Introduction to CT Head: Approach and Principles

This talk is intended to serve as a good introduction to CT head. If you're a new radiology resident, you must understand the basic concepts presented here so that you have a strong base to build on moving forwards. If you're a generalist, whether you be an emergency medicine, internal medicine, or whatever, this will give you more than enough to start looking at CT heads right away. It will also give you more than enough to be way ahead of your peers, and we know how important that is to internal medicine residents. Either way, the focus of this talk is the basics, but there will be pearls and some subtlety here which is going to become must-know for future radiologists.

Before we get started, I strongly suggest that you watch the first talk in this series called "A Practical Introduction to CT." Many concepts introduced in this talk will build on a basic understanding of what was discussed in the previous talk in more detail. To start off, we will briefly review some key concepts, discuss some basic relevant anatomy, and then we're going to get right into it by introducing you to a conceptual approach to CT scan. We will then introduce some basic key concepts, really just what you need to know to start looking at CT scans right away and efficiently.

Next, we will discuss windowing. Windowing is of utmost importance when looking at CT heads. It's very easy to miss key findings if you don't window appropriately. After that, we will learn by repetition. We will go through multiple cases with our approach and directly apply the things that we just learned. For those that are interested, the last five minutes or so will involve a demonstration of a more complete approach, including blind spots. This is mainly for those going into radiology.

Quick review again: I can't emphasize how important it is to know the basic radiographic densities and know some ballpark figures. The Hounsfield unit scale is a measure of density from negative a thousand (the least dense) to beyond positive a thousand (the most dense structures). By convention, zero is water and negative a thousand is air. Everything else is calculated based on that. Cortical bone and metal are very dense structures, near positive a thousand. Soft tissues are above zero. We will simplify and use a ballpark figure of positive 50 for now, and fat is less dense than water. We'll give a ballpark number to remember of negative 100.

Now let's focus on the main densities that we're going to deal with in the brain and the skull. We have the brain itself, sitting in a pool of CSF. The brain is made up of mainly gray matter and white matter. In the simplest terms, the gray matter is the thinking part of the brain, found mainly in the cortex or outer part of the brain and some deep gray matter structures. Gray matter consists of mostly neuronal cell bodies. White matter is made up of mainly myelinated axon tracts. The normal densities of these structures is important. CSF is low density; it's mainly water. It measures around 15 Hounsfield units. White matter measures approximately 25 Hounsfield units. Gray matter measures approximately 40 Hounsfield units. Knowing these actual numbers is not really that important. Instead, understand the basic principle here: CSF is near water density. Gray matter is slightly more dense than white matter. A simple way to remember the relationship between gray and white matter is that white matter is mainly made up of myelinated tracts. Myelin is a fatty substance, and fatty things have a lower density. Lastly, we'll introduce acutely clotted blood, which measures about positive 80 Hounsfield units. The principle to remember here is that acutely clotted blood is more dense than the rest of the brain. The last thing to take away from this slide is that all of the densities we're looking at here are in a tiny range of densities, and that's why windowing is so important when we look at CT scans of the brain.

I'll tell you right away, there is a lot of anatomy that one can learn in the brain, and how much you should know depends on your goal. This introduction to basic anatomy is not meant to be comprehensive, but it is more than enough to get you started. You may know most of this already and can treat this as a review, but if you don't, you can watch this portion over again to learn the basics. If you're going into radiology, you're going to need to know way, way, way more anatomy than this. I strongly suggest subscribing to eAnatomy, also known as IMAIOS. You can take your time learning the cross-sectional anatomy in detail.

So here is a normal CT scan of the head. This is the right side of the patient here. This is the left side of the patient. This is the anterior portion. This is the patient's forehead, and this is the posterior portion of the head. We've marked some of the major densities that we've discussed already. This here is CSF; it's the ventricles filled with CSF. The average density here comes out to be about 5 Hounsfield units. We said it's about water. This is white matter marked here. The average density in that selected area is about 27 Hounsfield units, and the gray matter is selected here with an average Hounsfield unit number of around 39 Hounsfield units. Again, the numbers aren't so important, but CSF is near water density, and the gray matter is slightly more dense than the white matter.

Okay, let's move on to basic anatomy. Don't get bogged down with memorizing any specific names for now. Just listen and take in whatever is relevant to your current level. In the midline here, we have the falx. This is an invagination of dura that separates the two hemispheres. As we scroll down, it's continuous with the tentorium, which we can follow down here. The tentorium looks like a tent. Everything above the tentorium is considered supratentorial brain, and everything inferior is considered infratentorial. The supratentorial brain is separated into lobes, which many of you guys know: the frontal lobes here, the parietal lobes back here, the occipital lobes back here, and the temporal lobes over here. The specific divisions between the lobes is beyond the scope here, but I do want to touch on a couple. First, these here are the Sylvian fissures. So, fissures separate the temporal lobe below from the frontal and parietal lobes above. Second, this here is the central sulcus. The central sulcus separates the frontal lobe anteriorly from the parietal lobe posteriorly. In front of the central sulcus, we have the precentral gyrus, which is where motor control sits, and posterior to the central sulcus, we have the postcentral gyrus, which is mainly sensory control.

There are a couple of ways to look for the central sulcus. Actually, there are many ways to look for it. I'll only talk about a couple very quickly. You're looking for this reverse omega sign. It looks like an omega flipped upside down. When you look for that, this little bulb of tissue here is the motor control area of the hand, or more specifically, the thumb. Second way to find the central sulcus, or another way to find the central sulcus that's practical, is looking for this in the midline, just inferiorly. Call it the mustache sign. It kind of looks like a mustache. These are the marginal sulci. If I scroll up, the sulcus anteriorly to that is going to be the central sulcus, and it is here as well. If you aren't going into radiology, don't worry too much about these details. Just know the basic lobes for now.

