this induces an inflammatory response which at the end produces central hypervolemia cardiovascular dysfunction and eventually a bride condition and this was this concept of end stage liver disease leading to cardiac dysfunction has been there for the last 10 years and lots of authors have tried to prove the fact that when you have end stage liver disease eventually something is going wrong with your heart so we tried to see what was happening with our particular method and we found was quite exactly the opposite what we found is that this is a control population these are different degrees of end stage liver disease as your liver disease get worse this systolic function of your heart gets better this is another metric of the degree of your liver disease as your liver disease gets worse your systolic function of your heart gets better and why is that? That is much related to the amount of your activation of the sympathetic nervous system so your sympathetic activity and increasing norepropyrin and other hormones in the blood increases your pump function the contractility of your heart and another important issue is that we showed that this was modulated by the fact you were taking better blockers of specific drug that modulates this sensitivity to SNS activation interestingly the authors who coined the term of this cyrotic cardiomyopathy which is smaller, he's a physiologist from Norway he recognized that the fact that we used the right tool made us the fact that we were seeing something that was actually completely changing around the paradigm that we were demonstrating that it doesn't seem at least at these stages that there's nothing wrong in your heart when you have chronic liver disease So I think that's a nice example of showing you the whole picture from bringing some software out of my lab and understanding a clinical question from the clinician is what's going wrong with your heart when you have end stage liver disease and do we have to take a special care for your heart function in those conditions? Remember some of these patients have to go under liver transplant and that's a very stressful surgery and it was believed that something could potentially go wrong with your heart when you were going that kind of situations Now let's move on to see other location and what goes on Look at here, I wanted to show you the fact at the end of the ejection obviously as this flow decelerates your pressure gradient is reversed you get a negative pressure gradient you get higher pressure at your aortic valve than at your ventricle related to the fact that flow is decelerating until your aortic valve is closed, right? So it was pretty straightforward that the intensity of this negative peak there of your pressure gradient probably was related to the rate of relaxation of the ventricle and we did this fancy engineering analysis you've probably heard about which is called wave intensity analysis have you heard about that? We understand the physiological phenomenon in terms of waves being moving out or towards or away from the heart and this is to understand this you have to bring up with this was a previous speaker was talking about this understanding mechanics in terms of waves, right? This is waves being propagated through the fluid from the heart towards the systemic circulation or from the systemic circulation towards the heart and what we saw here is this relaxation behaves as obviously as an expansion wave pulling your flow back to the heart as the heart is becoming relaxed and it's propagating from the heart outwards to the aorta and that relaxation wave was exactly taking place at the same time as the peak of that negative pressure gradient so therefore if time matches probably the intensity also had something to do and the peak of that negative pressure gradient was related to the capability of your heart to relax during the diastole and that we reproduced it in a number of conditions these are a bunch of patients in the cath lab where we can pretty well anticipate what is the degree of relaxation without the need of putting a pressure catheter inside the heart so that's another important physiological question we can answer doing something with a big, bad and ugly ultrasound scanner just doing, assuming your flow is 1D and doing very simple maths on that image other locations, I'll skip a little bit of this it has been published that obviously if you look at here, this is flow this is your left diatrum here, right? this is the mitral valve, this is the ventricle so this is flow coming from the diatrum towards the ventricle this is time, here you have what's going on with pressure if you look at this here you'll see that pressure is still going down while flow is going in you look at here, this time from there to there flow is coming inside the ventricle while pressure is still coming down that's proving the fact that the ventricle is pulling flow towards the diatrum because it's pressure is still coming down despite it's being filling so if you have a passive balloon if you start to get filling your pressure needs to be going up so if pressure is going down at the time you're being filled obviously you're pulling flow and you have a negative pressure you have suction from the ventricle to the diatrum and that again is something we can measure this is the effect and it gives you this corner here where the ventricle is actually acting as a syringe pulling flow from the diatrum and this pressure gradient is related to the functional class, the amount of this near and the outcomes of patients with a number of cardiac disease and as again as we showed it's something in another paper we showed we can measure again using this very very simple 1D technology we did the same from the right heart we can again measure, estimate this for the right ventricle of pressure gradient estimate something as weird as the rate of relaxation of the