 Up to the last session by Dr. Zishan Sheikh, who's a consultant pediatric radiologist at Birmingham Women's and Children's Hospital, NHS Foundation Trust UK. He's going to talk on pediatric marrow imaging goal of MR, which I think is a very important topic and something which is very confusing always. After this, we have a very interesting quiz session, which I would want all of you to wait for. Hi everyone, my name is Dr. Zishan Sheikh and I'm a pediatric radiologist at the Children's Hospital in Birmingham. I'm going to be speaking to you about the MR imaging of pediatric bone marrow. Looking at bone marrow on MRI is something that we do quite frequently, both in children who present with a limb or who are known to have pathology and need for their evaluation. I'll try and break that down during my talk just to let you know the kind of work we do at the Children's Hospital. We're a standalone pediatric hospital, so we only see children and we're a tertiary centre for both pediatric oncology and more specifically for pediatric bone tumours. But we also provide services for pediatric rheumatology and non-oncology orthopedics. And I'll just show you some examples of cases relevant to all of those with regards to bone marrow. So the aims of my talk are going to be to help get an understanding of the MR sequences we use in bone marrow imaging, to describe normal MRI images in appearances in marrow. For children, that is something that many MSK radiologists can struggle with. And also just to highlight the diagnostic utility of MRI in various marrow-based pathologies. So why do we image marrow? Trabecular bone does not generate MRI signal, but bone marrow does. And so sometimes we're imaging marrow as a surrogate for looking for pathology that may involve the trabecular bone. And malignant marrow disease is something that often a question in children that can present in a non-specific way. So that's another reason that when you're imaging a limb, particularly if you go on to MRI, you want to look at the marrow signal. And bone infection and injury can actually be a cult on other modalities. And again, that's another reason you like to resort to looking at the marrow. This is a table which looks very complex, but really what the message of this is is that MRI ticks almost every box that you want for looking at marrow for both the osseous abnormalities and sometimes in some pathologies the extra osseous abnormalities you get. It gives you good anatomical detail. It's possible to acquire whole body imaging. It's just not always accessible. And I guess in some contexts the accessibility is availability and children it can mean using an anesthetic or sedation. And obviously there's a cost prohibition as well in comparison to simpler things like X-ray. So what's a radiologist approach to looking at marrow? It's very similar to lots of other things. What is the clinical context? Where are the marrow signal changes? How are they distributed? And are there any extra osseous findings as well? That will all help narrow down what's going on. This is a very good article that gives a very nice baseline to looking at marrow in children on MRI, published in Radiographics in 2016, and I'll be referring to it in parts during my talk. So we're just going to start off, what is the basic imaging approach to imaging bone marrow? So bone marrow is structurally made of hemopartic cells, but in addition to the cellular material it also contains fat. Again, this fat is stored in adipocytes, and it lies in the medullary cavity rather than that's interspersed around the trabecular bone. It's got various roles. If you're making a rethrocyte, it's supporting oxygen transport, it producing leukocytes, it plays a role in immune defense, and obviously we're making platelets as well so that has a role in hemostasis. So those are all the different components your marrow is involved in producing. Now, when we come to looking at MRI imaging of marrow, the old wisdom that I was taught before I came into radiology was T1 sequences, fat is hyper intense, and that's for anatomy, and T2 sequences, fluid is hyper intense, and that's for pathology. And when you've been looking at bone marrow, you realize that that's not all quite true. T1 is not just for anatomy, particularly with regards to marrow. Because marrow has fat, T1 sequences are very equal for looking at marrow. And so what are the standard sequences if you're suspicious of marrow based pathology, whether it's a stress fracture or it's infection, you want to do some T1 spin echo sequences. And you'll also want some fluid sensitive sequences. So fluid sensitive sequences can either be PD or T2 weighted, but they will have some form of fat suppression. So they'll eliminate the fat to give you nice contrast to see where there's a change in the distribution of water protons. There are also advanced sequences you can do, chemical shift imaging. I'll show some examples. We may do post contrast, depending on what the indication is. And diffusion imaging can actually be very useful both in the context of