The gray matter structures you're going to need to know are as follows. I windowed it differently so that they pop a little bit more. Firstly, within each of the lobes, there is cortex, or the outer portion is made up of mainly gray matter. For example, this here in the frontal lobe is all part of the frontal cortex. This is a gray matter structure. Then you're going to need to know the deep gray matter structures as well. And this slice right here is going to be very important. I want you to burn it into your brains. Okay, these are the deep gray matter structures that you're going to look for when you're looking for stroke on any CT head. This here is the caudate, or the head of the caudate. This is the putamen, and the globus pallidus, which is a little bit more inconspicuous here. This here is the thalamus.

There are a few white matter structures you're going to also want to know on this same slice. This here is the internal capsule. This is the external capsule over here. As I scroll up, this here is the corona radiata, and this, as I scroll up, is the centrum semiovale, just above the ventricles.

The cerebellum is infratentorial, or below the tentorium here. For now, just know that in the midline, we have the vermis, and we have the right and left cerebellum. Here are the cerebellar tonsils inferiorly. Notice the normal location on a normal scan. This here in the midline, where my mouse is, is the brainstem. The brainstem is separated from superiorly to inferiorly into the midbrain, the pons, and the medulla. Okay, the way is the easy way to recognize these are by recognizing their shape. So the midbrain is commonly referred to as a Mickey Mouse head, with these being the Mickey Mouse ears over here. The Mickey Mouse ears represent the cerebral peduncles. Okay, as I scroll down, this is the pons. The easy way to recognize the pons is people refer to it looking like Darth Vader. More inferiorly, we have the medulla, and then the spinal cord. This here is the foramen magnum. Notice that there is CSF surrounding the spinal cord here, and that the cerebellar tonsils don't crowd this area on a normal CT scan. That'll be key.

Next, the CSF spaces are very, very high yield. First, the ventricular system. Most basically, these here are the lateral ventricles. This is the third ventricle here, sits like in the midline. This is the fourth ventricle here, behind the pons. The basic CSF flow is as follows: CSF is made by choroid plexus. We'll talk about that a little bit more later. CSF flows from the lateral ventricles to the third ventricle through the foramen of Monro, can visualize right here. The third ventricle, CSF flows through this spot here, which is called the cerebral aqueduct, and into the fourth ventricle. Okay, from the fourth ventricle, there's communication with the subarachnoid space through the foramen of Luschka bilaterally and the foramen of Magendie. Don't worry about all the names for now, just understand the principle of the flow of CSF. Any blockage of flow of CSF within the system is going to cause dilatation to the ventricular system. Approximately, for example, blockage of the aqueduct here is going to cause dilatation of the third ventricle and the lateral ventricles.

Next, we talk about the basal cisterns. These are just the CSF pools that sit at the bottom of the brain. These pools are assigned names that are important to know. I won't discuss all of them, but just the most high-yield ones for now. In front of the pons here, there's the prepontine cistern. If I scroll up, you guys will know that this is where the pituitary sits, is the pituitary fossa/sella turcica. As I scroll up, we have this cistern here, which is known as the suprasellar cistern. This is a very important cistern to keep in your minds. It's that famous star-shaped cistern. We already described this structure here as the midbrain. These are the cerebral peduncles here. This is the interpeduncular fossa. On the sides of the midbrain here are the ambient cisterns, and posteriorly, we have the quadrigeminal cistern.

Lastly, we talk about the vascular structures. You can see on plain CT head, you often look here for signs of acute clot. We already said that acute clot is going to look bright because it is more dense than the other structures in the brain. So a very dense vessel might signify an acute thrombus. First, we discuss the arterial structures. Start with the anterior circulation for now. Know the internal carotid arteries can be seen beneath the skull base. They pass into the skull. Eventually, they split into the MCA and the ACA. This is the MCA, or the middle cerebral artery. You're going to look here for acute clot and stroke. The MCA splits into branches that run in the Sylvian fissure, so you can also look here for signs of acute clot or acute clot in the more distal MCA branches. You can also follow the distribution of the ACA. It might be hard to see on plain CT head, but they should run in this region here and then up here. You're going to look for areas of clot there too, but that's more advanced. Lastly, these are the vertebral arteries here. This is the basilar artery here. It separates here the PCA arteries. Again, these might be hard to see, especially when you're not using thin slices. You're actually interrogating vascular structures, you're going to do a CTA anyways, but on plain CT head, it's important to know where to look for acute hyperdense vessel sign for a stroke, and we'll talk about that more later.

Blood drains from the brain parenchyma into veins and that eventually gets into the venous sinuses. The basic venous sinuses that you're going to look at on CT head are as follows. This here is the superior sagittal sinus. Superior sagittal sinus separates down here into the transverse sinuses, and the transverse sinuses become the sigmoid sinuses here and eventually the jugular vein beneath the skull base. Again, acute clot within these sinuses is going to show up as bright clot in the venous sinuses, known as venous sinus thrombosis. It can cause venous infarcts. You're gonna look for an extra bright sinus.

So I know that was a lot of information, especially if this is new for you. If you're not going into radiology, this is more than enough for now. Watch this as many times as you need to. Don't bog yourself down with too many names. If you keep looking at CT scans, you're going to pick up things as you go along. You can also go to headneckbrainspine.com where you can practice scrolling through images and identifying basic anatomy.