right ventricle not the left which is something obviously that had never been done non-invasively in the heart before now for the second part of my talk if I want to move a little bit further on now obviously okay if you look at it the way I was showing you this could be 1D this could be 1D, one and one but obviously there are many more things going on there why can't we look at flow in another way and we get the full picture not only filling or ejection and get more information out of that and there are a bunch of techniques you'll hear for quite a lot further on this is tracing by PIV contrast bubbles using ultrasound the problem is as we showed as you can see here you need very very very small interrogation boxes because if not you can't measure with enough spatial resolution and what we showed here is unfortunately for obtaining these pressure maps we need quite a lot of that temporal resolution we proposed a different approach in which what we did was solve continuity equation within a conventional 2D flow image and making use of this detection of the blood myocardium interface to obtain the boundary conditions for solving this continuity equation we did the technology we did some inventory simulations experiments we obtained we did some interleaving of frames to obtain much higher temporal resolution and this was after processing this is the raw data in a pulse duplicator we did some conventional PIV measurements there and we obtained that this technology of reconstructing the 2D velocity map out of conventional ultrasound images seemed to work pretty well and then this allowed us to obtain a full 2D representation of what's going on now what can you expect what's going on inside of the heart what is the conventional flow what's going on there you see this is a very intricate geometry where the ventricle has the aortic valve it's pretty much in the same plane as the mitral valve there so the flow has to go a very very weird trajectory to go from filling to outflow and it builds this there's a very huge change in inflow geometry from the mitral valve tips to the ventricle chamber so there's a sudden expansion for flow which obviously leads to the generation of a vortex ring that eventually advances through the ventricle from the mitral valve towards the chamber gets bigger unpings from the filling jet and eventually is taken towards ejection across the aortic valves because as I say they're pretty much in the same plane the mitral, the inlet and the aortic valve in the outlet so we have to do a flow does this very complex trajectory along the cardiac cycle and what can we do and what kind of information is there well we have this very very basic paper which is very provocative saying well it's not the same to get this kind of filling than this one and that might have its consequences on the physiology of the ventricle using this technology it's easy to understand what a vortex is because this is doing simple mathematics on the flow and then you can characterize exactly what we call a vortex using this Q index which you probably know much better than I do and there you can measure exactly the radius the position, the position and the energy that is there sitting in the vortex while flow comes in and flow comes out from the heart we did some validation how well do these 2D based metrics of vortex size, position, radius and so on correlate what's going on with MRI and we found they were doing pretty well and that although we're assuming in order to apply this continuity equation that flow is mostly 2D we got pretty good metrics and quite accurate measurements of what's going on with the flow and with the vortex and that gave us the possibility of describing for the first time using conventional echo without the need of any contrast agents what are the conditions of vortex you see very well here how the vortex is being built and that in every phase of the cardiac cycle you can obtain these metrics and get very nice representations in different populations this is something that can be obtained with a conventional ultrasound scanner no risk for the patient very cheap technique and you can get the characteristics of flow going inside the heart and we see the size, the circulation of this vortex structure and what is that related to obviously very importantly to the amount of energy that is coming through the inflow the more energy you get the bigger vortex and the more circulation the vortex you'll find and obviously the bigger this disproportion between your mitral valve tips and your ventricular diameter the this size of the flow expansion always brings up a larger and more energetic vortex so what are the consequences of that why has nature gone through this process of building up such a complex geometry for the heart and what is the advantage of getting such a complex flow dynamics inside the left ventricle there must be some reason there we need to understand what's going on this is a modeling paper which was done almost 10 years ago what they simulated different temporal activation patterns in order to screw up completely what was the trajectory of flow and what they found is was in terms of pressure and volume there was not much of a difference now the question is this a lot of people who have discussed this paper they've said well it's not a matter of not being able to build up the pressure volume loop it's a matter of how much pressure of the left atrium or how much energy does it take you to build up that pressure loop it's mostly a matter of efficiency okay if you you can develop the work the heart can build up external work but probably it needs more energy because if it doesn't make use of this mechanism I'll show you some data supporting that this is a study we did exploratory 20 healthy volunteers 20 patients with a dilated heart non-dyschemic dilated cariomopathy a very round and poorly contracted ventricle and