oncology and infection. So this is a nice T1 spin echo sagittal of the knee of a two year old child. You can see that the ossification centers are much brighter than the diaphysis of the femur and tibia. And that's because they've got yellow marrow. So they've got fatty marrow preferentially quicker than the long bones have. So we can see that the fat is hyper intense. There is actually a smaller fusion at the back of the joint, which is hyper intense. So you're not able to quite pick it out as easily. But there are other things that we know that can be T1 hyper intense as well. Hemorrhage contrast, some mentioned on the previous slide. Fluid sensitive sequences. So this is a PD weighted sequence on that image that you've got on the right. Fluid is hyper intense. So now you can see that area at the back is hyper intense in contrast to the previous image. But we want to fat suppress this. And that really helps us pick out that joint diffusion. And if you notice, there are actually tiny little loose bodies in that. So this was thought to be a case of a synovial osteocondromatosis, the cause of which we never got to the bottom of. So eliminating the fat, if there was any abnormality in signal in the marrow, that would jump out on a sequence like this. This is a post contrast image. So you'll notice that there's synovial enhancement along that pocket of fluid. So gadolinium, we all know is T1 bright. And in this case, we're not actually looking. We've kind of excluded marrow involvement. There's no abnormal enhancement of the marrow. There was no abnormal signal. This is all joint oncology. So this is another example of a T1 post contrast image. And I think this child is between eight to 10 years old has come with a hand lump. You can see that there's increased soft tissue swelling over the middle finger approximately. And these are both T1 post contrast images, but the image on the right is fat saturated. And that actually lets us see that there's this homogenously enhancing component in the soft tissues of the proximal phalanx, not just some signal change in enhancement in the phalanx itself. And this turned out to be a case of Ewing sarcoma. So fat suppression, again, this is very hopefully very basic for most of the listeners. It really helps pick out where the abnormality is. It's quite acceptable to do pre and post contrast that does not have fat saturation. But what I would just warn people not to do is to not fat saturate your pre contrast T1s. And then to fat saturate the post, it's sometimes quite difficult to compare. You have to be very confident that you know what you're looking at. So chemical shift imaging exploits the different phases between water and fat. And it consists of doing an in phase and out of phase. And at the minute I'm talking about T1 weighted chemical shift imaging, but you can do T2 weighted chemical shift imaging as well. So in the in phase, both water and fat adds to the signal in the image. In the out of phase, water and fat signal, they're not in phase together. And so the fat signal is canceled out and the intracellular fat loses signal intensity. We can then use those appearances on those two phases to create a set of images. So these are the simple in and out of phase images. You'll be used to doing these in things like looking at adrenal lesions to see if they've got frank fat. This patient's got an abnormality in their left proximal, sorry, their right proximal femur. You can notice that there is some loss of fat signal, but it isn't a lot. You can ignore my errors on the previous image. This is the same patient. And so this is a Dixon technique of fat suppression. This is a form of chemical shift imaging. We've got T1 in phase where both fat and water are adding to the signal. So everything that is bright could be a mixture of fat and water. Then we've got the out of phase and the MRI machine can do a calculation to remove all of the fat signal that it knows has been suppressed on the out of phase imaging and give you a water only image. So this is the fat you can see in this image in the bottom left is all suppressed. So the subcutaneous fat in the mid thighs that's suppressed. And there's also a fat image, fat on the image, so that only the fat signal shines through. And again, if you've got lesions that in the marrow, for example, the femoral head has nice normal fat, so it's nice and bright, but there's a lesion that's quite dark in the right proximal femoral neck. Unfortunately, you've got red marrow on the other side, so the difference is not so stark. But this patient's actually got a little lesion in their femoral neck, which turned out to be metastasis. So if you don't concentrate on the pathology necessarily in the previous image, but just remember the principle, diffusion weighted image, we're hopefully familiar with it in the brain, but it's quite useful at looking at marrow depending on what you're looking for. It's measuring the free motion of water molecules and we see restricted diffusion in a range of pathology. And it's the same in marrow as it is for