 The next speaker is José Rodríguez, he's a cardiologist, a clinical cardiologist by the Brown Hospital here in Barcelona. He's doing non-invasive imaging and for the last years he has been involved more and more in MRI, mainly in cardiology, going from using MRI for looking at the heart, mechanics, looking at deformation and things like that. And nowadays, more and more about flow imaging. Well, first of all, I will start my presentation. I would like to thank Bart for the invitation to participate in this summer course. And also I would like to thank the previous speaker for being so supportive and reply all the questions while I was struggling with my presentation there. The problem was that I had it on Dropbox and it was very heavy to download it. So as Bart said, I'm a clinical cardiologist. I work at the Brown Hospital. So I'm going to do a presentation that it's going to be a little bit different from the previous one because I'm going to explain physics in MRI but also why MRI and for the flow MRI is important for us in clinical cardiology. So, three more seconds. So the first question that we should answer is why flow imaging is important in cardiology? So why, most of you will be engineers. So why we need your support? Why we need flow in cardiology? So when we are trying to evaluate the severity of a valvular regurgitation, for example, a valvular stenosis in a patient that we have in our outpatient department, so most of the criteria that we have for the severity of disquantification of the regurgitation or stenosis are based on clinical flow. So, for example, we consider that a severe aortic regurgitation is when there is a regurgitant volume greater or higher than 60 milliliters per bit or the orifice of regurgitation is higher than 0.3 centimeters squared. So we need volumetric information in our cardiology department every day. So also once we have that information, we have the volumes of the regurgitation or the severity of the stenosis. We take these algorithms in clinical practice just to consider. So if we have a severe aortic regurgitation and the patient has symptoms, so we consider that the patient is ready for aortic valve replacement. So based on the information that we have from the echo, based on the information that we have from the MR and based on the patient that we are interviewing, so we decide if we have to operate or not to our patient. So, but all the volumes that we used to use were based on ecocardiography that has been the technique of choice that we have been using for a long time in the past and we are still using a lot in clinical practice. But the problem of ecocardiography is that we are assuming some anatomic patterns that they are not real. For example, in order to estimate, for example, the volume of the aortic regurgitation, what we do is we have to calculate the area of the outflow tract in the left ventricle. So we consider that the area of the outflow tract is totally circular but we know where we are using multimodality imaging that the area of the outflow tract is not circular, that it's more elliptical. So we are having some errors in the estimation of the volumes that maybe the information that we get from echo is not as accurate as we wanted to. So that's the reason why MR emerged as an important tool in order to solve these problems and to give us an accurate information about volumes and severity of regurgitation and also about velocities and flow. So what we have here is an MR sequence. So it's what we call phase contrast sequence. So this is an aortic at the aortic level. So this black image is the aortic flow. What we have here is just tracing the area around the flow as we can see here and just propagating all this ROI in all the different phases that we have in the cardiac cycle. We have a table like this where we have all the information. So we have all the total forward volume. We have the regurgitum fraction. So it's like 15% of the total volume that the stroke volume. So 15% of this volume is coming back to the left ventricle. So we are having lots of information just doing this trace. So just analyzing the information that we have for MRI. So we don't have to assume anatomic areas as we did in echo and all the calculations that we did with echo are more easy and we don't have to calculate all of them just using MR. So that's the reason that MR started to be the technique of choice in our clinical practice to evaluate volumes and fractions of regurgitation like for example as we have here. So it's much more reproducible and easier just to manage our patients. So what we do in MR is we use one sequence that is called like this is the phase-contrast sequence that what the sequence does is that it analyzes how the protons that are within the vessel change the phases and this phase is proportional to the velocity that the protons have. So for example we apply like two different gradients. We know the time that those gradients are applied. So we calculate how the proton has moved and this movement is how defaced and this defacing is proportional to the velocity that this proton has. So we can estimate using MR we can estimate the velocity of the protons that are just passing through one slice position as you can see here. So we can know the velocity and based on the velocity we can calculate the flow. So for example here what we used to do with MR was we analyzed one slice position so we selected we wanted to estimate for example the flow here so we decided to do a slice at this position. So we have an image like this. The aorta is here so this is the aorta. This is the ascending aorta, the aortic arch and the descending aorta here. So we have the aorta here and we can have all the velocity, the regurgitant flow in this slice here. But we can only analyze the flow that is going in two directions. So it's the through plane flow that is going forward direction or backward direction as you can see here. So based on this we can calculate the velocity of the protons that are across this aorta, this aorta, the lumen, but only just in one direction. Just