Neuro-ophthalmic complaints can be some of the most challenging and high-stakes presentations in eye care. Knowing when to order neuroimaging, what modality to choose, and how to interpret the results can make the difference between timely diagnosis and missed pathology. In this case-based webinar, Dr. Mackay, our neuro-op Cybersight Mentor, will walk through real-world cases and break down the clinical decision-making behind the neuroimaging for each case. Participants will learn practical indications for imaging, how to select the appropriate study (MRI vs CT, contrast considerations, etc.), and how to systematically review imaging findings relevant to common neuro-ophthalmic conditions. Join us in this investigative webinar as we provide clarity and confidence in neuroimaging skills. (Level: All)
Lecturer: Dr. Devin D. Mackay, Ophthalmologist, Indiana University School of Medicine, USA
Transcript
DR. MACKAY: Hello, everyone. Welcome to our international audience. So, my name is Dr. Devin Mackay, I’m the director of neuro-ophthalmology at Indiana University. This is going to be a very practical, and hopefully useful talk for you today. Talk about neuroimaging and neuro-ophthalmology. MRIs, CT scans, when to use them, order them, use them, and how to use the imaging for yourself. This will be hopefully very practical. I have no relevant disclosures for this talk. Our objectives today. Try to apply the principle of localization when interpreting neuroimaging. Localization is trying to figure out where in the nervous system the problem is in this case with regard to the vision. And we’re gonna learn how to select appropriate imaging decease, understand common MRI sequences and systemically review neuroimaging in neuro-ophthalmic patients and apply these principles to real cases. Why does neuroimaging matter? Has to do with high-stakes diagnoses, compressive optic neuropathy, we don’t want to miss, stroke, aneurysm, intracranial hypertension, tumors, demyelinating disease, and when we image, we can have certainty by excluding mimics. And monitor for progression or change, especially when following tumors that are affecting the visual pathways. One of our jobs and one of the messages that I really hope that you’ll take home from this webinar today, I want you to think like a neuro-ophthalmologist when reviewing imaging, ordering it and interpreting it. And the key is localization. Where is the problem that’s causing the patient’s symptom or examination findings? And we always focus on where before we focus on what. And that’s because the whats are different for any given where. That’s a lot of — a lot of kind of weird words there. So, let’s think about what that means. So, the whats are meaning what kind of diagnosis? Is gonna be different based on what part of the nervous system we are interested in. So, for example, if we know that an abnormality localizes to the optic nerves, we’re gonna be excluding diseases that are only found in the retina or they’re only found in the peripheral nerves or things like that. So, our first job is to really figure out where could this problem be coming from? And the list of different diagnoses that will apply that to location will be unique. That’s the first job. Figure out where is the problem and then what? And then, of course, when we know where the problem is, we also know what part of the nervous system to image and talk about scans and protocols and get into these details. So, a practical approach to neuroimaging I think would include the following. As with any patient and valuation, the history and examination are extremely important. And that is really what leads us to our localization. So, based on the history and exam, then we figure out where is this problem in the nervous system? Then once we know where the problem is, then we can select the best imaging study. The best protocol, figure out whether it’s with or without contrast. Know exactly what body part needs to be imaged. And then, of course, we really need to review the images ourself. So, radiologists are very competent and very well-educated, however, they do have one disadvantage compared to us. And that is that they are unable to examine the patient and interview the patient. So, that first part, history and examination, are so important because we are the ones that can do that. The radiologists cannot do that. Based on that, we can figure out where the problem is, and pick up things that radiologists miss, not because we’re better or smarter than them, but we have the ability to interview and talk to the patient. How do we know when we need to order neuroimaging? When is that appropriate or necessary to do? And it’s really important when red flags are present. And we’ll talk about what some of those are in just a moment. Also, if there are any signs or symptoms that localize to the optic nerve, the optic chiasm, the retrochiasmal visual pathways, the brainstem or the orbit. Areas that we can localize with neuroimaging. And also have a reasonable expectation that we would find something that would change how we care for the patient. So, if it wouldn’t make a difference in the patient’s care, then we need to rethink, do we really need to order this neuroimaging study? These are guidelines how and when it’s appropriate to order neuroimaging. I talked a little bit about neuro-ophthalmic red flags. These are a reason for imaging. Let’s go over what some will be. A new visual field defect, sometimes an indication for neuroimaging. An un-complained rapid pupillary defect, a optic disc edema. Not all of these findings always need neuroimaging, but these are the kinds of cases to consider it in. Unexplained optic atrophy, especially when looking for a compressive