The title of my talk is the Changing Landscape of Atherosclerosis Imaging. It is well known that atherosclerosis play a fundamental role and that some organs affected by atherosclerotic process are related with some pathologies. An example, also ancient Greeks know that carotid artery pathology was related with the occurrence of cerebrovascular events. You can see in the matop of the parthenons that the centaurs is compressing the carotid artery of the king of Lepidus. But 2,000 years were requested to better understand that atherosclerosis was pathologies involving the entire body and that the prevention and cure of atherosclerosis required the stratification of the severity of pathology. This concept in particular was lead by Fisher Miller and the problem was how we could stratify the severe of pathology. One answer was done in the 80s by the Nashet group that used the degree of stenosis for the quantification and stratification of the risk and the severity of pathology. The degree of stenosis was used because in that phase, the only imaging technique that allowed to study the carotid artery was the angiography. You are observing the graph of the Nashet study that showed that patient that underwent carotid endarterectomy are protected with the occurrence of several vascular events versus those patient that are not treated with carotid endarterectomy. This was a dogma and you see that also current guidelines incorporates the degree of stenosis as the lead parameter for the choice of the therapeutical option. This one is a guideline from the European Society of Vascular Surgery, but it's the same if we check the American Heart Association. The lead parameter is the degree of stenosis. This is the guideline released in 2021. The degree of stenosis is not the atherosclerosis, is the indirect parameter of the atherosclerosis. In the following 40 years, a lot of evolution has occurred in the imaging, but not only in the imaging, also in the biological understanding of atherosclerosis. We should better focus on the deeper knowledge of pathological process. To understand the technological evolution that we are currently facing, we should focus also that some concepts are not set in the stone, the impact of artificial intelligence and also the drugs impact, and their monitoring. Regarding the knowledge of atherosclerosis, emerging evidence has spurred a considerable evolution of concepts relating to atherosclerosis and has called into question many previous notions. What is atherosclerosis and more specifically the plaque? We can say that the plaque is the dynamic structure with an extremely high entropic level of heterogeneity. If we look one plaque into a patient with a high level, so electronic microscopy, we can see that there are a lot of components, a lot of architectures, and there is a lot of heterogeneity. These are microscopic electronic imaging from our institute, and you can see that in the same plaque there are a lot of levels and a lot of different architectural tissue representation. This is important because atherosclerosis is a pathology and is not characterized by the organized level of architecture of organs like liver, such as an example, the prostate, such as an example. But it's an extremely variable with multiple phases representing a spectrum of vulnerability and different pathological behaviors. Due to this level of heterogeneity, we should move to imaging biomarkers of vulnerability or stability. Another important point is the technology evolution that is present in CT, in MRI, in the ultrasound, and technology in CT is evolving in particular with the photon counting technology, with the evolution and creation of new and more advanced sequences, whereas in ultrasound with the hardware evolving 3D technology and algorithms. Let's pick one example. Look at this specimen from a carotid endarterectomy, and look at this image. It is an example of photon counting analysis, and you can see that the amount of information we can observe from the photon counting is much more than the conventional CT technology, because we have information regarding the atomic composition distribution of tissue. We can distinguish very well the iron from the calcium, and the different types of calcium configuration, for example, and also more understanding in the soft tissue composition differences. It's a breakthrough impact, and you see that it has obtained the Nature Reviews cover. It's also important to underline that some concepts are not set in stone, and one of these is the reproducibility of unfilled units value. That is a very important point because we use unfilled units to a metric system to classify the tissue, but there are evidence that is not so easy. Look at this plug, it's the same plug at different level of energy. In this case, 66 kiloretron volts, and in this case, 86 kiloretron volts. The attenuation is completely different, in this case, 95, and in the second case, 57. The same plug with different energy values shows different unfilled unit attenuation. This is an impact in the classification of the tissue, and we have also to remember that it's not a matter of energy in kilovolts, of kiloretron volts, but with the new reconstructive interactive methods, it's also a matter of ampere, and also a matter of body max index. There are a lot of parameters that change the equation of the unfilled unit attenuation. Plug categorization and the tissue, atherosclerotic tissue categorization using unfilled units is weak, and we need to find a standardization for that. The quantitative imaging biomarker alliance is trying to identify a new metric model, and the possible solution could be the texture analysis that is less dependent to the energy level. Currently, we are in the stage of consensus for this document, for the QIBA model. Another important point is the impact of the artificial intelligence. You are watching a plug in the carotid artery where the voxels show different color according to the unfilled unit attenuation. In the past, we needed to trace the inner lumen and the outer lumen in order to obtain the plug segmentation and the plug volumetric analysis and tissue composition analysis, so in order to have this type of results. But currently, with the evolution of the artificial intelligence, this step is very easy. You see, these are experiments in 2012, so 10 years ago, that compared the algorithmic green line and the human reader dashed white lines, and you can see that the algorithms are similar to the human reader. To obtain this type of level of recognition through artificial intelligence algorithms allows us to easily obtain this type of information, so an automatic segmentation of the atherosclerotic process. And this is important because we are moving from the qualitative assessment of the plug to a bio-quantitative assessment and analysis. And if we have information about the volume and composition and the