the rest of the body. If you've got something cellular like a primary bone sarcoma, if you've got pus or exudate, because of the complication of infection, blood products can sometimes cause a degree of restricted diffusion, but we don't usually see frank blood products unless it's in a tumour usually in the marrow. Red marrow in an infant can also restrict diffusion occasionally, so that's the only pitfall. So we're just going to talk about what the normal appearances of marrow are on those various sequences that we've discussed. So normal bone marrow is comprised of red marrow, yellow marrow, and trabecular or cortical bone. And the trabecular and cortical bone acts as the scaffolding for the marrow, although obviously they have another function as well. For our purposes today, the red marrow is contained, comprised of more cellular material, and it's more active metabolically. The yellow marrow is relatively inactive and comprises more fat, and that's why we call it yellow marrow. And you can see a little bit of both in this cut-through image of the femoral head. The trabecular bone is very difficult to see on MRI, that's the purpose of that image. You can just about make some kind of bony architecture out in the epithesis there, but that is something difficult and it's not the purpose of the talk right now. So this table may also look complicated, but really what I'm trying to tell you is the differences between yellow marrow and red marrow in the first instance. Again, yellow marrow is comprised approximately of 80% of fat, and so the percentage also increases slightly with age. We're talking about children, so we won't need to really go into that in detail. Red marrow also has fat, but much less fat, and for example, even though it says 40% on the screen there on that table, it can be as little as 20% if you're talking about an infant who has a lot of cellular material in their red marrow. And so naturally red marrow has more protein because it has more cellular material, and yellow marrow has less. Red marrow can be vascular, but yellow marrow is relatively a vascular, and that means when we give contrast, red marrow can enhance. Whereas yellow marrow, it doesn't really have as much or noticeable of an enhancement. Pathology obviously can contain cells. It can also disrupt the distribution of water molecules. There's really no pathology in the marrow that specifically incorporates fat, and that's in contrast to the soft tissues where soft tissue tumors or benign soft tissue pathologies can incorporate fat sometimes. That's not really the case with marrow, and obviously we know that pathologies either you can get blood, pus and cells as well as the edema as I've talked about the destruction of water, and you can get enhancement sometimes in various marrow pathologies. So you'll notice there are some overlaps basically between red marrow and pathology, and if you were struggling to see that there was a lesion in the proximal femur on that one image I showed you in the Dixon slices, I wouldn't blame you, and so this is often a struggle. Is it red marrow, or is it pathology? And I'll hopefully help you sort that. It's more easy to tell if there's red marrow or yellow marrow. So what are the different appearances on some of the sequences I've talked about? So I put a stir there in place of fluid sensitive, but you could replace stir with PD fat sat, or T2 fat sat, any fluid sensitive imaging. So yellow marrow has high T1 signal because it's got a lot of fat, and red marrow is either isotense or it can be hyper intense. Pathology is often low on T1, and again we're talking about low in comparison to a disc in the spine, an adjacent disc or an adjacent muscle in the periphery. The stir signal in marrow that has fat will suppress, red marrow doesn't often suppress completely, and pathology always we know can have quite hyper intense appearances, although sclerosis can sometimes look quite low and be a pitfall in MRI. We've talked about the overlap already between the two different red marrow and pathology both enhancing. I'll come back to chemical shift imaging later. Red marrow can really only restrict an infant, so if something is restricting you should be thinking about pathology outside of the infant age group. Pathology shouldn't suppress signal on chemical shift imaging, again because it doesn't incorporate fat. That's just to reiterate that. So the T1 sequences are the most useful. So if you're struggling with marrow on the fluid sensitive sequences there's a bright spot somewhere and you want to solve that. The T1 sequences are going to get you out of jail. It provides optimal contrast for marrow because both red and yellow marrow have fat, so they'll have some degree of hyper intensity. And as I've already alluded to, that's compared to an adjacent muscle or intervertebral disc. It is slightly different in younger children and we will get to that when we talk about them. So just what does red marrow look like that you can differentiate it from pathology? Well, one red marrow has a classic location, so