When you first start, the best way to approach a CT head is by thinking of things conceptually, based on the things that you never want to miss. Those are radiology. Do this too, but the list of things gets much longer. For now, let's keep this list shorter and more practical. First, take a couple scrolls through the head and look for anything big. You're going to get better at identifying things as you move forward, but you need to do more than just this. For most ER physicians, I've seen this is the end of their approach. They look for anything big on a couple of scrolls and call it a day. That can be dangerous, especially if you're in the community and won't get a read on the CT scan until the next day.

Once you've taken a few scrolls through the head to look for anything big, you need to specifically look for signs of mass effect and herniation. I put this first because this is the thing that beginners forget to look for and most commonly miss. It's easy to miss subtle herniation if you don't look for it. Then you specifically look for blood, then specifically look for signs of stroke, and then you're going to look at soft tissues and bones. But for now, especially for generalists, focusing on remembering the start points: look for mass and mass effect/herniation, rule out a bleed, and rule out a stroke.

There's some basic material that you're going to need to understand first about bleed, stroke, and mass effect in order to look for these adequately. Let's start with bleeds. We mentioned that acutely clotted blood is denser than everything else in the brain, so acute bleeds will look bright. So when you're trying to rule out an acute bleed, you're looking for things that are bright. I'm sure most of you guys know this, but the major basic types of bleeds are as follows: an epidural hematoma, which is a biconvex or lenticular shape because it looks like a lens and it doesn't cross the suture lines; a subdural hematoma, over here, it's crescentic and it crosses sutures but generally not the midline; subarachnoid hemorrhage, here, is high-density material in the subarachnoid space. This is blood on a suprasellar cistern, looks like a star, which we talked about earlier; you have an intracerebral hemorrhage, which is blood in the actual parenchyma itself.

Let's take this one step further. There are two more concepts that are crucial to understand about the density of blood products. Firstly, as we mentioned, acute clotted blood is high density. It's about positive 80 Hounsfield units, so it's going to show up as bright. As the blood progresses over time to subacute and then chronic, the density decreases over time. In the subacute phase, at some point, the density is going to be similar to the brain parenchyma, so it can be difficult to see subacute subdural sometimes, but looking for mass effect is going to help you with that. We'll talk about that later. This here is an acute subdural. Notice the blood is brighter than the parenchyma in this image. This is a subacute subdural. Notice the density is lower, but it's not quite CSF density. Eventually, a chronic subdural will approach the CSF density, like this one over here.

The second concept to understand is the density of hyperacute blood. We know that acute clotted blood is dense, but before it has a chance to clot, it's going to have a lower density. This is an example of a patient with an epidural hematoma. It is mainly high density, as you can see here. The areas of high density represent acute clotted blood, but there are also areas of lower density within, like here and here and here. These areas represent hyperacute blood that is not yet clotted. Some people refer to this appearance here as a swirl sign, but the name of the sign doesn't matter so much. Do you understand the basic principle that the areas of low density here represent hyperacute or ongoing bleeding, you'll be fine. If we apply this further, here we have a picture of a subdural hematoma along the bone that is mainly low density, but also high density inferiorly. This is a mixed-density subdural hematoma. This is the classic appearance of an acute on chronic subdural hematoma. This is the acute component here, and the rest of it was the chronic component. However, what we just learned, this could also be acute blood that for some reason hasn't completely clotted, so it won't be completely bright. The inferior part is the part of the blood that was able to clot, and the superior part is the not-yet-clotted blood. If a patient is on a blood thinner, for example, you can get this appearance in acute trauma. Keep this in mind as an alternative explanation for this appearance.

Very quickly, you should just be aware that not everything that's bright in the brain is blood. A lot of structures commonly calcify normally, and you shouldn't call them bleeds when you see them. Firstly, choroid plexus is often calcified. Choroid plexus, as we mentioned, produces CSF. It is found mainly in the ventricular system here. It's in the lateral ventricles in a sulcus adjacent to the thalamus. It is not found in the frontal horns or the occipital horn of the lateral ventricles. In the fourth ventricle, sometimes you can see it sticking out of the foramen of Luschka bilaterally, and you shouldn't be alarmed when you see that. Secondly, the pineal gland in the midline, which is calcified here, very commonly calcified. Thirdly, the globus pallidus is very commonly calcified and is calcified bilaterally here. Again, this is not blood. And lastly, I'll mention one more, which is not as common, but the dentate nucleus in the cerebellum here, mainly unilateral. So we can't be 100% sure whether this is calcification or hemorrhage given its appearance and unilaterality. You might have also heard the term pseudo-subarachnoid hemorrhage, which occurs when the entire brain is a diminished dark and the vessels in the sulci appear bright and can be mistaken for blood. Again, you don't need to worry about the details for now, just understand the principle. Not everything that looks bright in the brain is going to be blood. These are some common fake-outs.

Okay, let's move on to ischemic strokes. For now, there are three major things that you need to understand and remember about ischemic strokes on plain CT head. The first two are signs of acute stroke. The first thing to look for is the hyperdense vessel sign. As we know, acute clot is bright on CT. Ischemic stroke is caused by acute thrombus, including a vessel. So the earliest sign of acute stroke on plain CT is actually visualizing that acute thrombus. This is the case of a hyperdense vessel in a right MCA branch. We will talk about where to look for hyperdense vessels when we go through our approach later, but for now, understand the basic principle.