the direction forward and backward. So this is an image that we have at the aortic valve. This is the aortic valve of the patient. This is the face contrast image. That's the image that we have. As you can see the flow is seen in black here and this is the anatomic image that we get also from the MR. And what we do is as I said before we have to select one plane or one area when we want to analyze the flow so we can do a slice here or a slice there or at the aortic valve so we have to decide when we want to know how's the flow there. So we select the plane and we do like an image like this and we can estimate as you can see here the flow, the velocity, maximum velocity, gradients, being gradient, maximum gradient, etc. But what we used to do is we select one slice position so the patient did an apnea and we get this information from this specific slice position. One of the characteristics of the face contrast sequences is that we have to previously select the maximum velocity that we expect that the patient is going to have at this specific point. For example, this is a patient with an aortic stenosis that means that the aortic valve is reduced, the orifice of the aortic valve is reduced and as you can see here we selected a velocity that is below the maximum velocity that the patient is having at the aortic valve. So as you can see here we see that there is a black flow, the flow is black but in the middle there is like a white flow that means that the velocity that we selected is not the correct one so we have to increase the velocity, we have to choose another velocity encoding just to make sure that we are depicting the best for the appropriate velocity that the patient has here. So one question that why don't we start using a big velocity encoding so that we don't have to change the velocity, the velocity that we are trying to imagine, we are not depicting the correct one so we have to do one slice position with for example 100 meters per second, 100 meters per second if the maximum velocity is higher so we have to choose another one so we have to increase it slowly until we get an image like this with no aliasing in the middle. The problem is that if we start using a high velocity what we get here is that we have a lot of noise in our flow so we don't have an accurate measurement of the maximum velocity and we don't have a great or an exact determination of the minimum velocity either so the thing is that the higher the velocity that we are using the more noise that we are having in our images. So as I said what we used to do was one apnea for our patient we had to select before when we were scanning the patient where we wanted to know the arctic flow or the velocity at one specific point and then we get one image, one only image with the velocity at this place but in the last five, six years 40 flow sequences have started so we have improved our MRI studies our flow MRI studies with this new sequence that the properties that this sequence is the first one is that it's volumetric so we are not acquiring just one slice position we are analyzing one volume and as you can see here we can acquire the whole aorta all the volume of the aorta and all the velocities inside the aorta this sequence requires synchronization so we will have all the different phases in the cardiac cycle so we can have the different flows at different parts of the cardiac cycle in the diastole, in the systole, in the diastole whatever we want the thing is that these sequences this sequence also is related to the velocity encoding so as I said before we have to previously know the maximum velocity that we are dealing in that patient because we have to select the correct maximum velocity so we have the formation from the echo or we have the formation from D to the MR we have to select this maximum velocity because if don't we will have aliasing in this MRI too and as I said it's a volumetric so we have a volume and all the information, all the velocities included in that volume as I said before with 2D MR what we used to have was only we could only analyze the direction of the flow so in one plane, for example in this one in the z plane but with MR, with 4D flow MR we also have the information that we can codify all different velocities in all different planes so we can have the information through plane so in the z plane but we can also have the information the direction of the velocity in the y plane and also in the x plane so we can decompose all the different components of all different vectors of the velocity within the vessel so when we have one one segment when we have one vector of the velocity within the vessel then we can decompose all the different components and we can know exactly the direction of this flow so as I said we are not acquiring one slice so we are acquiring one volume that as you can see here we have like a box where all the volume of in this case we are doing a 4D flow MR of the whole aorta so we have all the information of the aorta in this vessel here and then we can select whatever we want so we can do an slice here or there we have that volume and we can work with that volume as we want and the information that we have is what we have here so we have the velocity the encoding direction in plane that it's right to left anterior to posterior so that would be in this direction here all the through plane velocity that would be in this direction like the z plane and also we have an anatomical image as we can see here just to know where we are doing our slash position at this image here this is the pulmonary valve the pulmonary trunk and the the refurcation of the pulmonary vessels the ascending aorta and the descending aorta so the advantage of this sequence is that we have a volume we have all the velocities of this volume that we are working and then we can select the slash position that we want we can know all the volume all the velocities at the plane that we want to analyze so the image that we have is like this one so this is one patient that is a normal volunteer so we have all the velocities that are