another 20 patients who have hypertrophic cardiomyopathy hypertrophic cardiomyopathy is exactly the opposite you get a very small ventricle very thick walls with a very small ventricular chamber systolic function is normal hypernormal in the early phases of the disease but there's nothing wrong with your pump function your systolic function is okay but your ventricle is small your walls are thick so you're getting some problems in the way you're being filled right because your stiffness your myocardial stiffness is much worse on the contrary you know on the contrary dilated ventricles the walls are thinner and it is believed that myocardial stiffness is below normal values because just because the fact that the myocardial walls are thinner than controls what we looked at is what was the role of the fill in vortex in this particular geometries so we measured this flow and then used is anyone here ever heard about something called the vortex panel method used in air spatial engineering well what we do is if you do some maths out of the flow and you have the geometry of the wall you can simulate if you what would be the flow being trajectories inside your chamber if you got this vortex and you put it inside the chamber right so this flow pattern we're obtaining here is the consequence just of the vortex and the rotational flow field is the consequence if we we did this exercise this would be vertical flow this is the flow we measured we subtract this from this this would be the flow you would get in your heart if your didn't have any kind of vorticity in there that would be sort of filling as a tube so what you measure we did this obviously artificial decomposition in what would be your vertical flow and your completely irrational irritational flow and that is what we see you can get this you can build the synthetic flow fields out of there and this is what we get in terms of velocity or probably more important in terms of volume so what are we seeing that this vortex being built inside the ventricle the bigger the vortex the more flow is that it is pulling from the left atrium just because of a phenomenon of inertia so as the filling phase is going through from the opening to closing of the mitral valve the vortex is being built in normal patient this is what it does the more the vortex ring is going round and round it's in training fluid from the left atrium pulling it through across the mitral valve if you have this small ventricle no expansion of the chamber your vortex is much smaller and you don't get this facilitation mechanism of filling just because of the fact of that vortex so in some way this vortex structure going on there it's acting as a crankshaft you get this pulsatil flow because the of the diastole has a number of phases it's not continuously the same you get intermittent flow due to the phases of the cardiac cycle you get this sort of crankshaft mechanism of inertia keeping the flow moving that it keeps it the way it facilitates filling and if you don't have the vortex you're not aided by that mechanism what else so what other kind of information can you get out of flow well you can do these fancy measurements of tracing where flow goes this was done first using these Lagrangian coherent structures you've probably heard about and then we can see very well what's happening with flow and you can do this at different phases of the cardiac cycle and you can get what's going to happen with your blood along the cardiac cycle for a number of phases you could also do this backwards what's going to go what's going to move inside what's going to happen with flow coming inside from the mitral valve you can look it backwards through time what where is the flow that's going to eventually exit the the ventricle this is a nice application we did for looking at different responses to different uh... adjustments of pacemakers and what we've seen sorry is that flow patterns of this what where is the blood sitting at the end of your cardiac cycle and what's going to happen for the following beats is different depending on how you set up these pacemakers fine tuning pacemakers is something that is also a pending issue in cardiovascular medicine you know we have very complex devices now where you can uh... fine tuned AV delay between the actual contraction and the ventricular contraction we also have to tune the interventricular delay what is the time you want one ventricle to follow the other one to start the onset of contraction and looking at these specific flow patterns can probably aid us in which are the best settings in the way to tune these pacemakers i'm going to end up with this very very important issue and we believe there's a lot of potential of looking at the flow in the heart in order to prevent this cardiombolic stroke is one of the major sources of disability in cardiovascular disease about thirty percent of patients with a stroke have there it is believed that uh... the clot in the brain came out from the heart went through the carotid arteries or the vertebral arteries and eventually uh... stopped flow in the brain right so if we are able to do and progress in preventing clot formation inside the heart it is understood that it had a it will have a major impact in disability caused by uh... uh... ischemic stroke in the brain right now the problem is current techniques for predicting and obviously you know you can prevent this by starting anti-coagulant therapy you use drugs to lower the coagulation capability of the blood and we can use these drugs to prevent clot formation inside the heart now the problem is we have to balance the risk benefit of anti-coagulation and then that is something that is not completely solved we know quite well we thought we knew quite well the way to go for atrial fibrillation a very specific cardiac arrhythmia and we have very precise indications of what to do for preventing stroke in