The second main thing to look for is loss of gray-white differentiation. As we have mentioned many times, the gray matter is slightly more dense than the white matter. In acute ischemic stroke, the cells in the gray matter distal to the occlusion are deprived of blood supply. As a result, they can't function, they can't make ATP, and their sodium-potassium ATPase stops working. As a result, the cells themselves swell up, what we call cytotoxic edema. In other words, their water content increases. We know water is less dense, so the density of the gray matter in the area is going to go down and look similar to the white matter. You couldn't follow the physiology. All you have to know is this: when the cells are deprived of blood supply, the cells themselves swell up, and as a result, look less dense, and you lose the gray-white differentiation. It can be extremely subtle early on and then look more obvious later. This is the same patient corresponding to the vessel clot we just saw. There is an area here with subtle blurring of the gray-white differentiation compared to the other side. Notice how crisp the gray-white differentiation is here, and it also is anteriorly on the right as well, but in this area here, you've lost the gray-white differentiation. This is a subtle case.

The third concept to know about ischemic stroke is the evolution of stroke over time, and it's very important to know competently calling a stroke old versus something acute or subacute obviously has clinical implications. In the simplest terms, in general, the infarcted tissue decreases in density over time. So here we have an example of an acute stroke in the left MCA territory. Notice the blurring of the gray-white differentiation on the left, compared to the other side, which is very crisp. This here is the insular ribbon, and notice how it's lost here as well. So that's an acute stroke. As we said, in general, the infarcted tissue decreases in density over time. So here, two days later, the infarcted tissue has, in fact, decreased in density over time. The second thing to know is that the mass effect is worst at around 3 to 5 days post-infarct and then decreases thereafter. This is the same patient on day 5. You can see how much worse the mass effect is. Got it. Lastly, old strokes will eventually have negative mass effect, often seen as dilatation of the ventricle close by, is called ex vacuo dilatation. Here's a different patient with a left old MCA infarct. Notice the density here is near CSF, and there is negative mass effect. There is ex vacuo dilatation of the left lateral ventricle. Now, I covered this superficially so you can understand the basics. I suggest further reading or watching the Radiopaedia articles on evolution of stroke if you want to understand this in greater detail, but this is more than enough for now.

So we've talked about bleeds and strokes. The third big thing that I want to stress in this talk is looking for mass effect and brain herniation. The reason I stress this is firstly because seeing mass effect can help you find more subtle pathology like a tumor on non-contrast CT or a subacute subdural, and even more importantly, it may be a critical finding that requires neurosurgical intervention ASAP. In my opinion, forgetting to specifically look for mass effect is the most common mistake that beginners make, and it can have massive consequences. We can't go through it all, but here are the basic concepts. One, look for sulcal effacement. When you scroll through an image, first you should be looking at the sulci and notice when they are effaced, like they are here. In this case, we know that there's going to be something on the left side that's exerting mass effect to cause this sulcal effacement. Secondly, look for midline shift. This is a case of subacute stroke that we just showed that demonstrates severe swelling in the left brain tissue. We measure the midline shift by drawing a line from the attachments of the falx anteriorly and posteriorly and then measure the distance from that line to a midline structure. Most commonly, we use the septum pellucidum.

The third thing we look for is evidence of other brain herniation. This is an extremely important slide. Again, if I get one thing across in this lecture, it is this: specifically looking for herniation syndrome is on every CT is unbelievably important. I'll repeat myself: in my opinion, it is the single most clinically significant thing that is missed or overlooked by beginners. Significant herniation can be easy to miss if you don't specifically look for it, and it's something that may require immediate intervention. The most basic herniation syndromes are as follows. The first is subfalcine herniation, called subfalcine because it refers to displacement of brain underneath the free edge of the falx. Here, in simplest terms, when you have midline shift, you're going to probably have a little bit of subfalcine herniation, as we do in this case here. The second place you're going to look for herniation is the suprasellar cistern. So we know this is the suprasellar cistern here, and this here is the uncus on the left. As you can see, there's displacement of the uncus slightly towards the contralateral side, or towards the right. The group of more severe herniation here are called transtentorial herniation, named because it refers to supratentorial structures descending through the tentorium. This is another case. If you have bilateral uncal herniation, you can get complete effacement of the suprasellar cistern. Again, if you don't specifically look at the suprasellar cistern for signs of uncal herniation, you might scroll right by this. It's harder to see something that's not there rather than something that is. This is another case here. We talked about tonsillar herniation. This can be subtle and is probably the most often missed on axial slices. Tonsillar herniation refers to the downward herniation of the tonsils through the foramen magnum. On every scan, you need to look at the foramen magnum, which is right here, to make sure that it's not crowded. If there is even a question, you need to get sagittal reformats to assess further. This patient has tonsillar herniation and requires urgent decompression of the posterior fossa. I've only mentioned the three most common and important herniation syndromes, but there are more to know. For example, ascending transtentorial herniation occurs when there is mass effect causing infratentorial brain to herniate upwards through the tentorium, as in this example here. Notice that the quadrigeminal cistern here, that should sit behind the midbrain, is completely effaced. As a result, other herniation syndromes include transcalvarial, which is through the calvarium, and transalar, which is herniation across the sphenoid ridge. If that demonstration was too much detail for you, for now, take away only these three: subfalcine, uncal or descending transtentorial herniation, and tonsillar herniation.