within this aortic volume this is the pulmonary the pulmonary trunk and the refurcation of the pulmonary so for example if we want to know the relationship in between the flow in the ascending aorta and the flow for example in the trunk of the pulmonary artery so we just do a slice at this position another slice at this position and we have all the information about velocities, flow gradients, so all all the information if we want to do another slice here or in a pulmonary trunk or refurcation or whenever we want and compare different ratios or velocities so we can select the position that we want so that's the advantage of 40 flow so we have all the information all the information in the volume and also we can analyze all the different direction of the different components of this flow one of the advantages is at the beginning one of the concepts that we have is well we are using 40 flow this is a different sequence from the previous one is this sequence the velocities and the flow that we have are comparable to the previous one to the 2D ones that we used for a long time so there is a first work that is here that compared how the velocities in 2D and 40 flow were and as you can see here the velocities are pretty comparable so the information that we are getting the values that we are getting are truly comparable with the previous one and these are the clinical indications that we have in current guidelines regarding the use of 40 flow so we can use for the flows in order to get the information of the valves in the heart so stenosis or regurgitation it's very useful for congenital complex congenital heart diseases and also when we have chants or we have collateral vessels because we have the volume and then we can compare the flow at different vessels or at different cavities so this is important when we have we don't know at the beginning all the information that we will use and we can get a volume and then analyze individually different vessels or different components and also it's very useful in patients with aortic disease diseases of the the aortic disease like aneurysm of aortic dissections so these are the clinical uses of 40 flow that we have in current guidelines in the heart there's not a lot of the uses of 40 flow in leth ventricle and in the heart is mostly used for research purposes so there are some clinical trials that they are trying to use how the complex flow within the leth ventricle can associate with the formation of thrombus in patients when the heart is enlarged and they have plotted like a dilated cardiomyopathy so there are some works based on leth ventricle but the most the most studied the most clinical application of 40 flow has been studying the aorta the aortic valve valvular heart diseases and also the aorta the whole aorta and the flow in the aorta so based on the main implication of the main applications of aortic the 40 flow in aorta so first of all we have I wanted just to remind you how the aorta how the whole aorta works so we know that when we have the systolic blood wave the first thing that happens is that we have the aorta what it produces is an expansion of the aorta what we call the aortic buffering that means that there is a systolic expansion of the aorta just to this all this flow that has been out of the leth ventricle and it's this process is mainly induced by elastin fibres and smoth muscle cells and then when the systolic happens after that we have the aortic reservoir phase that what it does is just try to couple all this flow distribute the flow within the whole aorta just to distribute it to the rest of the vessels to the rest of the body and that produces a normal wave propagation of the pulse wave velocity and the normal propagation of the flow within the aorta this is the normal physiology of the aorta but we know that there are some pathological issues that can interfere with these normal mechanisms of the aorta and as you can see here when we have mechanical or chemical stress like hypertension inflammation we have atherosclerosis that's going to interfere and that's going to alter the reparative process of the aortic wall so it's going to increase the extracellular matrix it's going to produce atherosclerosis it's going to increase the collagen and the calcification of the aortic wall and that for sure it's going to produce an increase in aortic stiffness and a decrease in the aortic compliance so what we are going to have is an alteration of the aortic buffering capacity and aortic reservoir capacity what are the mechanisms or why we have a mechanical or how are what are the causes of this alteration in the aortic wall but clinically we try to distinguish them in between genetics or classify them in between genetic causes inflammatory causes or induced by flow so the genetic one are the syndromes that are genetic like Marfan, Ehler Danlos, Lois Dietz they are like inflammatory diseases of the aorta like arthritis arthritis and induced by flow we have hypertension atherosclerosis, traumatic or patients with bicuspid aortic valve that they have like a mixed component in between genetic and due to flow classically we use different parameters just to know how the aorta works the stiffness of the aorta works so mainly we analyzed indirectly the presence of hypertrophy in the electrocardiogram or by echo we saw the thickening in the aortic wall or we have like different parameters but now we have lots of parameters derived from MRI and for the flow MRI so mainly sure you have analyzed or someone explained you how we know the formation and the mechanics of the heart, the muscle the left ventricle so we can analyze the deformation of the heart, we can analyze the strain the strain of the heart, the deformation the circumferential strain, the longitudinal strain the radial strain so now one of the advantages of 4D flow image and MR is that we can also analyze different patterns of parameters of the formation of the aorta as you can see here so the first parameter that we can analyze using MR is aortic distensibility that is like the radial distensibility of