you'll often see in adults persisting red marrow in the metaphysis of the long bones, particularly of the legs or the femoral or the humerus. And you'll see that they often have feathery margins. They can often be bilateral and they don't cause any mass effect. There's no other surrounding edema. Red marrow can also be endosteal. That means it can extend along the diaphysis of the long bone, just hugging the cortex. Whereas pathology, it's T1 hypo intense to muscles, so that means that the marrow fat has been either replaced or eaten away. It's margins are not particularly feathery. They can be very varied and it can be bilateral as well. I guess it's pitfall, but frequently we're looking for something really lateral. And it has mass effect. So these are just examples of what I've been talking about. So on the first image on the left, you can see there's an up and down to this area of, it's still T1 hyper intense to muscle, but it's less T1 hyper intense the area with a star in it than the remaining femur. So this has feathery margins. So this is red marrow. Red marrow can also be in the epiphysis in the subchondral location. And the image on the right is the endosteal red marrow. So again, it's less bright than the more central fatty yellow marrow. The normal progression of conversion from red marrow to yellow marrow has a predictable order. It doesn't have a definite age, so unlike things like myelination in the brain for children, there's no standard that you must have converted X marrow and X bone by a certain age. It varies and it's linked to nutrition and other variants. But I can give you some rough guidance that the main principle is that in an infant, almost the entire skeleton axial and appendicular has red marrow apart from the tips of the terminal phalanges. They have fatty marrow at birth and eventually you're going to get to an adult pattern where the red marrow is essentially located in the axial skeleton. But again, red and yellow marrow, they're not mutually exclusive. What we're talking about is a predominance of red versus yellow marrow, as seen in those images. So how does age-related marrow conversion occur? Again, it's reiterated that the age exactly is variable, but it occurs centrally to peripherally. That means it happens in the middle of a bone and then spreads from the center to the proximal and distal bits, if you're talking about a long bone, and it happens in the limbs before it does in the spine and skull. So this is a nice diagram just to illustrate that. So you've got red marrow in this femur and your yellow marrow conversion is going to start in the center. And your proximal epiphysis will finish before your distal epiphysis of that bone. And you can see the normal patches of remnant red marrow in this femur, which we've left. You can see on the distal femur that sometimes red marrow can just look like little pylons. And in places like the pelvis it can actually have a bullseye appearance, where it sits in a round little bit of a yellow marrow. So when you get reconversion of marrow, you get the opposite order. So if it starts centrally for normal conversion to go peripherally, once you've got the adult pattern of yellow marrow in your long bone, the reverse order is going to happen if you get red marrow reconversion. So the bits that converted to yellow marrow last, they're going to convert back to red marrow first. So that's the metaphysis. So in the spine, this process starts around the base of vertebral veins. So that's this little segment here. So you'll get increased T1 signal here first. That's where the conversion of red to yellow marrow is going to occur. Now, before the age of five years, on T1 images, normal marrow that's red marrow can appear iso intense or hypo intense to the disc. But certainly in children who are five years or older, in most children of that age, the red marrow in the vertebra should be higher in signal than the adjacent disc. Although it says, and this is from the paper that I quoted to you at the start of the talk, the vertebral enhancement, you can usually see it under seven. You can see it in sometimes in older patients as well. But actually, vertebral marrow can look quite heterogeneous even up to middle age, particularly in women. So we talked about this red marrow very briefly in that diagram where there was reconversion back from yellow marrow to red marrow. If you're under stress, you've got decreased oxygen carrying capacity because of chronic anemia or sickle cell disease. Or you've got, and this doesn't really apply to children smoking or chronic lung disease, perhaps cyanotic heart disease. Or if you're an athlete and you have increased oxygen demand, that can cause red marrow hyperplasia. So that means some of your yellow marrow will be converted back to red marrow to meet that systemic demand. This is an example of red marrow conversion in a three year old who's got aplastic anemia. You can see these little islands of low signal red marrow, very similar to that diaphragm we saw. And with treatment, there's been some red marrow conversion