atrial fibrillation but in other cardiac conditions which are related to stroke we don't have such a accurate metrics in order to predict who should go and through anti-coagulation therapy for for his life and who not for doing this we've brought up this new metrics and provide maps we use this fancy mathematical methods and again using data from velocity out of a ultrasound scanner or eventually from a 3D MRI scanner now we're not coloring velocity here we're not coloring pressure maps here we're coloring residence time of blood inside the heart so the longer your blood is sitting inside the heart the higher the risk it will activate coagulation and eventually will lead you to thrombosis inside the heart right which is exactly what you don't want and we've started some clinical studies looking at this using ventricular assistance devices where you know thrombosis formation is a major risk also looking at and obviously you don't just get the numbers but you also get the exact distribution of these high stasis regions inside the ventricle and now we can we can estimate what is the degree of contact of the static regions of blood with the myocardial wall and we can get very very accurate numbers this is a population of patients and who had a myocardial infarction and a large myocardial infarction in which some part of the of the heart is not contracting properly and we get these metrics of residence time region size of the stagnant region or the residence time of the stagnant region or the contact with the perimeter and as you can see here patients who develop thrombosis eventually in during or immediately after and myocardial infarction have disturbed and pretty nice nicely different patterns of flow this is an example these are two patients both of them have a abnormal motion in the apex due to myocardial infarction those are the measurements we get out of the scanner these are what we can see and flow going on inside the ventricle as you can see here you get an apical region of stasis there we don't like lots of blood lying there which eventually won't wash out not for the following four seconds which is something you wouldn't like definitely worse than what you're seeing here in this other patient who also has an apical infarct whose ejection fraction is exactly the same as this one but because of the distribution of flow flow is entering in a different way the ventricle this apical static region there is much smaller what happened with this patient he eventually developed an apical thrombosis this one did not so we believe this kind of information we've proved this in larger clinical trials will definitely tell us something about physiology something about the risk of stasis and this could be adequate to try to anticipate and target anticoagulation in this particular patients so to conclude I think this is the kind of things I'm trying to show you this morning recent advances in images now we can obtain very robust measurements of intercardial flows and this can be obtained in the clinical setting and you're going to hear about this probably tomorrow a little bit more you get these diastolic vortices which are very important and are long-standing and they have a physiological role if you don't get that working properly that would be leading to probably less efficiency in terms of filling and in terms of pumping and this is important in terms of mechanics and this could be instrumental for a number of diseases just to summarize I want to show you the kind of things other important authors have said about this clinical research this issue about dilatation, impairing diastolic filling this has been shown just simply by the fact of analyzing flow the fact that it's not only a matter of the mechanics of the material of the myocardial wall what's happening in the chamber also impacts efficiency these issues can be obtained using echocardiography and this is one of another editorials that were pretty quite encouraging that they said to us well in a wider scale emphasizes a shift in our thinking that the heart is just more the pressure-generating mechanism so we have to understand kind of actual physiology in terms of not only pressure but also flow on what's happening inside the heart I also want to acknowledge the people who work with me this is a bunch of MDs, engineers, physicists, vets I don't know, technicians this is a huge amount of work for a very crowded and busy clinical imaging lab this is probably the task for a lot of people from the department probably much beyond what you've seen specifically for the people and we're currently hiring and we are very interested in the kind of profiles where you are probably sitting there so if you considering a potential pre-doc or post-doc position in a clinical lab thinking about doing this kind of work please feel free to contact me and we'll be happy to give it a thought together, thank you thank you very much for this comprehensive overview maybe just a real practical question to start with is one of the things that you mentioned before is that you said like you're doing animal models and then you do millar catheters and this kind of things in general when we as modelers do this kind of work and we want to look at pressures and a lot of people are trying to do pressure volume loops and then we say like okay we need pressure measurements and you go to the cardiologist and the guy from the cat lab says like yeah I mean I take pressures every day and they give you the pressure measurements from the cat lab which is obviously from a fluid filled catheter so what's your experience with doing this kind of modeling research and then using fluid filled regular clinical pressure measurements uh... a tough experience uh... the way if you really if you want to get things done properly and if you want to I mean if you want to get accurate and science out of that these everyday measurements are