 This is the second part of imaging concepts and recent advances in imaging of stroke. In the first part, we covered the basics of stroke, which included the causes of stroke like subarachnoid hemorrhage, parancamal hemorrhage and what we are concentrating on today. That is 80 percent, which is formed by ischemic stroke. We know what is stroke definition. Stroke definition is reduced burqlo and perfusion caused by thrombus, embolus or hemorrhage. There are three zones within the area of stroke which are identified ischemic core, where the flow of blood is less than 12 ml or since 100 grams of brain per minute ischemic penumbra, where the flow is between 12 and 20 ml of blood and oligemic tissue, where the flow is more than 20 ml of blood. We also understood the saying which talks about what you want is what you get and how the stroke imaging has evolved over the last 15 or 20 years. In the last talk, we covered three questions about stroke on imaging. The first question being, is there an infarct? Second question, what is the cause of infarct? And the third question, is it reversible or not? Today we will cover the other five questions. One, what kind of intervention is possible in a patient of stroke if he or she reaches the stroke hospital within the given window period? When it can be done, what are the outcomes of treatments? Can we predict stroke before the complications happen? What are the complications and how we can detect upcoming stroke before it happens? So, let's start with today's first question. That is, what are the kinds of interventions available today in patients who are detected to have hyperacute or acute stroke? Again, they are the same. Family intervention, the concept is moving from just time is brain to physiology is brain and we'll understand that in next five or six slides. The type of intervention that are available today, if the patient comes within three to six hours of getting stroke, intravenous and intraartril thrombolosis. Within three hours, intravenous can be good enough. Within six hours, three to six hours, the intraartril is what is followed. Within intraartril thrombolosis, we have intraartril drug injection and mechanical thrombolosis or thrombic tummy. This is one example of mechanical thrombic tummy. This patient came to us with a stroke duration of one and a half hours. We showed hyperacute, a right MCA stroke on MRI. On BSA, there is a large acute thrombus in the M1 segment of right MCA. This is pre-procedure and this is post-procedure DSA showing complete opening up of entire right MCA and this is the thrombus which are located in the proximal M1 segment of right MCA. Next question that we as a radiologist need to answer is when these interventions can be done. The standard window period of zero to three hours and three to six hours is rapidly changing with better thrombolating drugs available and better understanding of stroke which we have acquired over the last 10 to 15 years. So, window period of three hours is getting extended and window period beyond three hours is also getting extended from three hours to four and a half hours and six hours and 12 hours in different locations and different sizes of stroke. What we also need to predict on imaging is what is likely to be the outcome of thrombectomy or thrombolysis if that is performed. That can be predicted by location of thrombus and also some certain imaging features which we will understand in next four or five slides. So, we know from history and predictions that mortality varies in the given patient by thrombus location. For example, if there is a large thrombus in basilar artery the mortality is almost 90 percent. If it is in terminal ICA then the mortality is about 50 percent. If it is in MCA the mortality is about 30 percent. Also, the outcome depends on types of stroke whether it is mild stroke, major stroke or stroke caused by acute lack of infarct. So, in mild CV when there is no large vessel occlusion and infarcts are small you have good outcome. If it is major LVO large vessel occlusion and infarct is large it has poor outcome whereas acute lack of infarct has got best outcome. Then there is concept of aspects or Alberta stroke program early CT score which divides MCA into 10 zones. Every area that is covered by stroke one point is deducted. So, if we have hypo intensity in two zones your score will be 10 minus 2 that is 8. If it is located in six zones your score would be 4 and based on the entire outcome changes. We will understand that in next slides. Also, because this aspect score became very popular in the supra-tentorial compartment aspects has come out with same type of scoring in post-sufficiency aspect. So, where if the hypodensity is present in thalamus, occipital load, midbrain, pons or cerebellum certain points are deducted and whether it is supra-tentorial compartment or infratental compartment a score of less than 7 has got bad prognosis score of more than 7 has got good prognosis. There are other several clinical variables which impact the eventual outcome of a patient with stroke and these are time-to-treatment, post-trombolosis, hemorrhage whether that