metrics of the plug, we can move to the radiomics, so the texture analysis with a lot of features that allow us to better stratify composition and hopefully outcome. And it's an important evolution field in this area. Another important point is the drug impact and the monitoring of the drug. We know that a lot of new drugs play a key role. In the past, we understood how statins and anti-lipids drugs could help, but we have also important and robust evidence that anti-inflammatory therapies play a role in the plug evolution, stopping, and regression. And this is one of the most important information we could obtain in the atherosclerosis domain. And we have also robust evidence that atherosclerotic disease can be changed with the drugs, and there is a variation in the composition, and in several cases, also a plug relation. You can see the change in this plug after statin treatment. This is important because we should track the evolution of the plug after a drug, and we also have to remember that atherosclerosis is a multivessel disease, and for the treatment decision-making and tracking, we should consider also the effects to multiple targets. So we check carotid, but we have effect in intracranial and in coronary, and so it's a unique circle of effects. And this plays a role because there is some evidence that carotid plug, for example, components and fibrous cap stop predicts the system cardiovascular outcomes. And we think that maybe it's possible that biomarkers of carotid plug vulnerability could be also a biomarker not only for stroke, but also for systemic atherotrombotic risk. Are all these evidences enough to change our mindset? These are papers we have published in the American Journal in 2021, issue of September, and we try to understand what are the steps necessary to incorporate potentialities from imaging and what is the landscape in terms of biology of atherosclerosis, impact of the drug, so strategy in terms of therapies. And in order to incorporate an imaging and better use an imaging, we should apply for different step, and different step are important because are the demonstration that our imaging play a role for the tracking, for the detection and for the strategies for the therapies. In conclusion, imaging is a rapidly evolving technology and there is a terrific potential for the detection and characterization of the atherosclerotic process. The target for the future is to use advanced imaging techniques to identify a model of risk based on the imaging features that allows to stratify the risk, to select the best therapeutic approach and to monitor the progression and regression of the atherosclerotic process. Thank you very much. So hey everyone, I'm Mahmoud Mosabasha from the University of Washington, and I'm going to be talking to you about stratification and outcome prediction in stroke imaging. We'll talk about embolic stroke of unidentified source and association with atherosclerosis, plaque characteristics associated with current and recurrent stroke, and blunt cerebral vascular injury and associations with stroke. So first, I want to talk about a couple studies associated with relating to cryptogenic stroke and ESIS. Specifically one study where they looked at 32 cryptogenic stroke patients with less than 50% carotid stenosis. What they found was that in 37.5% of cases there were advanced atherosclerotic plaques, ipsilateral to the stroke, while there were none on the contralateral side. And the most common findings in these complex plaques, advanced plaques, was intraplaque hemorrhage, fibrous cap rupture, and or intraluminal thrombus. And so in this setting, the thought was that this probably contributed to these cryptogenic strokes that were occurring. Another study, however, on the other side of the coin, was one that was published relatively recently where they looked at histology of clot in the setting of mechanical thrombectomy in which they looked at 77 cardioembolic, 46 non-cardioembolic, and 64 cryptogenic cases. And they analyzed the composition of those clots, and what they found was that in terms of fibrin content, the cryptogenic clots had similar fibrin content relative to the cardioembolic source. But there was a significant difference in fibrin content between cryptogenic versus non-cardioembolic. In addition, when looking at other factors such as RBC and white blood cell content, there were differences between cryptogenic and non-cardioembolic. So in this setting, their sentiments were that likely cryptogenic more closely associated with cardioembolic and likely had a common etiology in that setting. What I will mention is in terms of clot composition, there definitely is a spectrum of composition even depending on the etiology. So when you look at all cardioembolic clots, there's a spectrum of fibrin content, RBC content, and so on and so forth. Next we'll talk about vessel wall imaging and stroke risk stratification. So this was a study from the Toronto Group in which they looked at vessel wall imaging and its impact on stroke etiology. And they found that vessel wall imaging altered or modified the TOSE etiologic classification in 55% of cases of stroke patients. The most impacted categories was intracranial arteriopathy, not otherwise specified, went from 31% to 4%, and intracranial atherosclerosis went from 23% to 57%. And both these changes were significant changes in terms of classification. So in that setting, vessel wall imaging can certainly provide additional information to allow you to categorize the etiology of stroke. And here's an example from our own practice of a patient who was having right MCA, shower emboli. DSA was completely normal, subaclinoid ICA, MCA on the ipsilateral side. Carotid was normal, Holter moderating. Complete diagnostic workup was normal until we did vessel wall imaging on the ipsilateral side. There was an outward remodeling plaque that was growing outward, not much impact on the lumen, and this was the presumptive etiology for stroke events in this patient. And after maximum medical therapy, there were no subsequent stroke events. The next study we're going to look at from China, they looked at 243 ESUS patients and 160 small vessel disease stroke patients, and they found that plaque was significantly more prevalent ipsilateral to the stroke in ESUS cases than contralateral, than on the contralateral side with an odds ratio of 5.25. In addition, they found that plaque was equivalent on both sides for the small vessel disease patients, and the remodeling index, or the degree to which a plaque is growing outward, was independently associated with ESUS. And in that vein, talking about remodeling, plaques frequently can grow outward, and in that setting they can achieve a fairly prominent plaque burden and lipid core before they result in any appreciable stenosis. In addition, they can reach fairly advanced stages with complex plaque in that