not good enough well they're good enough if you want some lumped time-averaged values I mean if you want single scalars out of what is your endostatic pressure that's okay what is your wedge pressure that's okay if you want the waveform then it's not okay and probably you should ask for funding and train these guys to use high-fidelity catheters that's the way to go so that's my recommendation it's not only that the engineers don't speak cardiology language but the cardiologists don't speak engineering language neither no and probably nowadays the guys in the cath lab are not the most physiologically oriented anymore unfortunately but you can do that I mean it's they're not there now you have this you know about these uh... pressure guide wires used for the coronaries these are pretty cheap and they're sitting there in every cath lab in the country so you can put that in the ventricle and it'll provide you a very nice pressure waveform just as good as if you had put it in the coronary so that's a it's only rather difficult to get the digital data out at least in our experience maybe I can give you a hand okay partially related to this also is like what some of the things that you are emphasizing is that actually with the clinical data which is the Doppler data that you have you can do pretty much if you do it in a proper way so the question is a little bit like when you see that if you look in the echo lab people take like Doppler traces especially pulse Doppler traces from the valves every day and this is like routine the only thing they measure from it is maybe EA ratio or EE prime or whatever you would measure from it the question is a little bit what is your impression is this underused understudied data or is this old fashion that we need to go to 40 MRI because that's the only thing that can give us information well the way we've looked at it I think there is it's a continuum I mean there isn't some information out of the pulse wave Doppler data because you know regarding whatever no care whatever is going on in between they've already you know outcomes between EA wave and outcomes has been linked so we know from an engineer perspective most of the assumptions in the middle are wrong but the fact that there is some you know empirical relationship between metrics that is okay and tells well give me that information because I know it's going to help me decide what's going to be the outcome now for every what what is the the balance between complexity and clinically relevant information I mean the whole world what I'm trying to explain here the whole world of cirrhotic cardiomyopathy is a paradigm of wrong science out of the wrong technique so you were looking at diastolic function with the EA wave and the DTI data and an ejection fraction and strong volume which if you've ever seen a cirrhotic patient it's a completely rubbish I mean it's it's it's loading conditions are so abnormal that there's no info so if you want to go into there's a terribly important field nowadays where this kind of things that makes your life a little bit worse in the everyday echo lab is very is going to become very very important that is in oncology and you know nowadays most of the drugs being used to treat cancer have some sort of an impact on your ventricular function so you have to monitor your ventricular function very carefully eventually someone in the cardiac side says watch out something is going wrong with your heart then you're telling the oncologist stop the therapy and he's going to neglect therapy potentially able to uh... get rid of the tumor eventually cure that patient and you're gonna make him stopping because you're gonna say something's going wrong in your heart so there you have to make sure you got your metrics right and you got your reproducibility and you got the right indices and that's where we believe this kind of you know more sophisticated issues could be helpful the same for stroke prevention maybe maybe for you know some sort of subtle measurements of diastolic function definitely we have to go into more complex flow analysis that what we're doing and everyday clinical practice and maybe just again from your experience what you see is like you have been working in your lab on the Doppler measurement and trying to extract information out of there trying to do something clinically relevant and get that information on the other hand all of these tools at some point the only way to get them into clinical practice if the companies take them over but the experience seems to be that the companies want like fashionable new methods 3d even more sophisticated instead of going to some of these things that you can get hemodynamic information that's that's that yeah that that is a very very very important issue and i'm not sure we've done things right we've got we we have a step move i mean for when we started we did establish some relationships with some of the companies which actually brought up some patent filing of some of these tools by the company and eventually never put it into the scanner so that was a pretty nightmare and why did that happen i can't give you the answer as you know better than i do most of the companies don't need to demonstrate any clinical impact of putting a new a new bell and whistle in a scanner okay they just want to the all that they are asked by the fda or by the isi mark or whoever is that if they say three it's close to three and it does no harm if it's a diagnostic tool okay so nice new bells and whistles for diagnostic modalities is pretty easy to put it inside the scanners just make sure it makes no harm make sure that if you say three it's not thirty it's not three hundred it's something put something between zero and fifteen that's good enough and then you're allowed to put it in your scanner you can make yourself your machine better than other vendors now why is it more difficult to get