the aorta as you can see here in systole the the radium of the aorta increases and in diastole it decreases a little bit so we know for example that with different pathologies like for example in Marfan syndrome the distensibility of the aorta is is decreased so the aorta is more rigid the stiffness of the aorta have increased and also we can analyze for example the longitudinal deformation of the aorta we know that in diastole the aorta is like this but in systole there is a prominent displacement of the aortic annulus towards the apex of the left ventricle so as you can see here in diastole the aorta the position of the aorta is like this and in systole it moves towards the apex so we have an increase in the longitudinal deformation of the aorta and this deformation the longitudinal deformation of the aorta is based on the properties of the aorta but also in the mechanics of the left ventricle too classically we used how as you can see here this is a patient a normal patient a Marfan patient we have the aortic annulus here this is the ascending aorta we can see the left ventricle here and you can see how the the length of the aorta is a little bit based on the mechanics of the left ventricle so we used to analyze the longitudinal deformation of the aorta based on how the aortic annulus displaces in different views but now that we have for the flow and we have volumetric images in all different phases in the cardiac cycle we can also analyze the longitudinal strain of the aorta based on the volumetric changes of the aorta in different phases so we can see the changes in longitude that the aorta is having in sister Leanda Yasteli and we can analyze the strain of the aorta another parameter that we can analyze and it's very important in order to know the stiffness of the aorta or the elasticity, the sensibility of the aorta is what we call the pulse wave velocity this parameter reflects how the velocity of the arterial pulse propagates within the aorta so as you know we are analyzing velocities so we need a space and we need time so as we know those two different points, we know the velocity in one point, point number one and the velocity in point number two we know the time in between the transit time between this flow and this flow and we also now know the longitude in between or the separation of flow in between these two points will circulate the pulse wave velocity and we know that if the pulse wave velocity is increased so if there is an increase in pulse wave velocity that means that the aorta is rigid so the stiffness of the aorta is increased and we have like normal values, we can consider compared to normal values if the aorta of one of our patients is increased it's more rigid in between the normal range but also one of the things is that using for the flow as we we try to use a pulse wave velocity we only could analyze for example as we can see here the difference in two different points for example here one and two and if we needed the information for knowing the pulse wave velocity in another plane for example the aorta so we have to do another position, another slash position here and compare the distance in between these two different areas so we couldn't analyze all the different curves of the pulse wave velocity within the whole aorta so just could analyze one different slash position and we have to consider when we use 2DMR is that when we are acquiring one slash position the rate of the patient can change in another acquisition so maybe the different intervals of timing could be different so that was a limitation that we have to analyze the pulse wave velocity but using for the flow we have all the different planes of the aorta sending aorta sending aorta and we can have all the different curves and we can get the pulse wave velocity that is the pendant of that best fit in the pulse wave velocity in all the curves so now we can have a better and accurate pulse wave velocity curve that is based on all different planes in all in the whole aorta so we are having much more information and much more complete that we used to have before and we know that pulse wave velocity as I said it's a pattern that it has been considered the gold standard for aortic stiffness that we use in clinical practice so when we have an increase in the aortic stiffness we have an increased pulse wave velocity and there is a correlation in between the pulse wave velocity with all risk factors of the patient so patients with hypertension, old patients patients with hyper trillis eridemia, smokers males obese people so they have an increase in stiffness and they have an increase in the pulse wave velocity and also when there is atherosclerosis in the aorta as you can see here no plaque lipid plaque but an hyper ecogenic plaque that means that it's a more established plaque as we can see here the pulse wave velocity increases as atherosclerosis progresses it's important and it has been considered the gold standard for aortic stiffness because we have lots of studies this is just a review we have lots of studies that they state that the more rigid the higher the pulse wave velocity is so the more rigid the vessel is so the more risk factors or the more events that the patient is going to have so in terms of cardiovascular events mortality and all-cause mortality so this is a pattern that when we estimate this pattern in cardiology or in clinical practice is because we need it because it provides information to our patients in terms of that this is a patient with higher risk for cardiovascular events so now that we want that we have like a genetic inflammatory so that different factors can induce alteration in the aortic wall and this aortic wall increases its stiffness decreases the the compliance and that we have parameters and with different technologies but with MR and with 40 flow MR that we can estimate how the stiffness of the aorta is or how reduce the compliance so now we are going to analyze just a little bit why these changes in the aortic flow or in the aortic in the aortic wall induces changes in the components