as a response. So this is another case of red marrow reconversion. So on the right you've got a normal, and the normal is from a kind of young adult patient. And on the left we've got a 20 year old sickle cell patient and you can just see the difference in T1 signal. In some places in the vertebrae it's lower, it's certainly lower in signal than the adjacent vertical disc. And again, that's because we've got red marrow reconversion. So hopefully we've outlined the appearances of normal marrow. Now we're going to talk about some malignant processes and what the appearances of those can be. So if you have neoplastic replacement of bone marrow, that often affects areas where red marrow is. And that's to do with the blood supply. So we've said that you can get remnant red marrow in the metaphysis and that's something that you're used to seeing in adults. And because a lot of the vascularity in the growing long bones comes through the metaphysis and around the growth plate, that's often where you can get things like metastases. And you also have higher turnover there because that's an area of growth. And so again, a lot of bone tumors, particularly osteosarcoma, they can occur in that region. Neoplastic disease, as we mentioned in the table before, it's T1 hyper intense. Now, it's not always reliably so and so if you compare it to muscle, you're about 81% accurate in detecting and underlying neoplasm. But again, this just shows out shows to you that you will not let that doesn't mean you're going to capture all cases. So not all cases of neoplastic marrow replacement are going to be evident on MRI. And we'll talk about that a little bit later as well. So it's better to compare to the adjacent muscle, even in the spine if possible. It gives you a little bit more accuracy and sensitivity. So there are these different broad categories of marrow infiltration. And so that's focal, patchy and diffuse. And I put up images of each and these are all three different patients. I think they're all young teenagers actually. So they're not too dissimilar perhaps in an age group. And these are all T1 weighted images. So the image on the left is a focal pattern of involvement. And you can see it's multifocal because you've got many nicely well-defined lesions throughout the femur on either side. And there are areas of low signal. They're certainly comparable to muscle, if not lower. On the middle image, which is B, you've got more patchy low signal. So again, we've got these huge areas in the distal metaphysis that are low in signal but ill-defined. The kind of proximal thirds of both femur are quite low in signal. But it doesn't really have feathery margins or anything that we would normally recognize as yellow marrow, as red marrow, sorry. And the last is a diffuse pattern where you've got essentially replacement of all of the marrow. And then the image on the right is actually a case of diffuse neuroblastoma metastases. Now, the main processes that metastasize or involve the marrow, they can all have these appearances, but I've just put listed them under each image under the order in which this is commonest. So if you've got a focal involvement or a multifocal involvement, you're going to be thinking about metastases or lymphoma. Leukemia can also have this appearance, but more commonly leukemia has a patchy appearance or a diffuse appearance. And so actually the image on the right which turned out to be a neuroblastoma, we raised the possibility of leukemia, but on bone marrow we found that they had disseminated neuroblastoma. Now again, neuroblastoma is only one of the causes of diffuse metastases, osteosarcoma, rabdomyosarcoma, Ewing's, they're all there, but neuroblastoma and rabdomyosarcoma are just more likely to give you a diffuse pattern and that's why I've mentioned them, so just for you to have in your differential. And in the case of the neuroblastoma, we ended up doing an ultrasound in a CT to confirm a left adrenal tumor. I'm just going to talk very briefly about each of those. So acute leukemia is the commonest pediatric malignancy and ALL is the commonest subtype. As we've said, it usually gives you a diffuse infiltration pattern. The important thing is that the MRI changes can precede peripheral blood-filled manifestations. So that's if someone takes a peripheral blood test and tests it in the lab, you may actually spot the changes first on MRI. It's really controversial whether there's actually any imaging role in monitoring of leukemia and I think that's really limited to looking at complications like avascular necrosis from treatment. MRI does not tell you that leukemia is getting better, but it can help suggest that it is. Most cases of lymphoma are that involve marrow in children that are non-Hodgkin, so that's 25 to 40 percent. And osseous involvement means you've got stage 4 disease in those cases. You can often have a large soft tissue component, but you don't have to. Neuroblastoma is the commonest cause of hematogenist metastases in children followed by primary bone in soft tissue sarcomas, osteosarcoma, and