The next step, mainly if you're going to go into radiology, is to know the complications of these herniation syndromes, which you should specifically look for. I'll list them here quickly but won't belabor the point. If you're going into radiology, you need to know these cold, so learn them. The RadPrimer/StatDX article is a good reference. It's actually not that difficult to remember these complications if you learn the anatomy. For subfalcine herniation, the foramen of Monro, that drains each of the lateral ventricle, sits right there. The brain gets pushed over the contralateral foramen of Monro, gets compressed, and you get acute lateral hydrocephalus. The anterior cerebral arteries, or ACA, also run here. With severe subfalcine herniation, these can be compressed against the falx, and you can get ACA infarcts. For descending transtentorial herniation, think anatomically to figure out what's going to be compressed. The third nerve runs here, so you can get third cranial nerve palsy, which causes a blown pupil. The cerebral peduncles of the midbrain, or ears of the Mickey Mouse, right here as well. Compressing the ipsilateral cerebral peduncle would cause contralateral weakness if you remember your basic neurological pathways. However, the contralateral cerebral peduncles can also be pushed against the tentorium on the other side, so you can often get ipsilateral weakness, referred to as a false localizing sign, or Kernohan's notch. Other complications of descending transtentorial herniation include hemorrhage in the midbrain, known as Duret hemorrhages, or Duret hemorrhages, and PCA infarct. The posterior cerebral arteries run here as well. For tonsillar herniation, you can get obstruction of the fourth ventricle, which causes dilatation of the other ventricles, also known as hydrocephalus. Tonsillar herniation can also push in the brainstem, where you have structures that are important for respiratory and cardiac control, and this can cause respiratory or cardiac failure, aka death.

Okay, so take a deep breath. Pause if you need to, but that's it for the real hard content that we're going to cover here. The rest of the talk focuses on practical knowledge that you're going to need to know to look at CTs, and then some examples of cases. As I mentioned before, for non-radiologists, the best way to approach a CT is by thinking of things conceptually, based on the things you absolutely never want to miss. We already went through this once before, but let's go through it quickly again. First, take a couple scrolls through the head and look for anything big. Then you need dedicated searches to allow big pathologies. The first thing you should look for is any big mass or mass effect, and I put this first again because it's easy to forget to look for this, and the consequences can be huge. Then we do a run specifically to look for bleeds, and then for signs of stroke that we discussed. Looking at the soft tissues and the bones are also important, but findings can be more subtle. If you're not going into radiology, focus on the start pathologies. In order to search for these management-changing findings adequately, it's critical that you understand windowing. If you don't window, it's very easy to miss things.

The basic concept of windowing is discussed in the first video. In short, our eyes can only detect a finite number of shades of gray. If we were to distribute all these shades of gray across the entire scale of densities that we might see, for example, from air at negative a thousand to very dense structures at positive a thousand, like this, most of the tissues that we're interrogating in the brain in this region here will look like one or two shades of gray. So we're looking for blood or a stroke, like that, we're not able to see it. The principle of windowing is playing around with this distribution of shades of gray so that we can actually see with our eyes clinically important density differences. There are two numbers that define any window: the window width, which is 2,000 here, and the window level, which is where the window is centered here. It is zero. When we open a CT head, the default is a basic brain window. This allows us to look at the anatomy, evaluate for any large masses, mass effect, see if there's anything in the CSF, etc. This is the default window, and many abnormalities will present themselves on this basic brain window, but they can often be relatively subtle or even impossible to see on brain window. It is a relatively narrow window with a width of 70 and a center centered at about 30, near the main structures that were evaluated here: the CSF, the white matter, and the gray matter. All are represented by different shades of gray within this range. Anything above positive 65 Hounsfield units will be completely white. Anything below negative 5 Hounsfield units will be completely black. So you had acute clotted blood, which is around positive 80, it's going to look very bright, almost completely white, and it's going to stick out like a sore thumb. What if you have an acute subdural hematoma later along the bone? The blood, as we can see here, will be completely white, and the bones surrounding it will also be completely white. Despite such a huge difference in density, both are represented by white, and it's difficult to tell them apart. It's going to be super difficult to know if there's any blood layering along the skull. So how do we solve this? We use a blood window. Here, the width is 180 and the level is 80, centered at where we expect clot or blood to be. That way, clotted blood appears grayish, and bone appears completely white, leaving no question whether or not a subtle subdural is present.

So let's look at the brain window again. As we discussed, when looking for stroke, a key finding indicative of stroke is the loss of gray-white differentiation. So in evaluating for stroke, we really have to compare the densities with our eyes of the gray matter and the white matter. If we do that on this window, the gray matter and the white matter are represented by different shades of gray. But if you looked at CT before, you know that gray-white differentiation can be hard to evaluate well on basic brain window. Let's introduce the stroke window by narrowing the window here. The difference between the gray matter and the white matter is accentuated, making it way easier to compare the gray matter and white matter to look for signs of stroke. The gray matter is represented by a shade of gray that's more bright, and the white matter is represented by a shade of gray that's darker. Again, when we use different windows, we're looking at the same density information. Windowing is the principle of playing around with the distribution of shades of gray to help us see pathologies better. Then we use our soft tissue window to look at the orbits and other soft tissues. A more detailed approach for radiologists will be included at the end of this talk, and then we look at the bone window to look for fractures. Again, this will be discussed more in the detailed portion of the talk at the end.

Let's go through a CT scan and demonstrate our conceptual approach that we just learned, and then we will practice our simplified conceptual approach on a few illustrative cases. Okay, so I stress this is a conceptual approach. It's very simplified. If you're going into radiology, your approach is going to be more detailed, and we'll talk about that approach in a few minutes at the end of the talk. So conceptual approach: take a couple scrolls through the brain, look for anything big. As you look at more, you're going to notice more. Okay, there's nothing big here. Next, my first of three star pathologies is looking for our mass effect. So you take your first few scrolls down, looking at the sulci bilaterally and comparing them. You're looking for any sulcal effacement, and there is no significant sulcal effacement here. As I scroll more inferiorly, you're going to look for any midline shift or subfalcine herniation. Look at the suprasellar cistern. Look for any signs of uncal herniation. This is the uncus here. Then go down to the foramen magnum and look for signs of crowding here to suggest tonsillar herniation. If you have any question, get sagittal reformats.