this kind of information out well it's it's very interesting it once it's published then if the technique is published then one vendor won't have things so easy to put it being protected so other vendors don't have it so it's not as fancy as trying to make sure they only have this at least for the uh... following years they have some sort of advantage over the competitors so there are a number of reasons we are we have undergone some patent protection of some of the things i've shown you before and eventually i think there's a new market coming up that are companies who are not so much that it's sort of a smaller companies at least for the early phases they're going to bring up they're going to develop this tools they're going to hire engineers probably in the long term they're going to be you know bought by the large manufacturers but in the meantime they're going to provide some sort of and they're starting to to come up some sort of uh... processing algorithms and something that is uh... straightforward for the clinician you get the data out this is happening for the i don't know ffr measurements out of the ct there are small companies being built up where you get them data out of the modalities out of the scanners through the cloud you get a server locally or whatever they provide you the measurements and that is where the value is and i think there is a potential market there for these uh... post-processing based uh... technologies any questions everyone's didn't understand anything or is in shock or is simply hungry thank you for representation was very nice uh... a couple of questions the first uh... you partially answered before and it was about the balance between uh... uh... very complex technique and the assumption that we have to accept for using very easy technique and and also in what i think it's still uh... not mentioned was uh... the operator dependencies of uh... echo which is uh... quite a big problem and the second question was about uh... the resident times which is very interesting and uh... it was related about the fact that uh... maybe the the flow moves in perpendicular to the the field of view you have and so maybe i will ask uh... uh... answer about the possibilities for the flow to exactly yes this technique good thank you very very very good questions uh... we every time we get a new index out of these we do all the reproducibility issues and not only for the processing part which is usually easy we also go through the acquisition part right so we get the same patient one of the uh... sonographers come out comes out of the room another one moves in and starts the procedure from the very beginning again and so we get true retest reproducibility the more complex your image acquisition processes the more the lower your presuit ducability will be that is and it's also i mean a hundred percent but reproducibility is not true not even for either i would say uh... mri volumes which is supposed to be the cost and i can give you figures and you can find papers out of fifteen twenty percent uh... variability for managing insist on it or in the stock volumes out of mri so careful more complex technology is not necessarily equivalent to better reproducibility and regarding this uh... flat flow assumption for measuring residence time you're absolutely right and uh... our next uh... uh... approach and we have already done the software is ready is obtaining residence maps out of three d mri data yes definitely that's very important the problem is doing some in vitro validation for residence time uh... is not as easy because eventually we could be we've been thinking about different models maybe out using different dyes in in vitro simulators to see if we're able to to get some robust you know and to end validation when we say four seconds it is actually four seconds but uh... we're going through that we'll eventually probably make it but in the meanwhile yes i do definitely think there is going to be a mayor improvement if we're able to obtain uh... residence time maps out of flow measurements in three d out of face contrast yes other questions uh... in the technique where you calculate the vortices using the continuity equation do you have any kind of temporal uh... transfer of one image to another one or do you keep each page separated no uh... we do some interleaving we get a number of beats to build up and we do some interleaving to to build up the velocity so the velocity map does have have some temporal continuity and some temporal smoothing and some temporal filtering but the vortice detection is being done frame by frame so we don't uh... that that is based on the velocity field we don't do anything special or any uh... spline fitting or any smoothing for the vortex futures do you think the temporal resolution of ultrason plays a role in the results that you have meaning like miss some information uh... well uh... temporal resolution of ultrason is the best we can get and we've incorporated this into leaving which makes you even better uh... it depends what you want to measure i mean if this vortice thing the vortice uh... metrics are rather smooth because you know it takes almost the whole cardiac cycle whatever's happening isa volumic faces is much faster so their temporal resolution depends what you want to measure pressure gradients if you do that our first approach based on temporal and spatial derivation there you need very high temporal resolutions but as you probably know there are other methods which are not so dependent on temporal resolution so it depends thank you for your presentation oh sorry and uh... uh... regarding what we have just said about the MRI uh... according to your opinion uh... do you think that we need more uh... comparison study among uh... ultrasound and uh... MRI it depends what you're trying to measure for for some things you do for others you don't i'm not sure what the gold