setting. Conversely, and less commonly, plaques can result in constrictive remodeling in which a small plaque burden can result in luminal stenosis, but usually that's a progressive process. And this is from that same Chinese ESUS study, in which looking at the ROC analysis, they found that the remodeling index and all these curves up here, and I apologize for the small type, but all these curves up here on the ROC analysis all involved remodeling index, and the AUC for remodeling index alone was .74, outperforming a lot of other imaging characteristics associated with ESUS. Another category that we'll look at is the radial location of plaque along the vessel wall. And this study looked at that along the MCA and found that plaques along the superior and superior posterior aspect of the MCA, where the perforators, the lenticulostriate perforators come off, was associated with deep basal ganglia perforator territory infarcts. And so this is another thing to keep in mind, is that those perforator infarcts can be secondary to atherosclerosis in the setting of osteoocclusion. And here's an example from our practice involving the basilar artery in which you see an atherosclerotic plaque here along the posterior and parasagittal wall of the basilar artery, which is where the pontine perforators originate from. And when you look at T2, you can see that plaque again right here, and there's a little branch coming off of it right where that plaque is located. And you look, there's an old infarct involving the left side of the pons, likely contributory secondary to that atherosclerotic plaque. Next we're going to talk about plaque enhancement. And the first study we're going to talk about is a meta-analysis from the Cornell group in which they looked at eight studies with 295 atherosclerotic plaques, and they found that there was a significant association between atherosclerotic enhancement and stroke with an odds ratio of 10.8. The next study is from the Toronto group where they were looking at the relationship between age of stroke and degree of enhancement, and they found an association between acute stroke and strong culprit plaque enhancement. Subacute and chronic plaques enhanced less, and there was no plaques that showed an increase of enhancement over time where they had a number of longitudinal evaluation of plaques, which we can see in the lower graph here. And here you can see strong enhancement, mild enhancement, none, and as you get out further from the stroke event, that enhancement progressively decreases. The next study is from the Hopkins Group in which they looked at enhancement degree relative to culprit, probably culprit, and non-culprit plaques in patients. And they found that grade 2 or higher grade enhancement was associated with culprit plaques with an odds ratio of 34.6. Culprit plaques showed higher quantitative enhancement, which was a significant difference. But in addition, a lot of plaques do enhance between both culprit and non-culprit plaques, 73% enhance. But what they found was that in the setting of a non-enhancing plaque, those were only non-culprit. So if you see a non-enhancing plaque, there's a good possibility that it is a non-culprit plaque. The next study is a study that we did in which we looked at 176 patients with 702 plaques stratified into first-time acute stroke, recurrent stroke, or chronic stroke. And what we see on the left is the comparison between those three categories and the culprit plaque enhancement ratio or quantitative measure of plaque enhancement. The recurrent stroke group had a significantly higher degree of enhancement relative to the other groups. And on the ROC analysis on the right, we see that culprit plaque enhancement ratio, culprit plaque burden, and total plaque number all had AUCs above 0.7 in terms of their relationship with recurrent stroke. And then in terms of plaque burden and recurrent stroke, we did an analysis looking at 58 patients in which we followed them out for 48 months. And in this category, a number of characteristics were associated with the recurrent stroke, including triglyceride level, plaque burden, and percent change of plaque burden. The strongest relationship on the AUC curve you can see is the progression of plaque burden on vessel wall imaging. And on the Kaplan-Meier curve here, we can see that those patients with progression of plaque burden had a much stronger likelihood of stroke events that occurred relatively rapidly in the longitudinal evaluation period. In terms of fibrous cap assessment, we do see fibrous caps on vessel wall imaging, and they appear as T2 hyper-intense component superficial to the deep T2 hypo-intense component that represents the lipid core, which we can see here. And in the setting of thinning or ruptured cap, these are associated with stroke events as we've seen in the carotid literature and also some of the intracranial literature. So here are a few studies that looked at ex vivo imaging of intracranial atherosclerotic plaques. And one of the studies that I'm highlighting from Jeng et al. looked at 53 plaques from 20 cadavers, and they found that the fibrous cap was hyper-intense relative to the lipid core, as we just mentioned, on T2. And T2-weighted imaging was the sequence that was best at representing that fibrous cap relative to the rest of the plaque. And they found 80% accuracy of MRI in identifying fibrous cap and other plaque components relative to histology. So next, a couple studies, one of them from us, in which we looked at fibrous cap composition or fibrous cap visualization association with symptoms. We did a pair assessment of symptomatic and asymptomatic plaques, and we found that there was a significant association with plaque visualization and symptomatology, where thick fibrous caps, such as this one here, where you see a continuous T2 hyper-intense band, were not associated with stroke events, while if that fibrous cap went missing in any segment of the plaque or there was interluminal irregularity along the surface of the plaque, those were much more frequently associated with symptomatic events. In terms of intraplaque hemorrhage, this is an important designation, especially for carotid plaques, but some studies from Asia have looked at intraplaque hemorrhage in the setting of intracranial plaques, and these are mostly from China. But generally, intraplaque hemorrhage has been shown to be associated with enlarged necrotic lipid core and destabilization of the plaque. Two studies, one vessel wall imaging and one histology, found that anywhere between 20 to 30% of symptomatic plaques intracranially had intraplaque hemorrhage, while only 3% to 15% of asymptomatic plaques showed intraplaque hemorrhage, and these were generally represented by T1 or proton density hyperintensity greater than 150% of muscle tissue seen within the field of view. But as I mentioned, this is less common in Western