happened or no on imaging, severity of stroke, age of the patient, a race of the patient, weight of the patient and BP and several other comorbidities. So, that is what the outcome depends on. The next question that we need to answer when we are imaging these patients is can we predict post-trombolosis hemorrhage? If we can predict that we can avoid several complications which happen with aggressive thrombolotic therapy. So, post-trombolosis hemorrhage is divided by ECAS classification into four types. The simple ones are smaller hemorrhages which are divided to HI1 and HI2. HI1 or hemorrhagic infarction type 1 is when you get particular hemorrhages at the margins of the infarct like this borrowed patient. Hemorrhage infarction type 2 or HI2 is when the particular hemorrhages within infarct happen throughout the infarct. So, these are minor hemorrhages and then there are two subtypes of parent camel hematomas. PH1 is when less than 30 percent of the infarcted area bleeds with minor mass effect whereas PH2 or parent camel hematoma type 2 is where more than 30 percent of the infarcted area shows hemorrhage and there is major mass effect. Outcoming these patients is relatively poor compared to HI1 and HI2. There are several imaging productors which also show the eventual outcome of the patient. For example, this particular patient who has got massive left emsaturatory infarct on diffusion because it is covering almost 100 cc's of brain parent karma, chance of this bleeding post thrombolysis is almost 95 percent. On your right hand side, you have patient with massive right ICA territory infarct. The hypodensity is obviously covering more than one third of the MCA territory and chances of this patient undergoing hemorrhage post thrombolysis is also very, very high. Second predictor is the location of area of infarction. Cream matter infarction has relatively higher risk of hemorrhage compared to infarct in the white matter area which has relatively low risk. That is because there are several collatials in the cream matter and it is highly vascular. We saw an aspect score, an aspect ratio of less than 7 or equal to 7 has got a bad prognosis and chances of these patients bleeding are also very high. The third predictor is based on CT and MRI perfusion. So in the matched area of infarction or infarct core where there is irreversible brain damage, chances of hemorrhage are quite high and this is an actual example. If the location is beyond the infarct core, chances of hemorrhage are lesser. The fourth predictor is based on angiography. So if there is piled collateral formation on the DSA done before thrombolysis, if there is good collateral formation, chances of hemorrhage are 2.7 percent or lesser. If there is poor collateral formation, the chances are 25 percent and more. And here are two borrowed examples. This particular patient has got terminal left ICA occlusion with a massive infarct. Host thrombectomy, the ICA has opened up. Both ACA and MCAs are feeling well. But because this particular patient has poor collateralization, this patient bled in the left basal ganglia with extension of the hemorrhage in the subepnotic space and also in the ventricles. Compare that with this particular patient who has a right M1 segment thrombus with good collateralization on delayed DSA images. This patient underwent thrombectomy and host thrombectomy Cp does not show any evidence of hemorrhage. So good collateral formation on pre-thrombectomy DSA would have good outcome, whereas that with poor collateralization will have poorer outcome. The next question that we need to answer as a neuro radiologist is, can we detect upcoming strokes in cases of prior stroke or TIS before it actually happens. The basis of this is NASA study, which was a multi-centric, multi-country study performed in 1986, which talked about efficacy of carotid endotryctomy in patients which showed stenosis of between 70 and 99%. What the study said is if a patient has carotid atrial stenosis of between 70 to 99% and if you do carotid endotryctomy, these patients perform much better than those put on anti platelet therapy. Over the last 10 to 15 years, lot of researchers realized that this prediction was not as accurate as it was thought in 80s and 90s. What we now know is stroke genesis is an atheroembolic phenomena and not really related to low flow in carotid artery contrary to our understanding in coronary arteries. And lot of researchers presented their own work which said that almost 75% or more of stroke events happen in patients who have moderately stenosed vessels, lot of them showing less than 50% stenosis. So now we know that stroke genesis is an atheroembolic phenomena and not really a low flow phenomenon which is present in coronary vessels. Hence the new terminology for plaque came in. So plaques which are likely to cause stroke are considered as vulnerable plaques and the plaques which actually give rise to stroke are called as culprit. HRMR and Doppler are known to detect these culprit plaques before they strike. For this we must understand the plaque classification which was put forth by American Heart Association and