lesion as a cause. Painful vision loss that doesn’t have an obvious ophthalmic cause, diplopia with neurologic signs turn explained diplopia. Horner syndrome signs, especially with unilateral miosis and ptosis. That’s what’s shown in the paragraphs. And cranial neuropathies, as well as, of course, progressive symptoms. It’s worth mentioning as well, there are risks of overimaging. It might seem simple to, well, if we don’t know what’s going on, image everyone. Get an MRI of the brain or the orbits and that should solve our problems, right? Well, not exactly. There’s a big risk of finding incidental findings. And incidental finding are findings that are present, but maybe have nothing to do with the clinical presentation or not a sign of disease or abnormality that needs to be pursued. However, many incidental findings end up leading to further testing and investigation and sometimes even interventions that can be unnecessary and actually bring harm to the patient. And it’s a significant risk of iatrogenesis. We can actually harm the patient sometimes by previously ordering neuroimaging when it’s not indicated. It’s burden? To the patient and the health care system. It’s costly. It ends up consuming resources that can be used for other things. Some examples of some things that may be incidental finding that sometimes get over-evaluated might be a mild asymptomatic Chiari malformation, a pineal cyst, which is what we see in the image on the right. Obviously the arrow pointing to this cystic structure. This is a pineal cyst, often benign. Arachnoid cysts are often benign as well. And a lot of assigns, meningiomas, not pressing on anything important involving the skull base or the brain, those can be examples as well of an incidental finding that doesn’t need further evaluation. There was a study called the Rotterdam Scan Study, which looked at the incidence of incidental findings and found that 9.5% of patient who is had a scan had an incidental finding. And 3.2% referred to the specialist, and 76.6% of those were observed or discharged from the clinic. In other words, the specialist didn’t need to do anything in those cases. We’re finding, again, a lot of healthcare over-utilization in some of these cases again, with respect to over-imaging. So, we want to make sure that we order just when it’s appropriate to do so. And on the flipside, there’s also risks of under imaging. We can miss a vision or life-threatening condition. False reassurance from a normal study, or a study that’s read as normal when it really isn’t. For example, a head CT has about 16% sensitivity for posterior fossa infarcts. That’s infarcts of the brain stem and cerebellum. Only 16% detected on the CT scan. It’s reassuring, everything is fine. Both over and under-utilization of neuroimaging can be harmful. Today I would like for us to consider when is it appropriate to for us to pursue neuroimaging? And how do we interpret those results? So here are some situations where neuroimaging may be low yield. One is with typical migraine aura. If it sounds like a migraine aura with jagged edges, comfortable, scintillating scotoma, and grows or shrinks over time. That’s a typical migraine aura and probably doesn’t need neuroimaging in most cases. Longstanding stable congenital findings, those may not be necessary. And improving classic microvascular cranial nerve palsy. And anything that’s clearly an ocular pathology that explains the symptoms. In a lot of those cases, we really don’t need neuroimaging. Now that we have that out of the way, let’s say that we find that it’s appropriate to order neuroimaging for a patient, how do we know what study we should order? We’re gonna focus just mostly on MRI and CT scan. I realize there are other neuroimaging modalities, but we’re gonna focus on these two as some of the most helpful for the diseases that we’re going to see. MI is best for soft tissues. And CT scans are best for bones and blood. MRI — soft tissues that would be helpful to evaluate with an MRI would include the optic nerve, optic chiasm, the brainstem, looking for any demyelination in the system, and tumor characterization. Those soft tissue issues. CT scans for trauma, acute hemorrhage, looking for fractures of bones. Emergency imaging, CT scans are easier and cheaper and faster to get, and it can also look if I can calcification well. And of course the study, we need to decide does it need contrast or not contrast? Things that help us decide that. You often do need contrast with a possible neoplastic or inflammatory condition, demyelinization, meningeal disease, or an MRA or CTA with blood vessels, that’s helpful in those cases. And without contrast, likely acute ischemic stroke, low likelihood of pathology and doing the scan to rule out things that are not super-likely. And if there’s a contrast allergy or contraindication and some other way to contrast. So, I’m not gonna bore us too much here, but we need to understand some very basic things about MRI before we get into more details. An MRI, the principle here is that the body is placed in a strong magnetic field. And that strong magnetic field ends up aligning the hydrogen protons abundant in water and fat. They will look kind of like this, all oriented within that magnetic field. And then a radio frequency pulse is applied and that tips these protons out of alignment. And then after that pulse, the protons relax back to their original state. And different tissues have different relaxation times, and the MRI coil is able to examine that and can give contrast to T1 weighted images. And by contrast, I don’t mean intravenous contrast, I mean, the one contrast where the signal looks different from another in terms of its signal intensity. And