populations. It's much more frequently seen in Asian populations. So in terms of extracranial atherosclerotic disease, similarly, extracranial disease can be a source of ESUS, and characteristics associated with it are intraplaque hemorrhage, ulceration, disrupted luminal surface, lipid core volume, and neovascularity. Intraplaque hemorrhage is the most established risk factor for carotid vulnerable plaque. It's detected the same way that we just described for intracranial plaque, 150% signal. And it's associated with a four to eightfold likelihood of lipid core progression with intraplaque hemorrhage, 35 hazard ratio for future stroke, and lipid core generally does not regress with statins as it does with other plaques in the setting of intraplaque hemorrhage. And one thing I'll mention in terms of risk stratification, luminal imaging, as we were hearing in the prior talk, does not always associate with plaque progression and potential events. And in this example, we can see in the 3D rendered view that the plaque progresses. It grows in size over a year. But that lumen isn't really affected. And what you can see on the vessel wall imaging is here's that plaque. You can see it grows in volume, and in addition, there's intraplaque hemorrhage that develops over time, indicating the instability of that plaque. Next, we're going to talk about blunt cerebral vascular injury quickly. And so in terms of blunt cerebral vascular injury grades, we can start off with a normal vertebral artery. And then once you get hit, there's some stenosis, grade one, where there's less than 25% stenosis. Grade two, where there's intraluminal thrombus adherent to the injury, or greater than 25% stenosis, as we see in this example. Or there's a dissection flap in a false lumen, which also falls into grade two. Grade three is where there's a pseudoaneurysm that develops. And then grade four is where there's occlusion of the artery. And lastly, grade five is when there's transection of the artery. So in terms of stroke rates in blunt cerebral vascular injury, this is a study we published not too long ago, where we found that there is a significant association with the number of injured arteries and the likelihood of stroke. In addition, the highest grade of injury is also associated with stroke events in a significant manner. Now in terms of stroke and BCVI, we looked at stroke rates in inpatients as well as stroke rates once patients are discharged on long-term follow-up. And we looked at Washington State rehospitalization records and vital statistics from the DOH. And what we found was that in a fairly large cohort, we found that the hazard ratio for inpatient stroke and BCVI is higher than non-BCVI patients with a hazard ratio of 4.98. And the mortality was higher in BCVI patients as well. Conversely, post-discharge, there was no increased rate of stroke in injured patients and no increased risk of mortality in these patients. So in this setting, on the inpatient side, there's an increased risk of stroke. But once they're discharged, there's no increased risk or they may potentially have a reduced risk. And lastly, just to mention dissection and BCVI in the setting of trauma and imaging, CTA is really the standard that we use, the reference standard. It has its limitations in terms of detection. We found in a study using vessel-wall imaging that there is increased accuracy with vessel-wall imaging, especially for low-grade injuries, grade 1 and grade 2. And in that setting, vessel-wall imaging could potentially, as a second-line imaging modality, save the patient from additional follow-up, additional imaging, and medication. So in summary, plaque characteristics on vessel-wall MRI are associated with current and future stroke events. Intracranially, plaque burden enhancement are most substantial, while extracranially, intraplaque hemorrhage. Vessel-wall imaging can indicate plaque characteristics associated with ESUS and cryptogenic stroke. And BCVI, higher-grade injuries, and number of injured arteries are associated with stroke events. Thank you so much. Good morning. Welcome to Chicago. I'm sitting here in Rosenheim. And thank you very much for the invitation and for the opportunity to give this talk. I will talk about controversial topics in carotid plaque imaging. So controversial topics which I identified are what are the indications for plaque imaging. And so I want to talk about different concepts of patients with symptomatic and asymptomatic carotid disease and for the general population. Also what is the best imaging marker? How do you define the best imaging marker? And what is the best imaging method to do carotid plaque imaging? So I will talk about these topics today. So if you look at stroke, it's one of the major causes of death and leading cause of permanent disability. Symptomatic risk is really high, 25%. And the number of incident strokes is expected to increase by 30%. So actually, we do want to improve stroke prevention strategies, like for primary and secondary prevention in symptomatic and asymptomatic patients. So this is why we are all here. So you have this guy with this carotid plaque. And actually, this guy wants to know what he needs to do, whether he needs surgery or if he can have a stent, or if maybe medical treatment is the best option. And at least in Germany, we have this increasing problem that people now trust more the YouTube channels than the doctors, but that's maybe a topic for another day. So the European Society of Cardiology and Vascular Surgery, they published guidelines in 2017. And they are interesting for one reason, because at least to my knowledge, it's one of the first guidelines that included in asymptomatic patients the use of plaque imaging. So if there is one feature suggesting a higher risk of stroke, then actually, the likelihood of revascularization should or may be considered. Other interesting subgroups here in this chart are patients with less than 69% stenosis, which are symptomatic. And we'll talk about those groups later on. These features that were mentioned in this paper or in these guidelines that are associated with increased risk of stroke in patients with asymptomatic carotid stenosis are actually interplaque hemorrhage by MRI and lipid-rich necrotic core, and also large plaques, echolucent plaques, and hypoechogenic areas by ultrasound imaging. These are some of these features of the carotid vulnerable plaque, as you can see here. You have these different imaging modalities. And some of these features, as I already mentioned, others are neovascularization, carotid plaque thickness, surface morphology, and the carotid plaque volume. And actually, these high-risk plaque features are very prevalent in asymptomatic carotid stenosis. This is something that I want to point out. So in this really nice paper published in Chamer in 2020, almost 27% of all asymptomatic patients had one of