of the aortic wall and they are related to atherosclerosis so classically we know that we have flow in in a vessel when there is a normal flow when there is no plaque so we have a low a normal a laminar flow and there is a low shear stress but when we have an atherosclerotic plaque the flow the the lumen of the the vessel is reduced so we have an increase in the velocity and increase in the high shear stress but after the atherosclerotic plaque what we have is like a turbulent flow because of the flow has released after the atherosclerotic plaque and we have oscillatory shear stress so we know that this oscillatory shear stress when this oscillatory oscillatory is low flow or shear stress happens they are going to activate different mechanisms of the aortic wall in terms of like metalloproteinases and all these molecular changes, monocyt activation procoagulant a procoagulant state so they are going to happen different mechanisms in the they are going to induce some changes in the aortic wall that they are going to start the atherosclerotic process as you can see here this is a endothelial image of a patient a normal patient or a patient with a normal flow with a low shear stress and here we have the endothelium there is a disarray in all the endothelial cells when we have an oscillatory slow flow and now we using for the flow we can also combine the information of the flow so we can analyze how the flow distributes within the aorta and this can be also translated in a map as you can see here that can show us the wall shear stress in the aortic wall so you can see here there is an increase in the wall shear stress at this area here there is a lot of turbulence of the flow at this part here we will see more examples later and also we can also analyze where are the places or where we localize areas where there is a higher turbulent flow not in terms of wall shear stress but when there is an oscillatory changes in the turbulent so what is called the oscillatory shear stress so we can also analyze areas where there is oscillatory shear stress we can also estimate the oscillatory shear index and we can analyze areas where there is this increases values of the oscillatory shear index so these are maps that we get from fro the flow images we can now we can get images when we can know in this patient for example where are the areas of high wall shear stress and these are areas that they are going to increase the size and this patient with different areas of increased wall shear stress is more prone to dilate the aortic wall and we can also have different maps with oscillatory shear index so that means that also patients with an increased value of oscillatory shear index has more probabilities to atherosclerosis and also aortic dilation you know better than me what wall shear stress is that you know that it's the displacement of the different components of the different layers of the aortic wall so as you can see here and the higher the wall shear stress is the higher that the tension of the aortic wall is and then the more probability to enlarge or to have that much of the aortic wall and increase the produce dilation in the aortic wall so this is what the wall shear stress is as you can see here this is a point here and you can see that the wall shear stress is just the displacement of the different layers of the aortic wall and it's important to note that the wall shear stress is based on two different components we have the longitudinal wall shear stress there are what we call the axial wall shear stress that is the wall shear stress produced in this direction in the longitudinal direction and we also have another component of the wall shear stress that it's called the circumferential component of the wall shear stress that it's produced because of the presence of turbulence of flow where we have flow that it's going not only in the laminar direction but also in a circular distribution as you can imagine wall shear stress is going to be low when we have a laminar flow but when we have complex flow when we have turbulences so we are going to increase the circumferential component of the wall shear stress and this can also produce alterations in the aortic wall and aortic dilation so here is an example of a complex flow using for the flow MR as you can see here we have like flow that is going up in this direction forward in this direction but also we have like flow that is going in it's going in different backwards in a backward direction as you can see here we have flow up but also flow that is going backwards and then up again so we are having like turbulent flow here that this flow that this turbulent flow it's going to produce an increase in the oscillatory shear index it's going to increase the wall shear stress index and it's going to produce enlargement of the aortic wall so that's be an important component of aortic dilation also we can analyze as we can as I said before we have like flow that it's in a forward direction but also we can analyze the flow that it's coming backwards and what we call the regurgitan fraction so it's like the fraction that it's even in systole is the flow that goes in an opposite direction in the lumen so and this flow the presence of backward flow the more backward flow that we have in systole that means that the most probability the more likely is the aorta to enlarge and this is also related to enlargement so the more enlarged the aorta is the more regurgitan flow in systole the aorta has and we know from different histologic papers this is one important paper that it was based in patients with bicuspid aortic valve yet here is that the presence of an asymmetrical wall shear stress so again pointing out the importance of wall shear stress in atherosclerosis or in the pathogenesis of aneurysmal of the aortic of the aorta production so when we have an asymmetric and an increased wall shear stress what it's produced when we have an elevated wall shear stress it's a reduction in the elastin fibres and increase in the collagen fibres and increase in the metalloprotein axis and also