Ewing's being the commonest following, followed by rabdomyosarcoma. And neuroblastoma at our centre we assess by MIBG. So that's an image of an MIBG scan. You can see uptake in the proximal femur and the skull. There's no uptake in the growth plate, so that's the real advantage of MIBG. You don't have to be confused by uptake in growth plates that you get on a bone scan. There are some cases of neuroblastoma that don't take up MIBG and those cases can be quite difficult and you may resort to using MRI for things like that. Just some things to outline. When you're looking at bone tumours on MRI and you're trying to look at what the length of the tumour is to try and plan their surgery and to also see if there are skip lesions, MRI can overestimate the tumour extent and that's because merodema can look very heterogeneous around a tumour. This is a telangiectetic osteosarcoma, so you'll note on the T1 image on the left there are actually some fluid levels and there's some high signal in this because there's some blood products in this tumour in the proximal humerus and it looks like it extends quite well down the shaft but you'll notice that on the bone scan the area of bone turnover is not quite so extensive. It's very difficult in some cases to know whether or not to believe the bone scan or the MRI. I would be more partial if we'd done a PET scan to believe the true extent of tumour on a PET scan but again a PET scan doesn't give you really the resolution in terms of muscle compartments and so on involved. Often in these cases it's a case of treating and then doing a follow-up MRI with or without other imaging to see what things look like before their surgery. So just be careful about commenting on tumour extent. If you're seeing lots of heterogeneous merid changes just be careful about where you're thinking or with what degree of confidence you're saying you can see where the tumour goes. So this is another example of a skipped lesion in a 14-year-old with an osteosarcoma you can see the primary lesion in the tibia and there's this area of high signal in the shaft but there's nothing corroborating on the T1 and because there's nothing on the T1 we can say that that is perhaps related to altered walking or stress in the bone but there's no lesion there. This is not a skipped lesion so just be careful when you're looking at the fluid-sensitive sequences. Radiation can also have varying effects on marrow. A radiation insult causes different changes so in the acute stage after two weeks of treatment or within two weeks you get hemorrhage and edema. Then in the next three to six weeks you can get the disappearance of red marrow and eventually you get fatty marrow replacement at six weeks from beyond. And so these are the kind of signal changes you'd expect to see. You'd expect to see marrow edema as stir hyperintensity which may be heterogeneous if you've got hemorrhage as well. And in three to six weeks you'll lose the T1 hyperintensity before you finally get fatty marrow replacement and high signal. And the degree, whether or not you're going to get marrow recovery depends on the dose you give and in higher dose, if higher dose has been given you may not necessarily get normal fatty American version that's homogenous. Things can look quite heterogeneous in some cases. So this is a case of radiation-related change. So this patient has a medulla blastoma and they've had whole cranius neural axis radiotherapy and you can see image A is before treatment and image B is after treatment and you can see the yellow marrow conversion and you can also see that it's not homogenous. Part of this may be that some areas have received a little bit more dose or overlap and so those areas for example in the C spine perhaps have got a little bit of overlap of dose but these are usually things that we try and avoid with modern radiotherapy. This is a pathological fracture in the distal femur. It's something that I have not seen in my own working practice and it's felt that with more targeted and lower radiotherapy doses this is something we see less frequently but avascular necrosis is still a real complication of both radiation and other treatment that patients such as people who've got, children who've got leukemia get and so you may see a fracture complicating cases of bad avascular necrosis. This patient had popliteal fossa treatment. I'm just going to pass this slide. This is just to say that again to emphasize that chemotherapy can also cause similar heterogeneous changes and red marrow hyperplasia and that if you've been given GC-SF it can also transiently cause while your own treatment also cause heterogeneous marrow changes and you can see on the image on the right once we've stopped granular site stimulation factor it's those proximal femoral changes have resolved. Whole body MRI, I'm not going to talk in detail about this because it's probably a talk in and of itself but the short of it is that MRI is 98% sensitive for looking at bone disease in certain tumor groups including Ewing's Rhabdomyo sarcoma, so the