these different plaque features that are believed to be high-risk plaque features. And what I found interesting in this publication is actually, first, they looked at the different decades of publication. If you look at the odds ratios for events, if you have one of these high-risk plaque features, actually, they didn't, or like they rather decreased when increased. So actually, in the last 20 to 30 years, we were not better in identifying high-risk plaque features as we were like 30 years ago. Another interesting point that I do want to point out is that the risk of, if you have these features, is increased independently of the degree of stenosis. So in all stenosis subgroups, there's an increased risk if you have a high-risk plaque feature. So that makes it so worthwhile to really look deeper into these type of features. And okay, these are the risk increases. And what I just want to point out, so actually the number of events here, it's in many studies, not so high, like only like in total 200 events in these high-risk groups, even if it looked at all these studies done in the last 30 years, so, and including TIA and amaurosis fuga, so it's not really a lot of events. So what is the best imaging method to identify vulnerable plaque features? And actually, this is a paper published from a carotid imaging consensus group. And this group of experts, they looked at all these different features. And as you can see here, MRI is a really very good imaging method for most of these features, better than all the other imaging methods, and so it's a really good imaging technique. In this other paper, we also looked at what is the best carotid plaque imaging method in prospective outcome studies. And we were interested not in like individual small studies with a couple of like 100 patients, no, we only wanted to look at the existing meta-analysis at this time. And what we were able to show is that MRI has several plaque features that are interesting, like IPH, fin or ruptured fibroscap, and lipid-rich necrotic core. And also there is good evidence that ultrasound can detect certain features such as plaque echolucency and IMT, which are the increased risk of stroke. And here are the hazard ratios and risk ratios, and I don't want to go into more detail at this page, but I want to show you one more study in more detail. But also what I think is really important to point out, we do have these imaging techniques such as CT and chiropractic, PET-CT, contrast enhanced ultrasound, OCT and IVAS. But at the moment, there are no studies that look at, or no meta-analysis, no larger studies that look at plaque imaging features detected by these techniques in prospective outcome studies. And especially with CT and chiropractic, it's so widely available, like there are like thousands and hundreds of thousands of patients, probably stroke patients, can get this technique. So if there are good features, I think it should be possible to find out which are those, such as, especially like large soft plaques. But at the moment, these studies are missing in action, so I hope that we will see more of those in the future. And I want to look a little bit more in detail into a meta-analysis that was published the same time that we saw the review in Lancet Neurology. And it actually, the strength of this study is that it's a meta-analysis of individual patient data. So all of these authors, they did their separate studies and they provided their really individual patient data. So we were actually able to do like more detailed analysis, and we only looked at stroke patients. So all of these events are strokes. We didn't include TIAs or amaurosis fugors. First, IPH, black haemorrhage, was present in 52% of patients with symptomatic stenosis and only in 29% of patients with asymptomatic stenosis. Again, 66 strokes occur. And if you look at these hazard ratios, in symptomatic patients actually the hazard ratio was 10.2, and in asymptomatic patients 7.9 for stroke during follow-up, which is quite high. And actually, this was true independently of the degree of stenosis in the symptomatic patients. And this stenosis group seems to be really interesting, patients with less than 50% stenosis who really have a high risk of stroke during follow-up if they have IPH at baseline. So if you look at the major causes of ischemic stroke, I mean like we as black imaging people, we are usually interested in large artery atherosclerotic stroke, which is defined by more than 50% stenosis. But I think a very interesting area of research is these patients that have an undetermined stroke etiology, which is actually up to 30% of all patients. And we started to look into this like a couple of years ago when we had this patient who was categorized as a cryptogenic stroke patient, and he actually got his stroke while squeezing a pimple in the morning while shaving. And so as you can see here, these are his small thrombobotic DVE positive lesions. And you can see his carotid artery here. So there's some luminal narrowing, but it wasn't stenosis by ultrasound. And he actually has a really large plaque here of four to five diameter, and he has a fibroscapular rupture here, as you can see on these top images and also on this lumen irregularity. And like one slice further up, he has a very large plaque with interplaque hemorrhage. So the hypothesis was that he didn't squeeze only his pimple, but he also squeezed his plaque and got his stroke due to plaque content. So this was one of the major reasons why we actually started this carotid plaque imaging in acute stroke study, CAPILE study, which included patients older than 49 with acute ischemic stroke. In the anterior circulation, we can last seven days with carotid artery plaques of more than two millimeters, and patients with carotid artery stenosis more than 70% were actually excluded. So our starting hypothesis was that patients with cryptogenic stroke had more complicated plaques ipsilateral than contralateral to the stroke, and also compared to a reference group. And these are our results, and actually what we were able to show is that patients with cryptogenic stroke had a significant higher number of complicated plaques ipsilateral to the stroke than contralateral, and also compared to the reference group. And also there was a positive control group, so these patients with large artery atherosclerotic stroke, they had almost 70% rate of complicated plaques compared to 31% in these patients with cryptogenic stroke. So actually both of our primary endpoints were positive, and actually this is a study which is currently ongoing, and the review is ongoing. We actually submitted the data of this follow-up and looked at the event rates. I want to show you two cases. I'm not able to share all the results. But this is a patient with the index stroke and then he had a recurrent stroke on the same side three months later. And as you can see here, he