standard is for depending what you're counting for flow the problem for 4D MRI obviously is is sensitivity you make sure you get the right bank for exactly what is your window flow velocity you want to measure if you're and that's not it may be not that easy to adjust and obviously your uh... the resolution of your 4D dataset is very very very closely related to acquisition time so if you want quite a lot of temporal resolution spatial resolution you're going to go through hours of imaging acquisition time or some of the fancy new algorithms where you can do it a little bit shorter but it's for these 2D ejection fraction that's every every clinician i mean if you go to the hospital tomorrow uh... everything is based on 2D ejection fraction we know it's rubbish we know we have from 3D echo biplane echo MRI CT uh... nuclear medicine where you get more robust measurements of ejection fraction the clinicians said will answer you so what you know clinical validation papers were done with n-mode derived ejection fraction and we're still guiding a lot of decision making with techniques what are not which we know do not meet a number of assumptions but a lot of things they're good enough so it's a trade-off how much validation you need where to obtain it how much you need to make sure you match the MRI results i don't know it depends and the second question about uh... your work uh... how do you think it could be possible to improve communication among physician and engineers i can give you the answer to that i mean if you want you can have a beer with me later on the main answer is you want the engineers sitting in the hospital you want the engineers sitting as close to the patient as you can that is my particular view uh... you want the patient we want the engineer to understand the way we work the way we think the way the kind of questions we need to and you know sort of to sense the impact it has on uh... everyday clinical practice so my personal view is you have to move by your engineers closer to the clinical field and interact together definitely yes that's my personal answer to that question and that's my personal experience plus that it takes ten years in order to talk it takes a while it takes a while maybe just also still has a comment on what you're asking with regard to the MRI i mean when you look at the modeling engineers the ninety five percent of them use MRI as starting images but if you look in cardiology ninety five percent of the images is echo and less than five percent is MRI so i think using the echoes to get information out we really have to start thinking as an engineer in that direction i think that's quite important well going back to your previous question another thing that is important is you have to move uh... medicine out of exactly mean the problem with medical school it's very very close we're having we're doing the same training for the last fifty years uh... you know the kind of question i get from my fellows on my residence at the hospital they said well i'm a cardiologist you know and i always have to reply yeah you know you come from a scientific field okay this is not an art this is not about literature you know maths you know physics you went that through school so we also have to do a lot of work on our side and i think you guys in the engineer work are probably uh... getting closer to open up your mind and understanding that your the kind of materials you're studying in university has to be mostly oriented to different tools you're eventually going to need to make your way out as soon as you get out of school in the medical path you just want to you know get the knowledge specifically of what you want to treat your patient tomorrow and you're not caring if probably that's got nothing to do what the kind of skills you're going to need twenty years on which is probably more related to i don't know dealing with computers being with physics physiology and not just you know learning the basics of the clinical guidelines which is something that eventually will come out of the computer thank you very much for coming thank you for the clinical insight i just want to ask you five more minutes uh... before having lunch uh... because one of the important things is the way we can the only way we can organize this type of events is also if we have some help and if we collaborate with some organizations and Jerome is going to tell you something quickly about this so uh... thank you both well uh... thank you all uh... first for being here at this summer school uh... so you've seen the price of summer school is extremely low because the objective cannot make it for free but uh... the objective is to make it so as cheap as possible so to have a maximum of opportunity to uh... to attract students here and uh... we have very prominent speakers that are coming from different areas of Spain and Europe and uh... so all those people are coming uh... for free for we pay their travel and they stay here but it's really uh... there because they're willing to share uh... their extraordinary knowledge uh... with us here and and we have to be thankful for that and uh... here's uh... scientific societies play a huge role and and this is what i would like to introduce you uh... shortly so there are two main uh... societies that's uh... underlining uh... underlining this uh... this event so this uh... first is the ph institute uh... who is co-primitor of the event and then the european society of biomechanics and then and the local part of it for uh... so sponsoring the also sponsoring the event then we have other sponsors uh... so cardio function so i don't know if but later on we'll speak about this it's Marie Curie training network and uh... and then so we have uh... simula and foundation qs who are uh... who are funding uh... awards so having a winning an award is always good for the cv so so it's it's very nice to have