Next, you'll assess for bleed. So starting again at the foramen magnum, you want to scroll up, and you're looking at all the CSF spaces, so all the basal cisterns we talked about: prepontine, suprasellar, etc. Then you want to look at all the sulci up here for any hyperdensities in there to suggest bleed, and there's none here as well. Blind spots for blood are more subtle and often dependent in the often in the dependent portions of the brain, such as the interpeduncular fossa, right here. You can often see very subtle hemorrhage there in trauma, and in the Sylvian fissures here, in the most dependent portion, you can often see hyperdensity there representing subtle blood, and also in the occipital horns of the lateral ventricles, you can see intra-axial blood. Then we look for signs of extra-axial hemorrhage. So when we scroll up, we're looking at the anterior portion here to look for any extra-axial blood. We can come back down and look at the lateral side here on the left, then go back up on the right, and then look posteriorly, and again, we're looking for extra-axial blood. As we mentioned, because the bone is bright and acute blood is going to be bright as well, it can be very easy to miss blood layering on the bone. So we need to put on our blood window. We do the same thing. So look anteriorly first for any extra-axial blood, laterally on the right, laterally on the left, and then posteriorly, and there's no extra-axial blood here.

Then you assess for stroke. First, start by looking for hyperdense vessel. So I start in my simplified conceptual approach by looking at the M1s bilaterally, and I'm looking at the Sylvian fissure for any hyperdense vessel there, and looking at the basilar artery for a bright basilar suggesting clot there. Then I look for gray-white differentiation. So put on the stroke window and then I sit back far off my computer, or I just split it into four here so you can see the gray-white differentiation better. So in a simplified approach, you can just take a couple scrolls through and look in the ACA territory here, the gray differentiation, and the MCA territories here bilaterally, and then most importantly, we look at the insula for signs of early stroke, the putamen, the caudate, and the thalamus. A more detailed approach, you look for PCA strokes, well, because cerebellum, the brainstem, but that's beyond this right now. Then you look at the bones for any fractures or any bone lesions. We'll talk about that more in a bit, and look at the soft tissues, and to include the orbits, etc., and again, that's more in the detailed approach.

Okay, now let's practice our approach. We're going to review the concepts that we learned today in these cases. This is the first case. Let's say you're in the emergency department, you get a CT head on a confused patient, and you open it up. You scroll down here quickly, take a couple scrolls. I don't really see much. Okay, oftentimes, if you were just to leave it alone, you might miss some big things. Okay, let's go through our conceptual approach. We took a few scrolls. Now let's look for big mass effect. So all sulci look okay. There's no midline shift, there's no uncal herniation, the suprasellar cistern is patent, there's no tonsillar herniation here. Okay, I don't see any hyperdensity in the sulci. I'm looking specifically for bleeds and certain locations as well, and the interpeduncular fossa. Okay, I don't see any hyperdensity in the sulci, so I don't see any blood here. Guys, make sure I remember to put on the blood window. Put on the blood window here. You're looking for any extra-axial hemorrhage along the right side here. I don't see anything. Anteriorly over here, I don't see anything. Laterally here, I don't see anything. Posteriorly here, I don't see anything. So there's no big bleed, at least.

Let's move on to look for a stroke. To put our brain window, look at the M1s. I don't see anything. Look at the M2s. Some questionable things over here, but nothing that's definite. Look at the basilar artery. Looks okay. Don't really see any hyperdense vessels. I don't look for more subtle signs of stroke. I'm looking for loss of gray-white differentiation. So change the window to stroke window, split it into four so it's easier to see, and I scroll down from the top. Okay, it all looks preserved here. It could be ACA territory. First, it's great. Look at the MCA on the left. It's very crisp gray-white differentiation. Notice how crisp and how bright the gray matter is compared to the white matter, and look along the right side here, and I notice this kind of looks a little bit blurred. We've actually seen this case before, so I'm concerned that this area has lost gray-white differentiation, and this is a sign of an acute stroke. So again, very easy to miss if I don't go through and know how to window appropriately, and even subtle when I went and I want to do it appropriately. So even with our quick conceptual approach, we caught this. It also demonstrates the point that the hyperdense vessel can often be very subtle. So you don't really see it here on the thick slices, but when you put it on thin slices, this is again a non-contrast, but in thin slices, and we look at the M2 branches in this area here, you actually see evidence of a vessel here. Compare this. We kind of thought this is a hyperdense vessel that was not well appreciated on thick slices. So sometimes you can put the thin slices, look for it. But again, this is a subtle case. The point here is not the pathology, but the fact that they go through our quick conceptual approach, we can catch things that are more subtle, and we won't miss important pathology.

Okay, case two. This patient comes in with a bad headache and some neurologic symptoms. You take a couple scrolls through. So first scroll through, you're just looking pretty big, and wow, there's a big hemorrhage here centered in the right basal ganglia. And you still want to go through your full approach. We won't do it here for time's sake, but in this case, this is a classic hypertensive hemorrhage. Now, other things can cause blood in this area, but this patient was hypertensive, and we did vascular imaging to make sure there's no vascular malformation. The most common location for a hypertensive hemorrhage is right here.