has a very large plug with interplug hemorrhage. Another patient, this patient was published actually, has like you can see very normal lumen in the internal parotid artery. But he has a plug with interplug hemorrhage at baseline, high signal on T1-rated images. And during follow-up, he developed a new ulceration in this plug. So part of this plug went further downstream and this actually probably caused his amaurosis fugax, which he experienced. So the best imaging method, in my opinion and in the opinion of a consensus group of carotid imaging experts, is MRI to identify vulnerable carotid plugs. And the most promising imaging marker is IPH. And IPH imaging can be performed with standard coils during routine clinical brain imaging. And you just need like two to five minutes scan times. Possible indications for plug imaging, the most promising are patients with asymptomatic stenosis more than 50%, symptomatic patients with less than 75% stenosis. So in the future, what we actually want is to have this risk assessment and treatment recommendation tool. And I think this tool should include plug vulnerability as one marker. And I hope that we will have this in the near future. I think it should be possible if we all work together. And to do this, actually, Professor Sava, he outlined his roadmap or an idea of a roadmap to implement plug imaging into clinical routine. And I think we are still in an early phase of doing this. But I think the potential is there. And we'll see what the future brings. So I would like to thank you for your attention from Rosenheim. So actually, Rosenheim, the name of my town where I live in, if you translate it, it's the home of the roses. So you can see this is my garden. So actually, I just wanted to share this with you. And now I hope you have a nice discussion. Thank you very much for your attention. Good morning. I am Hediye Baradaran. And I'm going to be discussing using vessel wall imaging to evaluate the intracranial vasculature. So in this short talk today, we will review the basics of intracranial vessel wall imaging by answering very simple questions about it. First of all, what is vessel wall MRI? So in its most basic sense, vessel wall imaging is imaging that uses high resolution to evaluate the vessel wall rather than just luminal narrowing. So before the use of increased vessel wall imaging, we have really been relying on the degree of narrowing to evaluate the intracranial vessels with the gold standard, of course, being digital subtraction angiography and other commonly used techniques like MR and CT angiography used daily in the clinical setting. So vessel wall imaging, as the name implies, involves imaging the actual wall of the intracranial vessels to get a better sense of what is happening outside the lumen and to evaluate potential causes of narrowing and wall inflammation. There are a number of situations in which imaging the vessel wall is important and can be used clinically. So I'm going to first talk about what I would argue are the two most critical and commonly performed uses of vessel wall MRI, and then we'll go on and discuss some of the others. So the first reason is to differentiate between various etiologies for vessel narrowing that we can identify on angiographic techniques. So basically patients who are presenting with intracranial stenosis of unknown etiology. So we're going to use a couple of cases to illustrate. So recently we had a 44-year-old woman who came into our ER and she was having transient right-sided weakness and some speech difficulties. So since she was young and her symptoms had already resolved, the clinical team had a relatively low suspicion for a true infarction. She still underwent a neuroimaging workup. Luckily she had no evidence of acute infarction on her MRI, but we did see that she had focal severe stenosis of the M1 segment of her left MCA seen here on this coronal head CTA. Because her symptoms were also referable to the left MCA territory, she was thought to be having a TIA. So this degree of narrowing in someone who's in their 40s is quite unexpected. So this was a perfect case to recommend a vessel wall MRI to further evaluate why she was narrowed here. So on the MRI, we saw that she had focal area of predominantly T2 hyper intense plaque that was positioned somewhat anteriorly in the vessel wall that we can see here on this axial T2. On the sagittal reformat, we see an eccentric T2 hyper intense plaque narrowing the lumen. As a comparison, here's another sagittal image proximal to the area of narrowing where we see the normal caliber of the M1 segment right here. This area of T2 hyper intensity was actually also associated with avid enhancement on the post contrast sequence. So putting all of her imaging findings together, she had an area of eccentric T2 hyper intense plaque, which was severely narrowing the lumen. It had avid enhancement. So together, these findings are consistent with an atherosclerotic plaque, and this plaque was thought to be the cause of her recent TIA. So it was considered a so-called culprit plaque. So despite her young age and her relative lack of any known risk factors, based on the vessel wall MRI, we were able to conclude that she did have findings of intracranial atherosclerosis leading to her symptoms. She went on to be appropriately treated for this area of narrowing. So here's another illustrative case of how vessel wall MRI can be useful in clinical practice. So a common clinical conundrum is differentiating RCVS and vasculitis, because sometimes it can be difficult clinically. So in this case, we had another young woman. She was in her early 50s. She was presenting with a headache. She was found to have small volume, some arachnoid hemorrhage, and some small strokes. So on her time of flight MRI, which is shown here, we can see that she had areas of vessel narrowing and irregularity, including in her ACA branches here. So the differential considerations with her clinical and imaging presentation at this time was RCVS and vasculitis. Atherosclerotic disease was considered less likely for a number of reasons, but specifically her pattern of narrowing. There was this preferential involvement of the distal vessels with relative lack of involvement of her proximal vessels. So because of the clinical uncertainty between RCVS and vasculitis, she went on to have a vessel wall MRI, where we can see on her post-contrast T1, we have an area of circumferential vessel wall enhancement associated with an ACA branch, which was narrowed. So this finding, in addition to the fact that these areas of narrowing persisted on follow-up imaging, was thought to be diagnostic of vasculitis, and she was treated as such. So there are a number of other clinical indications to differentiate between narrowing, like Moyamoya disease versus Moyamoya syndrome, or identifying radiation-induced narrowing. So vessel wall MRI can be