enlargement of the aortic wall so first of all these complex flows using for dmr have been described in patients with aneurysm of the ascending aorta so as you can see here there is a normal patient with no turbulent flow there is all laminar flow in the ascending aorta but as you can see here a patient with an enlarged ascending aorta as you can see here there are turbulent flow in the ascending aorta so that's related to the dilation and increasing wall shear stress adlation of the ascending aorta and also for the flow MR has been used also in patients with aortic dissection that even though that we have lower flows in the false lumen what we can see here is the false lumen is also associated with turbulent flow these areas where the turbulent flow is higher is associated with areas with increase in the dilation of the aorta the most important thing of the most commonly studied process in cardiology so far because I said that for the flow MR is quite recent so it has been used we have been using for the flow MR in recent years so the most the pathology that has been more widely studied using for the flow has been the bicuspid aortic valve we know that you know that the aortic valve is formed by three different casps but there are some patients that they instead of having three different casps they have two different casps so and this is associated with an asymmetrical opening of the aortic valve and because of the asymmetry in the opening of the aortic valve what we have is more probability to have complex flow and at the beginning what one of the first studies showed that patients with bicuspid aortic flow bicuspid aortic valve they have more complex flow compared to normal subjects that they have a more normal distribution and also they also have a different pattern of waltz stress in the ascending aorta these cubes represent the ascending aorta and as you can see here there is like an asymmetric increases in the areas of waltz stress in this part here lower waltz stress in this area here so we have like different components of waltz stress then we try to identify the presence of the different morphotypes of aortic valve how they change the flow in the ascending aorta so we have like different morphotypes of aortic valve they can generate different patterns of flow and for that reason as we will see later different morphologies of ascending aorta so different patterns of dilation and the only the first paper that established a little bit the correlation between the turbulent flow with aorta dilation was this one but it's from 2012 as I said we are using recently for the flow MRI and this study was based just in 13 patients with bicuspid aortic valve and 12 patients with tricuspid aortic valve and we saw that progressive dilation was higher in patients with bicuspid aortic valve was higher in patients with eccentric flow and correlated with the displacement the centricity of the aortic flow so we know from this study that at least in a small series of patients that the more complex flow the more severity turbulent flow the higher eccentric flow the more probability the patient has to enlarge the ascending aorta or the aortic diameters based on that we started to work with 40 flow 2 or 3 years ago we have like different papers that some are under review and some of them will be published sent in recent next month or before summer so we studied the different components of 40 flow in more than 100 patients with bicuspid aortic valve with not extreme ascending aorta dilation as you can see here so this is a patient and as we can see here at the beginning is that this is a normal patient that you can see that the isocenter of the flow is very central but when we have patients with bicuspid aortic valve we can see that there is an eccentric flow more anterior in one of the morphotypes of bicuspid aortic valve what we call and more posterior in patients with type 2 and as we can see here we correlate it we analyze the different patterns of the flow in ascending aorta as you can see here this is a patient with a morphotype 1 so and as we can see here the flow is eccentric we have turbulent flow here but the flow goes all the time in the anterior wall of the ascending aorta as you can see here this is the maximum flow that it's related here so this flow starts to be higher in the aortic root that's why as we will see here patients with morphotype 1 also increases the aortic root diameter because the flow goes in the direction at the aortic root and as you can see here all the flow goes to the anterior wall but in another subtype of patients with bicuspid aortic valve is what we call morphotype 2 so what we saw here is that at the beginning so in the aortic root the flow starts its posterior so it's more posterior than anterior and then rotates from posterior to anterior until the ascending aorta where all the flow becomes anterior as you can see here it starts posterior and then shifts to the anterior part and at the end of the ascending aorta the flow is totally anterior so those patients won't have an increase in the size of the aortic root but they will enlarge the ascending aorta distally before the aortic arch and as we can see here we can have a 3D representation of the how are the components of the wall shear stress the axial wall shear stress that I said before and the circumferential wall shear stress that is involved with the turbulent and the circular flow and as we can see here the wall shear stress is maximum and is higher in all the ascending aorta and in the anterior wall when we have a bicuspid type 1 and it's posterior and then anterior when we have a patient with type 2 bicuspid aortic valve and also because of that shift in the flow in patients with type 2 bicuspid aortic valve the circular components and the circumferential wall shear stress is increased in patients with type 2 bicuspid aortic valve higher than patients with type 1 aortic valve so as I said before there are some patients that they enlarge the aortic root as you can see here so this is a bicuspid aortic valve patient with an