primary bone sarcomas and that's better than PET. The downside to MRI is that it can often overcall small lesions and if you're doing whole body MRI imaging, it's quite a long exam. It can be anything from 40 minutes to an hour and so you don't always have time to do additional MR slices in a child to solve and see whether or not the high signal you've seen on a stir sequence for example is real pathology and of course in children we're involving the dose of PET CT and it's certainly in some tumor groups. So what protocols do you use particularly for malignancy? You're going to use coronal stir and coronal T1 and you want to see the stir because it's going to help draw your eye to the area of marrow signal abnormality as high signal and you want to look at the T1 just to make sure that you've got marrow replacement so you're looking for that hypotensity relative to muscle. Some centers also use DWI but we don't do that in my practice and I think that's an area of ongoing research. We have now instated a policy of using this routinely for Ewing sarcoma and Rhabdomyosarcoma whereas osteosarcoma depending on the scenario we can rely on bone scans as well and it's ideal for diseases that are multi-centric even though LCH is not strictly a malignant disease and CRM certainly isn't. So I've spent most of my time on the malignant parts because they're obviously the areas of concern where the stakes are high. I'm going to talk briefly about benign disease as well. So osteomyelitis is probably something that you'll see very commonly and that we see very commonly. The organisms that cause osteomyelitis vary and so your mix may be slightly different to the Staph aureus and Kingella particularly that we see. Most children are quite young who get osteomyelitis at least in Northern Europe and North America and most of it is seeded hematogenously. It can be diagnostically challenging for the clinician because not all of them have raised inflammatory markers and white blood cell counts and the radiographic features are not helpful usually. So what's the sequence of events? The insult usually starts with hematogenes seeding into the metaphysis where the blood supply is and then your infection builds from there. It can cross the growth plate, it can cross into the joint, it can extend into the soft tissues and into the subperiostal region. So these are all the areas that you need to really be able to look at when you're thinking about imaging or infection. So in the early stages of marrow infection you're going to have high stir signal, so you're going to have displacement of water protons and you'll have enhancement. This is not a malignant process but if you really have osteomyelitis then you'll get replacement of the marrow with pus and other material so you'll get low T1 signal and this is a case of acute tibial osteomyelitis. As the pressure builds within the medullary cavity you'll actually get devascularization of bone and you can see in this distal tibia there's an area which no longer enhances but there's marrow around it which is enhancing. Eventually you'll end up with this, so this is an intraosseous axis within the right acetabulum. You don't need the contrast sequence to diagnose it. The reason I've showed you the post-contrast T1 coronal on the right is to show you that actually there's a right proximal femoral epiphysis growth center which is not enhancing. So there's actually septic arthritis as well because you'll notice that there's a joint effusion on that side and the pressure from that joint fluid is actually affecting the vascularity of the epiphysis. So this is something that's going to push the surgeons to intervene overnight rather than waiting to the next day. Subperiosteolapsis, so this is just to remind you we're looking for all the different complications. So we're looking for subperiosteol collections or anapsis. If you have chronic osteomyelitis you'll get a sinus which drains into the subperiosteum and that's called the cloaca. So this is a cloaca and again it's extending from the vices into the anterior tibial cortex but there was no tract from the bone to decompress pus into the skin. And so this is the bit near the vices, the brodysapsis and the tract that goes down from it, that's a cloaca. I don't want to talk in detail about sequestra, but I just want to say that they can be very difficult on MRI. Often you see low signal within a cavity and sometimes I think I over-call sequestra and perhaps sometimes we should be suggesting CT if it's making the difference between having sequestrum removed and having a conservative treatment with IV antibiotics because a sequestra is a bit of dead bone which will not be treated with IV antibiotics. This is the last thing I want to touch on really before I, just before I finish very quickly. So this is a case of, I think it was a six or eight year old boy. He's come with some swelling over the medial clavicle and there's a PD-fatsan image on the left that shows high signal in the medial clavicle but there's also some low signal interspersed within it and when