This is the third case. This is a patient who fell, and we want to rule out a bleed. So again, a couple scrolls through, don't really see much on the first scroll through. Let's go back. Don't see much. Case rolling down, looking for subtle signs of mass effect. No sulcal effacement here, no significant midline shift/subfalcine herniation, suprasellar cistern is patent, no tonsillar herniation. Now looking for blood. Looking for some regular blood. First, looking in the basal cisterns: prepontine, suprasellar, interpeduncular, ambient, quadrigeminal. Look at the Sylvian fissures bilaterally dependently, looking at the ventricles, looking at the dependent ventricles, look at all the sulci up here to see if there's any signs of signs of blood, and there's not. And I look for extra-axial blood. So I look anteriorly first, I don't see anything here. Look laterally here. Anteriorly, see anything laterally? Don't really see much. Posteriorly, don't really see much. There's really no blood on this window. But remember to put on your blood windows. So for my blood windows, and again, I look for extra-axial hemorrhage. So I look anteriorly on the right here, laterally, again, nothing on the left here, laterally, and look, there's a decently sized subdural hemorrhage right here that was nearly impossible to see on normal brain windows. But because we windowed it, it sticks out. So again, very important to know how to window so you don't miss this. It can be bigger and still blending in with the bone if you just use a normal brain window. So this really reinforces the concept of the purpose of looking at blood windows.

Okay, then we look at signs of stroke. We look for, look at the M1s and the M2s, frontal hyperdense vessel sign, look at the basilar artery, and then look for loss of gray-white differentiation on the stroke windows and all the vascular territories: ACA, MCA is included, deep gray matter structures, PCAs, cerebellum.

Okay, case four. This is a trauma. First, take a couple scrolls through it, look for anything big, and then take our dedicated runs. Our first run again is mass effect. As I scrolled down, I noticed the sulci are a little bit tight bilaterally. There's no significant midline shift. I look for the suprasellar cistern, and I can't really see it. The reason is because this brain is effaced, with sulcal effacement and obliteration of the suprasellar cistern. As I scroll more inferiorly, still looking for mass effect, and there is fullness at the foramen magnum suggesting tonsillar herniation. As a bonus point here, there's also ascending transtentorial herniation. The infratentorial brain is herniating superiorly through the tentorium and causing effacement of the quadrigeminal cistern here.

Next step, I look for bleed. And there is scattered blood on in this brain, most importantly in the posterior fossa. It's kind of difficult to see on this window, but if I use the blood window, you can see blood layering posteriorly in the posterior cranial fossa here. This was causing the mass effect that is causing tonsillar herniation, and this patient needs urgent decompression. I finish off my approach by going through to look for signs of stroke, but there was nothing on this case, and to save time, we'll skip that.

For now, okay. So, this is the end of the introductory CT head talk. After the slide, I will demonstrate a more complete approach to CT head on a more subtle case. You can listen to this if you wish, but the focus of this talk, which is a more conceptual approach for non-radiologists, is complete.

So, take-home points: Number one, know your basic concepts of stroke, bleed, and mass effect. They were introduced here. Know them well, and always look for those three things on CT head.

Number two, understand the basic densities. You don't need to know the numbers, just know that CSF is basically water, grey matter is slightly denser than white matter, and clotted blood is denser than everything else in the brain. Do you understand the densities of the different ages of blood? It can really help you interpret scans accurately.

Three, use the conceptual approach. If you're working in the ER, especially in the community, looking at the brain a few times isn't enough. Make it a point to go look for the big things that we just discussed. Start with mass effect. Look for the sulci, the effacement, look at the suprasellar cistern, and look for midline shift, and look for tonsillar herniation. Look for blood, specifically using the blood window, and then use a stroke window to rule out strokes. Lastly, window appropriately. Understand how to window appropriately on your PACS system so you can go through your conceptual approach relatively quickly, well at the same time making it more likely that you won't miss pathology that is going to change management. We gave a couple good examples of situations where windowing well helped us identify the pathology accurately.

If you're going to go into radiology, or you're already a radiology resident, it's very important to have a good, detailed approach that looks at everything. This can serve as a good example, and you can modify it as you get more practice and see what you prefer, but this is a good starting point.

Okay, for the future radiology residents and the radiology residents out there, let's go through a more complete approach that's a little bit more detailed and looking for more subtle findings. This is a subtle case. This is a patient who came in with a trauma.

Okay, so again, similar principles to what we talked about in our conceptual approach. We'll start by looking for anything big. Take one scroll down, one scroll up. Don't see anything big. Next, we're going to look for signs of mass effect. We look at the sulci bilaterally. You look okay, no significant midline shift. Suprasellar cisterns patent with no uncal herniation and no tonsillar herniation.

Then we're going to look for signs of bleed, and we start by looking at all the CSF spaces again, but in more detail. We scroll up here, looking at all the basal cisterns, including the prepontine cistern, the suprasellar cistern, the ambient cisterns, and the quadrigeminal cisterns are all patent. There's no hyperdensity within them. We scroll superiorly and look at the ventricles bilaterally, looking for hydrocephalus or signs of hemorrhage as well. And specifically, we're looking at the most dependent portions of the ventricles for more subtle signs of hemorrhage in the lateral ventricles. We don't see any. Looking at the third ventricle, the aqueduct, and then the fourth ventricle here. Okay.

Then we're going to take a scroll through to look at the sulci specifically again for subtle locations of blood, like the posterior portion of the Sylvian fissure is here, and any subtle blood in the sulci more superiorly, and we don't see any hemorrhage.

Then we're going to look for extra-axial blood. You can do it on brain window first. A blind spot for extra-axial blood is also along the falx and along the tentorium here. So always look at the tentorium, the falx for extra-axial blood as well. Always have to remember to look at the blood window. So here we look anteriorly first, and we don't see anything. Posteriorly, next, we don't see anything. We look around along the right side here. There's no extra-axial blood. Call them blind spot is just anterior to the temporal lobes here for extra-axial blood. Similarly on the left, look in the same location, as well as along the left side of the skull here, and we don't see anything. So there's no signs of extra-axial blood.