enlarged aortic root enlargement the ascending aorta is not enlarged and as you can see here the flow at this area is very higher this is a bicuspid aortic one bicuspid aortic valve so the flow is mostly anterior and also there is turbulent flow also in the aortic root that it's related with this increase in size in the aortic root and this is a type 2 patient that has an increase in the volume in the ascending aorta and as you can see here the flow starts more posterior and then shifts to anterior and that's why the reason they increase mostly the diameters in the ascending aorta we analyze the different components of the associations of the different morphotypes of aortic dilation we can see that patients with aortic root the morphotype, the ones that enlarge mostly the aortic root they have less turbulent flow compared with patients that they increase the size in the ascending aorta that they have more turbulent flow and also more reversal flow compared to normal patients of type 1 patients with the morphotype the root morphotype when we analyze the different components you can see here that as I said patients with enlargement in the ascending aorta they have more circumferential wall shear stress they have more normal values in the aortic root compared to patients patients that enlarge the aortic root that they have more values higher values in the aortic root and lower components of circumferential wall shear stress so just to conclude for the flow sequences constitute a useful alternative to quantify flow especially in case of complex cardiopathies dissection by cuspid aortic valve congenital heart diseases they present an excellent correlation with 2D phase contrast sequences with the advantage that they can perform flow study at every aortic segment because we have a whole volume of the aorta so we can analyze the flow whenever we want they allow a detailed study of complex flow and their functional repercussion in arterial structures and they allow a greater knowledge of aortic pathophysiology as I showed you before however at present it's post-processing it's complex requires experience qualified personnel team and it's very tight consuming because we don't have lots of software so far to analyze all these parameters so most of the parameters that we are analyzing are based on math lab and different software that engineers usually use and so far as these are new technologies new techniques we lack we don't have many prospective studies that determine the real pronostic impact of these variables that are being by 4D flow in our patients that for sure that in some years we will have more data just to correlate what we have at the beginning so these parameters that are increased with 4D flow MR and the correlation with events and also the aneurysm formation of aorta so thank you very much thank you very much for this extensive overview just maybe a bit of a practical question that this is like a complex acquisition is like maybe you can elaborate a little bit and how many patients can you do it and how much time does it take do you more regularly still use the 2D flow or in the end do you rely on ultrasound what's the practicalities of things so the practical thing is that we always manage our patients based of cardiac ultrasound that's for sure so we are analyzing we are using 4D flow MR just for research purposes or when we have doubts there are some patients that we have doubts for example in quantifying the aorta regurgitation because there is an excentrical flow we are not very precise with echo so in those patients we perform a CMR just to quantify the real regurgitant volume so we use it in complex cases that we have doubts in echo so mostly echo when we have doubts MR compared one we have selected our patient and we are doing a CMR in that patient because we have doubts in echo so we usually start doing is 2D MR 2D phase contrast because we need to know very precisely how is the maximum velocity so we have that velocity from the echo report or we have doubts we are doing a 2D acquisition at the aortic valve for example because we need to know the maximum velocity because it's the encoding velocity what I explained before that it's very useful to know because if we are using a low threshold of velocity we will have aliasing and the velocity the values won't be real so we can we do a 4D flow so it's an acquisition the advantage of that is that the patient is free breathing so it lasts the whole acquisition is like 8-10 minutes but the patient they don't have to collaborate at this time they only have to lay down breathe normally and it's more comfortable for them compared to the 2D acquisition that they have to collaborate they have to hold it breathe they have to hold it for 10-15 seconds and again and again every slice that we have to do so it's an apnea that the patient has to do so it's more stressful for the patients the 2D acquisition than the 3D and in total how much time would it take to do a full acquisition of a patient we are using a strict protocol and it's about 40 minutes the whole acquisition since the patient starts we have to take into account that the patient first comes to the MR room they have to lay down we put an IV for the contrast in some cases so all the process it's about 40-45 minutes and what about availability of the 40 flow sequences is it available on every scanner do you need a special one do you need special sequences the thing is that you need a special sequence you cannot you can't acquire a 40 flow volume a 44 acquisition if you don't have that sequence and most of the softwares of the MR scans at this time they don't have it so you have to to have an agreement with the company just they can give you the sequence mainly for research purposes it's not already commercial so most of the for example I don't know how many centers in Barcelona or in Spain are working with 40 flow but we are not many in the whole country working with 40 flow I think here in Barcelona there are two centers us and the people from Sampao the ones that are working with 40 flow so it's not