that too so uh... i will start with european society of biomechanics i don't know if someone you know this site it's very well established has been funding in nineteen hundred seventy six it has started with twenty scientists so today it has uh... over uh... one thousand uh... members and then the goal is to encourage research to see me in knowledge and progress in biomechanics so large part of you've seen and as you will see large part of uh... computational uh... research in biomedical engineering and it involves biomechanical uh... component so so having a representation of society so makes makes uh... fully sense and uh... so this society does indeed a lot of things so uh... most important events in annual conference so this year it will be in savi in uh... by the beginning of july and uh... so most of the money of society is is used uh... to fund awards so and most of the awards are indeed awards for students awards for junior researchers so uh... we have beach best phd thesis awards uh... which is uh... very competitive uh... our art and very well-recognized across uh... europe and we have the student awards which are what's for the best abstract presentation at the at the congress so we have travel awards so too uh... to fund the travel of uh... of people who deserve it and to the to the conference and uh... we also have a cheer mobility awards so mobility awards uh... so funds phd students to spend three months three to four months of their phd time in another european uh... laboratory in order to uh... format collaboration and if it's interdisciplinary so even better so it fully follows and the philosophy of virtual physiological uh... human and it does so give access to junior scientists to very prestigious awards so this uh... stephan awards stephen peron awards so sometimes it's a very senior scientist who wins it so uh... nearby uh... retirement and sometimes it's uh... phd students so i think that's that's a so very nice and to have this this kind of uh... things and then we have student workshops mentoring awards job first so in the society everyone pays a fee uh... fees very low i think for students it's uh... twenty euros a year and uh... and then this is this money is very important because this money allows uh... so the promotion of organizations like the summer school uh... allows the organization of uh... awards if you apply to competitive fellowship in the future so that's very important then uh... to do to have an award or couple of awards in your in your city if you don't have it's not dramatic but if you have much better so the more opportunities you have then of course uh... the better and virtual physiological human uh... institute so it's currently following uh... a very similar path uh... in order to uh... support research so broader than biomechanics really what we're covering here and uh... it has started uh... it has started as an initiative so it's back nineteen uh... ninety seven and has followed several steps as you can see here so to promote the use of computational tools in in medicine so for personalized care solutions reduced need for experiments on animals more holistic approach to medicine and preventive approach to treatment of disease so then the objective is is to take so what is uh... available in basic science basic numerical uh... tools and seri coupling it to biology and then for the medicine and their uh... really foster uh... really foster so an improved uh... an improved health care systems of both those patients and and for the society so here so there's a huge effort of uh... a lot of persons uh... that are uh... giving uh... time and personal resources to the society there is some membership uh... that's for students ten years per year so it's it's really symbolic all those of you who have registered you have your membership uh... incorporated in the uh... in the fees so uh... so you're contributing to uh... nice project and as for the activities that there isn't the annual conference where you still have uh... student mentorships uh... artworks etc uh... then we have those annual summer schools that uh... right now we are uh... doing here in Barcelona and we still have uh... two mister webinars so uh... those webinars uh... so if you remember of the VPH Institute you're receiving all the news and uh... those webinars so aim uh... to uh... bring to everyone uh... researchers who are normally not really available you go to conference there are these researchers you can not get in touch with them because they all have very important things to do meetings etc and they don't answer the emails if you email them in general so and so here through this webinar so we want to create special opportunities and uh... for example you have the next webinar this is a clinician who is using uh... computational tools i will have a webinar so special webinars so uh... in life uh... from the VPH summer school on uh... on Wednesday afternoon so given by an officer of the food and drug administration and uh... and uh... so what makes a difference uh... between uh... common scientific societies and the VPH Institute that the VPH Institute is really committed so to uh... to foster our activity and our opportunities of uh... of funding and uh... of finding uh... jobs later so through the exploitation of what we are doing and of course this can only be possible if we pass through uh... policy uh... so policy exercises so they're working very closely to the commission that's a lot of years of of efforts and now uh... basically the European Commission is recognizing this importance and we can see it so through the H-2020 uh... programs and other programs uh... which generates a lot of opportunities then for us okay thank you very much so i think it's important to have a look to these websites you can get a lot of information there especially for young students to say awards and things that are important