how we can show that on Doppler and on MRI geography. So American Heart Association has divided plaques into eight subtypes. What we should be aware of are type four, five and six plaques which contain lipid-rich necrotic core in the center with fibrous cap or plaques which contain hemorrhage within the core of the plaque because these are unstable plaques and likely to embolize in the brain compared to type seven and eight plaques which contain calcification and fibrous tissue. HRMR has excellent soft tissue contrast, it shows accurate in view depiction of plaque and it also quantifies the narrowing. So there's a new concept of plaque load which looks at the content of plaque and the size or volume of the plaque and HRMR can do that very well and here is an actual example. The MR brain images of this particular patient I showed in part one of this talk where this particular patient had a huge right MC stroke. If you looked at this high-resolution plaque imaging which are done subsequently you can see a hemorrhaging plaque which is right on T1 and fat set T1 both and part of the plaque has ulcerated, the cap is denuded and it is probably embolized creating a big crater over here. So we know we're dealing with an unstable hemorrhage type six plaque which again is likely to embolize and cause another stroke. So this particular patient although the narrowing is close to about 50-60% is the candidate for carotid endodontomy or stenting. Coming to what are the current status of treatment and imaging of hyperactive stroke. So most of the researchers are now moving from wall clock-like situation to tissue clock-like situation because we have to understand the pathophysiology and not only the anatomy of stroke. So the idea is to identify group of acute stroke patients who could benefit from thrombolytic therapy, do a normal plane CT, look for hypertensive C-sine with involvement of less than one-third of MC. These are right candidates in the given time for either thrombotomy or intravenous thrombolysis. If you're doing MR, look for microhemorrhages, leptomanager enhancement which are relative contraindications for active treatment of hyperactive stroke. The issue that is always raised is whether CT is better or MR is better for hyperactive stroke. I'm going to answer that little later but there are several proponents of MR like me who talk about how MR can be done on good machine within a matter of 8 to 12 minutes getting much more information than what a CT, which CT perfusion can give. So a lot of researchers say that if you're going to give intravenous thrombolysis or breech therapy just do a quick T2 diffusion and circle of wills angio which can happen in 6 to 7 minutes, give half the dose of thrombolysis in the MR suite itself and then complete the rest of the study which will give you complete information. What are the current ACR guidelines for imaging hyperactive stroke? This is a relatively busy slide which is kind of simplified in the next slide and if you look at the ACR guidelines of 2016, whether you're dealing with patient of PIA, hyperacute, acute or early subacute stroke, all four that is MRI, MRA, CT and CT angio get marks of 8 on 8. So that means all are the most appropriate investigations in patients with hyperacute, acute and early subacute stroke. So it really depends on how you have prepared yourself to treat and image these patients with hyperacute or acute stroke. What we as radarists must know is time is brain and in the typical patient with large vessel occlusion, almost 1.9 millions of neurons get damaged every minute the patient's treatment goes delayed. So the onus is on us to make sure that we do not waste any time and a good collaboration of teamwork between emergency department, neurologist and radiologist is the way forward to minimize the brain damage and give earliest possible treatment to appropriate candidates. Let's now move on to a few advances and what is going to happen in the future for stroke imaging. There are several new modulators which are coming in in MRI itself which hold promise of further defining patient classification to improvise patient management and hence the outcomes. There are four relatively new things which will completely change the way we image patients with stroke. Two of these we are currently using two of these we may use in future much more. So let's look at all these four options which will change the way we may stroke. Let's start with ASL imaging or arterial spin leg link in stroke. Currently it is used for evaluation of penumeral zone and look at diffusion perfidumous match. Also it can localize the intra arterial acryl thrombus. It is very very useful in demonstrating post therapeutic hyperperfusion and reversibility of ischemic core and penumbra and also it is very very useful to exclude stroke mimics typically postepial state. And here are two examples on your left hand side you have diffuser abnormality and perfuser abnormality almost same. So outcome of this patient is not going to be as good as the patient on your right side where diffusion abnormality is small and perfuser abnormality is really large. So that is