the reason this is important and we’re going over this, every sequence on the MRI, different sequences, they measure a different tissue property. So, we need to think about that based on what decides we’re looking for. And we can see what sequences might be appropriate to look for those specific properties. Those are based on the different properties of the soft tissues. now a few words about contrast. CT scans use iodinated contrast. Iodine-based. And con Infrastructure-induced nephropathy. MRI uses forms of gadolinium. And allergic reactions are less common with the gadolinium contrast media. Some types, not all, are associated with nephrogenic systemic fibrosis in patients with severe renal disease, causing severe — and normally contrast separated from the brain and optic nerves by the blood-brain barrier. It keeps the contrast within the vessels. When contrast leaks out, it tells us that there’s an impairment of the blood-brain barrier. Happens with tumors, infections, and inflammation. Those all affect the blood-brain barrier. And if there’s a contrast imaging study, there’s a leakage of contrast, abnormal contrast enhancement, think about these, tumors and inflammation especially. Now let’s understand a few MRI sequences before we apply these to actual cases. Some of the sequences we’re gonna go other will include T1 pre- and post- contrast images. Something called FLAIR, a sequence. DWI and ADC, explain what those are. STIR, T1 fat suppression which subtracts fat so you can have a better view. And most are available in different views. Axial, a slice through the head this way, coronal, slice through this we way, and sagittal, which is a slice through this way. So, let’s look first at T1 pre and post contrast and what do we use those for? The T1 measures how quickly protons realign with the magnetic field after excitation. And different tissues have different relaxation time and give the different signal intensities. This tends to be really good for looking — and also for post-contrast enhancements. We can see, are there areas where the blood-brain barrier is dysfunctional? On T1 weighted images, fat is bright, and fluid is dark. Which is the opposite of T2-weighted images. So, fluid is dark here. So, for example, fluid within the ventricles will be dark, fluid in the cisterns around the brainstem,al be dark. And here, this is a post-contrast image. And you might think, well, there’s a lot of fat in the orbit, why isn’t that bright? That’s because there was fat suppression here, go over in a little bit. But what pointing out with the are a resource, there’s here on the little bit of brightness there that’s not there on the other side. That’s a clue that there’s some abnormal contrast there, and tells us, again, we want to suspect things like inflammation or tumor or infection. And here we are in a coronal view. And same thing, see brightness of the pre-chiasmatic optic nerve and not on the other side. This is a case of optic neuritis that was acute. And a T2 FL AIR-based sequence. Think of it like a T2 sequence minus the CSF, it’s interacted for you. FLAIR stands for fluid-attenuated inversion recovery. And it’s a T2-based sequence that nulls the signal from flee-flowing CSF. Myelin and fat are dark. And great for looking for edema and inflammation or gliosis, which is scarring. And notice how that’s brighter than the sub-cortical structures. Myelin and fat are dark, that’s why where that area of the skin is darker. And the arrows, it’s excessively bright. That’s edema. It’s fluid, and fluid is bright on a T2-based sequence. We can see there’s something edematous here in the right occipital region. This was a patient who had an acute ischemic stroke in the right posterior cerebral artery territory and that led to this finding. Fat saturation. There’s a few different ways to look at that. One is with a STIR sequence, which stands for Short Tau Inversion Recovery. It’s like a T2 sequence minus the fat. It’s great for looking in the orbits. Here is an example of a coronal T2 STIR. And one of the pearls to know is that the optic nerve has a lot of myelin in it. It’s got a lot of fat. It should be the same as the myelin in the frontal nerve. See how dark it is and the nerve? That darkness is the same, it’s normal. Notice the left optic nerve, it’s a little bit brighter, not as dark as the frontal white matter. That’s signs of increased T2 signal in the optic nerve. You could have it from edema, optic neuritis, or prior damage that’s not active as all. Those can all cause brightness of the nerve. This is not contrast enhancement, this is T2 hyper intensity. This is a great sequence for looking in the orbits, it subtracts out the orbital fall and you can look at the opt ecosystem nerve. T2 with fat suppression is a sequence, known as fat saturation. Here is the T1 sequence without fat saturation. Notice how bright that is in the eye socket. That’s a problem looking for signal in the opt ecosystem nerve as well. One way to get around that, on the T1 post-contrast images, right here you can see that the fat has been subtracted away. It’s not bright anymore. This is the lateral rectus and the medial rectus muscle, and those normal lip enhance. And we can see the optic nerve without being subtracted by the fat. This is the image that we saw earlier with the left pre-chiasmatic optic nerve is inflamed. And this is great for looking for abnormal enhancement within the eye socket. That’s usually a feature of MRI orbits protocols. You should get that. DWI and ADC are helpful for looking for acute ischemic stroke. It stands for diffusion weighted imaging and apparent diffusion coefficient. I won’t go over what those mean. But these measure random motion of water