The next thing we look for signs of stroke. And a more detailed approach involves looking at all the vasculature that we can see. So starting posteriorly, looking at the vertebral arteries, scrolling up, looking at the basilar arteries, looking at the PCAs, we see the branches off, looking at the M1 branches bilaterally, looking at the M2 branches here in the Sylvian fissure bilaterally, and looking at where we expect the ACAs to be. We can kind of see them there as well. There's no hyperdense vessel sign as well.

So now that we've looked at all their intracranial arterial vessels, we can look at the venous sinuses. Venous thrombosis can cause a venous infarct. So we want to look for hyperdensity in the superior sagittal sinus and the transverse sinuses bilaterally. Follow them down to the sigmoid sinuses and the jugular veins as well. And then we look for signs of gray-white differentiation. So we put on stroke window, zoom out, and again, look at the same location. So the ACA territory, the MCA territories bilaterally in the cortex, all the way up to the top, and then looking at the deep gray matter structures, the putamen, caudate, the thalami, looking at the insular ribbon bilaterally. Never forget to look at the PCA territory here. It's a very commonly missed location for loss of gray-white differentiation, and then the cerebellum as well. Lastly, look at the brainstem. Be aware that there is a lot of artifact going through here as a result of this, as a result of the skull base. We can catch subtle strokes if you look here. Every time.

Then I look at these soft tissues. So I'm looking at the soft tissues. I start superiorly and look along the anterior portion of the soft tissues here, looking for any signs of trauma or any lesions, and look laterally here, posteriorly, looking for any lymph nodes back here as well, and then laterally as well on the left. I don't see anything.

Next, I look at the orbits. Always look at the orbits slowly. Look through each of them, compare them for symmetry as well. I look at the retrobulbar fat, look for any haziness or hemorrhage behind the globe, look for any air behind the globe, look at the globe itself, the lens. I look at the optic nerves bilaterally and compare them, and the extraocular muscles to see if they're thickened. And the last thing I look for is the superior ophthalmic veins. I compare them bilaterally and look to see if they're dilated. They can be dilated in cases of carotid cavernous fistula or cavernous sinus thrombosis.

Then scroll inferiorly, and there are five pairs of fat that you always want to look at at the skull base. The first is the retromaxillary fat here. It's relatively clean. I look at the pterygopalatine fossa bilaterally to see if they're filled. They're relatively clean. I look at the superior orbital fissures posteriorly here. That should be clean fat as well, which it is. This is the pharynx here. I look at the parapharyngeal fat over here bilaterally for anything filling it, and they look normal here. Then look at the stylomastoid foramen posteriorly here, and the fat beneath this area here should also be clean. And look at the nasopharynx itself for any asymmetry. And then lastly, but definitely, I look at the carotid arteries beneath the skull base.

Okay, and the carotid arteries beneath the skull base can be hard to find on an individual slice. The best way to find them is to find the carotid canal itself here and scroll inferiorly from that location, and naturally, the carotid artery will be able to be followed beneath that point.

Then we look at the bones, looking for fractures or any big lesions to make your life easier. Most fractures are associated with either soft tissue changes overlying them or blood in the sinuses adjacent to them. If you see soft tissue changes, look at the bones beneath or in that region closely. If you see blood in a sinus or in the mastoid air cells, then look more closely at these bones.

With that in mind, here's my approach to bones on a routine CT head. You've already looked for soft tissue swelling. If you see any, focus or search in that region to start. If not, move on to the sinuses, starting with the frontal sinuses here, the ethmoid air cells here, the maxillary sinuses bilaterally here, and the sphenoid sinuses here. Next, look at the mastoid air cells here and on the left as well. If you see any fluid or blood in any of these structures, you need to focus on the walls and look for fractures closely.

Once you've done that, there are a few checkpoints that you need to look at. Look for calvarial fractures anteriorly, laterally, and posteriorly. Scrolling down from the top, make sure you follow each of the bones down to the skull base. So when you scroll, when you're looking at the lateral bones, make sure you follow it down to the greater wing of the sphenoid and the skull base bilaterally. When you're following it down posteriorly, make sure you follow it down to the posterior cranial fossa until you reach foramen magnum.

Then I look at the orbits, specifically looking at the lateral orbital wall superiorly, and then the medial orbital wall, and I look at the other side, looking up the medial wall and then down the lateral wall. Then look at the zygomatic arches bilaterally. Look at the pterygoid plates bilaterally. All the fractures or the four-part fractures involve the pterygoid plates. Then look at the nasal bone. The nasal bone is the most commonly fractured bone, so always look at the nasal bone. Make sure you look at the clivus, always, always, always look at the mandibles. Make sure they're congruent and then follow them down as far as you can. You can always see part of the spine. This is C1, it looks okay, and then you can also see C2 and the dens here as well. So always look for that. You can see subtle signs of fractures there as well.

So that's it for a more detailed approach. There are always other blind spots and things you can look for, but that's a similar detailed approach to the one that I use. This case was subtle, is pretty much normal. There was nothing abnormal in the brain that was traumatic. However, if you notice when we're looking at the soft tissues beneath the skull base, I didn't mention at the time all the pairs of fat looked okay, but if you look at the carotid artery on the left here, it looks okay. We've looked at the carotid artery on the right. Notice the hyperdensity of the walls here as I scroll down. It's more obvious here. As we know, acute clot or acute blood is bright. This is a case of a dissection. When you have an acute dissection, you get acute mural hematoma or an intramural hematoma, and this is the most subtle sign of an internal carotid artery dissection. A CTA was done thereafter and confirmed that this was, in fact, a dissection. An interesting case with no other findings except for this.

Okay, that's it for this talk. Thanks for listening. I hope it helped.