the usefulness of ASL. Here is another patient where you have thrombus in right MCA, acryl thrombus in right MCA which is very very demonstrated on ASL imaging. So bright vessel sign is one thing that we really do not look for but if you look for it you will see it very very often. Other area where ASL is very useful is post therapeutic hyperperfusion and here is our patient with left MCA stroke which I showed in last lecture and here is an MRI done after 24 hours of thrombectomy. The diffuser abnormality is significantly reduced. The perfuser abnormality you see a lot of revascularization in the posterior half of this left MCA stroke. Next new advance in stroke imaging is swan imaging or susceptibility weighted imaging. This was first described by hackay in 2004. It is now an essential part of any hyperact stroke protocol. It is very very useful to differentiate hyperperfusion from direct perfusion in patients where diffuser abnormality is suspect. It can also detect thrombus like ASL. It shows better depiction of hemorrhagic transformation than any other sequence and there are two signs which are described. I am going to show you one example of each of these. One is susceptibility vessel sign and two is prominent vessel sign. susceptibility vessel sign is useful to show hyperacted thrombus. So this thrombus which is paramagnetic produces blooming artifact and here is an example of the same. When this hyperacted thrombus is present at the site of thrombus diameter of the vessel because of blooming appears larger than the diameter of contralateral artery. So this is one way of making sure that you are dealing with a hyperacted thrombus. Second thing that we look for is prominent vessel sign. That happens due to increased deoxygenated blood in the draining veins which appear prominent and hyper intense in the infected area and here is a borrowed example of the same. This is 63 year old woman who had diagnosis of left emsatoratory infarct. The area of infarct shows multiple prominent veins which are showing susceptibility effect which is excellently correlating with the diffusion abnormality and CB way abnormality seen on the top of these four images. Next new important sequence which is coming up is diffusion kurtises imaging. What are the difference between normal diffusion that we look at versus DKI or diffusion kurtises image? The standard diffusion measures Gaussian distribution of water diffusion in biological systems whereas DKI is based on non-Gaussian distribution and this effect is taken into account when you are looking at diffusion abnormality. All of us now know that diffusion abnormality does not show the true core because there are several studies which show that subsequent scans show decrease in diffusion abnormality. So DKI is supposed to be changing the way we look at diffusion. DKI will actually show the core and not what the diffusion abnormality shows and here are two borrowed examples of the same. Here is the patient with a small M1 territory left emsatoratory infarct. This is the diffusion abnormality. This is the kurtises abnormality and this is the follow-up scan showing gliosis. All three are matching. Problem is when these do not match and here is another borrowed example of left emsatority and posterior watershed hyperacute infarct. This is the diffusion abnormality. This is the kurtises abnormality and this is the follow-up scan showing gliosis on T2. If you look at these three images, gliosis is absolutely correlating with kurtises abnormality and not so much with diffusion abnormality. So we now know that DKI is more accurate in detecting core of the infarct and not so much diffusion abnormality. Next, we look at PHMRI which is useful in hyperacrystalt imaging. The disturbing metabolic activity that is associated with hyperacrystalt infarct correlates very well with PHMRI because PH drop is supposed to precede drop in CBF. Currently we look at drop in CBF when we will be able to look at drop in PH that will precede the drop in CBF. We can detect hyperacrystalt using PHMRI much earlier. With DKI and PHMRI coming in, we will have to probably revisit the mismatch model. So two abnormalities, diffusion and perfusion abnormality zone 1 and zone 2 that we are talking about will then be divided into four sub zones. The diffusion lesion will be divided into kurtises lesion and diffusion kurtises mismatch lesion. The perfusion diffusion mismatch that is zone 2 would be further divided into ph diffusion mismatch and perfusion ph mismatch and will then have better destination of ischemic core, ischemic penumbra and benign oligemia. So to conclude, imaging plays a critical role in store management. Diffusion is the stroke sequence. With ongoing advances in imaging protocols, we will get imaging which is more robust and more rapid. Penumbra mapping which is currently a problem and it should be a part of imaging protocol would get more refined with PH imaging and kurtises imaging coming in future. What we must remember is as a radiologist, we must insist on getting these patients imaged as quickly as possible and never forget time is brain. Thank you for your attention.