an incredibly helpful tool when patients are presenting with narrowing in their intrapreneal vessels with an unclear etiology. Okay, so moving on, the next primary reason to perform vessel wall MRI is to potentially identify areas of symptomatic but not synoptic disease in cases where angiographic imaging is not helpful or doesn't reveal any significant findings. So when we're solely using conventional angiographic techniques like CTAs or MRAs or DSAs, we're really only detecting plaques that are resulting in luminal narrowing, but we're not seeing the vessel wall. But we're in this midst of a paradigm shift when it comes to evaluating stroke risk, and we know that the degree of luminal stenosis does not necessarily equate to stroke risk. There are so many other factors to consider when evaluating stroke risk, in particular, the actual plaque we're imaging. So we're not, with vessel wall imaging, we're not only assessing the degree of narrowing, but also the characteristics of a plaque, which may indicate more details on the level of risk. So in this example, this is another real patient who had had a left MCA stroke. He had a completely normal CTA of the head. There was no narrowing that was detectable on the head CTA. So the patient was thought to be having a cryptogenic stroke or a stroke of unknown etiology. Because of that, he went on to get a vessel wall MRI, and on that exam, shown here, we see an area of eccentric T2 hyperintensity with associated enhancement consistent with an active atherosclerotic plaque thought to be the so-called culprit plaque. So by using vessel wall MRI, the patient went from having a stroke of unknown origin to knowing that he had an active atherosclerotic plaque and thus was able to begin appropriate treatment. Okay, let's just move on to some of the other many reasons to perform vessel wall in clinical practice. So the first two really have probably the most evidence behind them. These other reasons do have some evidence, but it's a little bit more mixed, and so we'll talk about that briefly. So first is just to evaluate atherosclerotic plaque itself to determine its activity and likelihood for future or recurrent stroke. So let's just take a really brief minute to discuss what we're actually imaging when we evaluate atherosclerotic plaque. So many pathologic studies have found the basis for our imaging, and so let's talk about the plaque components. So first, the most common plaque component, or one of the most common plaque components we see, is the lipid core, which is typically T2 hypo-intense and non-enhancing, as we can see by this area of crescentic T2 hypo-intensity and non-enhancement here. So another plaque component that we can image at times is interplaque hemorrhage. So this is another finding that's indicative of higher risk and can be seen as an area of T1 hyper-intensity. Overlying the main plaque components, we usually have a tough fibromuscular layer called the fibrous cap. So this fibrous cap separates the internal lipid or hemorrhagic plaque components from the flowing luminal blood and is usually seen as an area of T2 hyper-intensity. So the cap can enhance, especially in active or symptomatic plaques. Also, if we can see any irregularities in the fibrous cap, that can expose the plaque itself to the flowing luminal blood, increasing stroke risk. So another finding that's often associated with atherosclerotic plaque on vessel wall MRI is its location. So usually they're eccentrically positioned with respect to the lumen. And beyond that, beyond these findings, we can also see a peripheral thin rim of enhancement surrounding the plaque due to increased vasovasorum in patients who have active or culprit plaques. So these images showing these examples are really just the ideal beautiful appearance. And many times our imaging is not quite as clear as these. We often kind of see this eccentric clump of enhancement, but it is something that this is the basis for what we're striving to look at. There's a few other points on atherosclerotic plaque. So plaque enhancement is strongly correlated with stroke, regardless of the degree of enhancement. So they're thought to enhance because of neovascularization and inflammation. And usually what the enhancement that we're seeing, as we mentioned, is kind of the fibrous cap and the peripheral wall of the vasovasorum. In clinical practice, it's kind of hard to separate these two components, but we can identify areas of enhancement by looking at enhancement similar to the degree of the pituitary inflandibulum as kind of an internal control. Another finding in intracranial atherosclerosis is a change to the vessel wall in response to plaque formation called remodeling. So positive remodeling is the outward bulging of the outer surface of the artery. And this is usually occurring as an adaptive response to the presence of a plaque in order to preserve blood flow. And so it explains why we can see these, you know, very large atherosclerotic plaques without significant narrowing. So, you know, we can see patients who have normal angiographic imaging but have very large plaques that could be a source of stroke. So positive remodeling is separately strongly associated with ischemic stroke. Okay, let's talk about another indication to use intracranial vessel wall MRI. So we kind of talked previously about using it as a means for differentiating vasculitis from other types of vessel narrowing, but some people can use it also to evaluate the activity of vasculitis. So similar to active atherosclerosis, we frequently see enhancement in active vasculitis. Also similarly, we can use the pituitary infundibulum as an internal control. So here's an example of this one of many patients we have who's followed over time to evaluate the degree of enhancement as a means of assessing disease activity. So just as a note of caution that the degree of enhancement does not always correlate with to disease activity. So while it is nice to be able to image vessel wall vasculitis, it can be considered kind of like an adjunct to other surveillance techniques in this setting. Okay, and another area of great research, of active research right now in vessel wall imaging is evaluating the stability of aneurysms and also identifying which aneurysms have ruptured recently in cases where patients have multiple aneurysms. So enhancement in an aneurysm wall is thought to reflect inflammation and there are several studies that have shown that enhancement is more commonly seen in growing or recently ruptured aneurysms. In fact, it's thought that more than 95 percent of recently ruptured aneurysms demonstrate enhancement. So it's really strongly correlated with instability and rupture prone aneurysms. So it's an area that is currently being used in cases where there's multiple aneurysms and in fact we often use it for aneurysm surveillance as an adjunct to standard MRA techniques