extremely popular and you say the post processing is also complex and how long would it take there is it like 10 minutes of processing or 10 days of processing the thing is that you have like a learning curve so when you start you start everything so with the process you are you develop tools that make this post processing more automatic so Lydia started doing that that for sure she spent more than 1 hour per patient for sure so she can give you more information about that if you want to say something and now half an hour so yeah we started and how important is it also to have the anatomical information and for example have a segmentation of the aorta? Yeah so the anatomical information it's very important because what we have here is that patients with 40 flow they don't have the when we are analyzing dissections or bicuspid artery valves aorta male that's what we are studying so far is that what happens is that there is a symmetry of the components of the water stress the flow is directly to one direction so we need to know the anatomical references in order to know that the flow is in the anterior wall or is in the right part of the aorta so we need the all the anatomic references and the flow because we will know where the symmetry is based on the anatomic references yeah. It's a question there. Hi thank you for the presentation I have two questions. Both questions are related with the wall shear stress one of the first one is you talk about the wall shear stress how is the resolution close to the wall and how is accurate your segmentation because in the MRI when you segment the wall it's not quite well defined so I don't know how you compute this wall shear stress because it's just the gradient of the plotity and the second one is related with the OSI this is the temporal I don't know I see correctly the wall is the move is rigid so how is just the wall shear stress the wall is move of the aorta and if you I suppose if you want to compute the OSI value okay it's just along the cardiac cycle and the wall is move so I don't know see if the wall is move in the video I don't know just to know how to compute this the values for the first question I will require the engineers because they know about segmentation much more than I do what we do for wall shear stress and for oscillatory shear index we give the most the most information that we need is what is happening in the systolic phase when the flow is moving forward so we use the peak of systolic the systolic flow just to calculate the oscillatory shear index and the wall shear stress but because of the noise and just to mitigate this phenomenon what it's considered or well established in the community that we are using for the flow is just analysed also not only the peak but also the previous one and also the different cycles I mean one previous and one posterior and just to make a mean an average of the values just to know the wall shear stress the maximum wall shear stress and the oscillatory shear index so mostly the maps that we are getting are systolic or an average of the systolic so previous and posterior so the systolic more or less the systolic phase okay but how is the resolution close to the wall because okay I can answer the resolution in general it's 2.5 millimeter and box so it's a cube and this is not very good for wall shear stress computation because it's being a derivative a special derivative we all know that it depends on the resolution but about comparative study in clinical setting you the main thing you know as long as it's not becoming a clinical tool is a relative so if you use the same sequence you will get the same problem and will propagate the same but in general we use spline to get read of a lot of noise close to the wall in general we use common common of the same yes function to velocity along the wall in the radial direction so we approximated the point we have as close to the wall with spline or also other equation but I repeat the point here is that being a research we don't really need the absolute value but much more relative so we have different patient with different pathology what we are looking for it's much more how different phenotypes or morphotype or pathology relate each other so the day that this will become it does a clinical tool there will be need a consensus on how to behave in this aspect I understand what you say but the thing is for example you have a method to compute the worst stress okay maybe in other hospital and other tools they have another method to compute the worst stress there are consensus about that so how for example the data you have in the Balder Braun can you compare versus against sample for example yeah that's a problem that we are having there we know that different machines they have the different sequence the different sequence they have different values we know that there is no consensus about normal values just because of these irregularities for sure but that's why we are relying more mostly in maps I mean or in a general representation of what is happening because we don't exactly want the exact value of this parameter so we want to know that for example that an increase in the circumferential worst stress we are very sure that an increase in circumferential more than the axial worst stress it's important and patients with an increase in the worst stress in every part proximal aorta or distal aorta so they are going to enlarge the aorta much more than patients with lower values of worst stress and mostly the results that we are getting mostly based on these are regularities asymmetries different patterns much more that in the clinical value that we don't get that because of these irregularities and these inhomogenities that we have in different vendors different machines different yeah any other questions if not then it's lunchtime it was a long morning so for the ones that doing the hands on please come back at 2.45 and it's in the rooms of yesterday additionally the ones that were doing the lumped model for the fetal that will be in room 30 but there will be indications there