we did an x-ray we actually knew that he had some sclerosis and expansion there and he had a biopsy which didn't show any organism and his inflammatory markers were completely normal and so eventually the diagnosis was made of CRMO. So what is CRMO? It's also called CNO but CRMOs stand, they're both the same thing and CRMO is chronic recurrent multifocal osteomyelitis and it's essentially a non-infective osteitis that is idiopathic. We're not quite sure what the cause is. It's self-limiting in most cases. Most patients present before the age of 10 years but it can occur either in the long bones or in the spine and we've certainly seen a few cases in the spine where there's a bit of vertical body height loss and where you might actually be concerned about LCH as a differential. If the medial clavicles are involved in our local practice and there's no other concern, there's no other particular feature to point to anything else there's no rays, no inflammatory markers or anything aggressive on a plane film then we don't do a biopsy. We're very happy to say that that is CRMO and the rheumatologists take these patients on and follow them up. The tibial metaphysis in the literature is also a characteristic location. It's possible in the future if we've got proximal tibial metaphysial involvement we may also be looking to skip biopsies but that's not our practice at the minute. I'm not going to talk about fractures and stress injuries. I think I'll skip this because of time because this is very similar to adults. Again, biomechanical changes is something that's also quite similar to adults. This is these patchy areas of high signal on stern imaging they can be reproduced in trial patients if you put an insole in them so we know altered walking and weight bearing can cause patchy areas of high merus signal. I'm going to skip on your crosses as well and this is really what I want to talk about transient merodema. So we see a lot of patients with possible cancer or referred as possible cancer who have subtle periostal reaction. This case is a 14-year-old with JIA who actually had an MRI done to look for synovitis and there isn't any synovitis in this ankle but there's some heterogeneous merodema in that distal fibula and there's some periostal reaction. But there's nothing aggressive otherwise on the x-ray this is nice smooth periostal reaction and so we chose to do some repeat imaging and you can see the repeat imaging at two months shows that this abnormality is now consolidating and almost resolved. It's very difficult sometimes to know what to label these as is this transient osteoporosis is this a stress fracture that's healed. A stress fracture is always something worth thinking about but the important thing is just to take a good clinical history examination and to have everyone relevant involved before jumping into telling patients that they have potential cancer and doing biopsies. I want to finish on that note and I want to summarise by saying in malignancy MRI can definitely help distinguish benign from neoplastic disease with a reasonable amount of sensitivity it can identify neoplastic marrow replacement and sites of metastatic disease and it helps tell us what the anatomy is for biopsy and treatment and identify complications. We can't delineate bony changes or matrix calcifications so we still must think about doing radiographs and we don't always get a definitive diagnosis sometimes we need tissue for that. And MRI and MRI cannot provide a diagnosis without a clinical context that's very important and hopefully that case of transient bone marrow demons emphasise that. We cannot also differentiate osteonecrosis from infection there's a lot of literature out there but at least that is my position and that of my colleagues it's very difficult for us alone to differentiate infarction from infection. I hope that was useful to you I wasn't able to join you live unfortunately but if you have any burning questions or queries or if there was something that I said that's not clear I'm very happy to have an email and answer any queries from there. Thank you very much for listening. Thank you Dr. Zeeshan for taking out time and giving us this extensive talk and clearing the concepts. Malima would you like to add something? Yeah so I think a very important topic marrow is always confusing and Zeeshan did a very simple meditator pattern and approach as to how we can do that. And with this we come to the last session Mitrucha you can continue. Questions please put in the Q&A and also Zeeshan has given us an email so the one in the Q&A will be answering them. So with this we come to the last session for the MSK Masterclass and this will be by Dr. Jathali Parik. She is consultant musculoskeletal radiologist at Nawaran Diagnostic Centre and she has finished her fellowship and interventional radiology at Innovation Imaging Mumbai. She has received Dr. Aran Goyal Gold Medal for best paper presentation on MSK Interventions and Jathali is quite famous even on our YouTube channel of Indian radiologist for her wonderful talks.