to see if there's any new areas of enhancement. This is an area, again, that has ongoing research and we'll get more updates on this as the years go on. Okay, lastly, just as a quick note, vessel wall MRI is also helpful in identifying areas that could be amenable to biopsy in patients who have suspected CNS vasculitis. Just to find an area that is more amenable to targeted to biopsy is helpful to reduce morbidity associated with brain biopsy in these cases. Okay, so now that we've talked about kind of what the basics of what intracranial vessel wall is and just a little bit about why we do it, let's talk about how to do it. So there are really four main components for successful vessel wall imaging. So first is suppressing the luminal blood flow in CSF. Next is having high spatial resolution. Next is using multi-planar acquisitions. And lastly, using multiple tissue weightings. So let's first talk about the suppression of luminal blood flow in CSF. So this is a critical aspect of vessel wall imaging because the whole idea is to make the wall pathology more conspicuous and to maximize the contrast noise ratio. So interestingly though, it's one of the most critical aspects of imaging intracranial vasculature. It's an area where there may be the most variability and least standardization of protocols among institutions. So different techniques are used depending on if 2D or 3D sequences are being acquired. Most institutions are primarily acquiring 3D sequences just in order to increase coverage of the intracranial vessels. So common methods for suppression of the blood flow in CSF include variable refocusing clip angle sequences with T1 or PD sequences. So like VISTA on Philips, SPACE on Siemens, or CUBE on GE. There are a number of other techniques. In fact, we use Dante or delayed alternating mutation for tailored excitation, which is what is shown here. And I think that is becoming increasingly common. The important thing here is finding a technique that it can be done. There are a number of different techniques across different vendors, but the critical component really is suppressing blood flow and CSF in order to maximize the conspicuity of the vessel wall pathology. Okay, next, spatial resolution. So really having adequate spatial resolution is fundamental to accurate imaging of the intracranial vasculature. So most institutions, I would say nearly all institutions are doing at least a 3T MRI in order to maximize the signal-to-noise ratio. So vessel wall sizes based on pathologic studies are around 0.2 to 0.3 millimeters. Realistically, not many people are imaging to that degree of size. Most people are doing around 0.4 to 0.9 millimeters. And kind of 0.5 millimeters is kind of like a target voxel size because it kind of balances acquisition time with image quality and cost, but still providing excellent diagnostic confidence. Using 7T MRIs in an area of active research, it has really improved signal-to-noise and contrast-to-noise ratios compared to 3T, but still maintains a reasonable signal-to-noise ratio. But still maintains a reasonable scan time. There are of course some limitations to 7T imaging, but I think as more and more institutions are using 7T, we're going to be seeing it more commonly in the clinical realm. Okay, so determining the pattern of enhancement of a vessel wall is critical in identifying its etiology. So just as an example, we can see two different patterns of enhancement on top. We see an area of concentric enhancement around the vessel wall, which is a pattern that is commonly seen in vasculitis. On the bottom, we see an area of eccentric enhancement here, which is a pattern that is commonly seen with atherosclerosis. So being able to see vessels on FOS is critical to evaluate patterns of enhancement. So really multi-planar acquisitions are really useful, just because of the normal tortuosity of the intracranial vessels. So most institutions are requiring 3D isotropic sequences, which allow for reformats and any obliquity and kind of provides an overall reduction in scan time because you're able to cover a large area. Some people do use targeted 2D sequences if there is an area of particular concern, and that's okay, but I think most commonly people are doing 3D acquisitions. So another, the last thing I wanted to briefly talk about was multiple tissue weightings to evaluate the composition of the vessel wall plaque. So different pathologies have different compositions, and these are just the basic sequences that are required for a vessel wall. So a time-of-flight MRA, which is kind of the basic roadmap for all the vessels and obviously can help evaluate and flow dynamics. Some also are performing a post-contrast time-of-flight as well. And then the kind of workhorse for vessel wall imaging, as we mentioned a couple of times, are the pre- and post-contrast T1 or PD weighted sequences. And again, these require blood flow and CSF suppression, as mentioned earlier. Many also include fat suppression in these sequences, just to kind of evaluate the small ECA branches, which may enhance in cases of temporal arthritis, for example. So, and T2 is another sequence that's very helpful for discriminating between various disease pathologies. It's also very commonly performed. It's helpful because T2 hyperintensity is often seen with atherosclerotic plaque, so it can be helpful for discriminating atherosclerosis from other types of vessel narrowing, such as vasculitis. And there's a couple of small notes on vessel wall imaging on the time delay specifically for vessel wall imaging. So most people delay, have like a five to seven minute delay from contrast injection to scanning the post-contrast T1 sequence. We do, and I think many others do, another sequence in between. Sometimes we do a T2 flare just to kind of space it out and maximize the contrast visibility on the post-contrast sequence. And then depending on where you are with incorporating vessel wall imaging into your practice, some people choose to monitor every exam, especially if there's a very small specific area and a region of interest, then it's helpful to have a radiologist kind of monitoring each step along the way. We kind of do a pseudo-monitored versus unmonitored where, you know, we're doing 3D sequences, so we're able to acquire most of the intracranial vasculature, but we usually do a check before the checks are completed. And I think this is something that's going to be institution specific. So I hope we were able to review some of the basics of intracranial vessel wall imaging so you can incorporate this into your practice as well. And I thank you so much for your attention and I look forward to answering any questions at the conclusion of the session.

atherosclerosis imaging

pathology

technological advances

systemic issue

angiography

photon counting CT

AI in imaging

radiomics analysis

risk stratification

imaging biomarkers