molecules, and when it’s random, it’s dark. And with acute ischemic stroke, you get brightness. That’s what you see here with the stroke in the right occipital lobe. That’s from cytotoxic edema that restricts that diffusion of water. And then on an ADC image, that should be dark in acute ischemic stroke. And the DWI portion should be bright. DWI bright, ADC, bright, you know that’s an acute ischemic stroke. And I alluded there are different ways to look at the sequence in different planes. Axial, coronal, sagittal views. Sheer a coronal, good for optic nerves. And the sagittal, the infarct we were talking about earlier in the occipital lobe. And it sometimes helps to view an abnormality from different perspectives. And you can see if a mass is contacting the visual pathways or look for other details. So, there are some common ordering mistakes that you want to be aware when have ordering a scan. You want to make sure that you don’t scan the wrong area. Another mistake is not giving the radiologist enough information. And again, even when you’re ordering, you want to think like a neurologist or a neuro-ophthalmologist, localize first, then decide what scan and protocol you need. If you’re not sure, you can always call the radiologist and ask them. You want to describe the suspected localization in your requisition for the radiologist. Here’s an example of a radiology requisition. Say you’re concerned about optic neuritis potentially and order an MRI of the orbits with and without contrast. And you want to see a coronal STIR sequence. You can say that. You can specify if there’s certain sequences you think you need, you can let them know that. And Rene Deida reason for exam, that’s where I put the relevant clinical symptoms. Here, a right optic neuropathy. Lets the radiologist know, that’s where this localizes. Please pay special attention to that area. The question always comes up or often come up of do I need an MRI of the brain or the orbits in this particular season? It’s helpful to know the differences between a brain protocol MR I and a orbital protocol MRI. The brain has a larger field of vision and lower spatial resolution, the orbits is smaller field of view and higher spatial resolution. The brain has thicker slice thickness, the orbits thinner. The fat suppression — or co-al imaging in the orbital, and brain may or may not include that. Fewer sequences in the orbit because we’re usually not looking for acute ischemic stroke and things like that there. To summarize briefly, MRI orbits, higher resolution, fat suppression, and thinner slices, and that can be useful in certain situations. CT angiogram, MRA and MRV are things we can order. Identify vascular pathology. Aneurysms, heaven us identify if aneurysm is compressing the third nerve and causing a palsy. That’s a common indication to get neuroimaging with the vessels. Horner syndrome. Stroke or TIA, make it necessary to look at blood vessels. And venous sinus thrombosis, looking at the veins. If I were to assemble a cheat sheet for you to figure out, okay, well, for certain clinical problems, what kind of imaging? This may be a helpful reference. Optic neuritis, MRI brain with orbits is helpful. Chiasmal syndrome, se LHC a region, pituitary. CN, brain and CTA. And pap Pi edema, and MRI of the brain and head. And MRI of the brain is test of choice for homonymous hemianopia.
Now how do we read the scan? A redding strategy may be to first confirm the correct study in the correct patient. There’s times when I’ve looked at a debris and start reviewing it and realize, oh, whoops, this is the wrong patient. Make sure you have the right patient and study. Always think, based on my history and my examination of this patient, what am I looking for? And more importantly, where? Where is the problem gonna be? Or likely to be? And then select the sequences to review that you want to review. And then focus on the area that you’ve localized, or if you don’t have a specific localization, you may just try looking through the visual pathways generally. For anything that you’re evaluating where there’s bilateral structures, always look for symmetry with the normal side. That applies to the optic nerves or the optic tract or the occipital lobe. Those are paired structures. If up an abnormality on one side, look for symmetry. And read the radiology report and see what they have to say. I think it would be helpful for you to accept a challenge here for me today. And that is for you to on your own time review some neuroimaging studies and try to trace the visual pathways through the study on your own. And the goal there is to practice identifying things like the optic nerves, the optic chiasm, the tracts, the occipital lobes and try different views, axial, coronal, and sagittal. If you can do that on normal patients, that will increase the ability to do that effectively with an abnormal patient. And all these different visual field pathways have different manifestations in terms of the abnormalities. That’s one of the keys with local fields. This is an example of someone tracing the visual pathways on an MRI. There are some common imaging pitfalls to be aware of and to try to avoid. One is to — is missing subtle enhancement. Subtle abnormal contrast enhancement can sometimes be missed. You can minimize that knowing where to look and what kind of disease you’re looking for. And a motion artifact, making it blurred and harder to see the relevant abnormalities. Don’t assume a normal report equals a normal scan. We can interview the patient, and localize, and sometimes find I think this is that radiologists might not know about. And on those same lines, blindly trusting the radiology read can sometimes not, the right approach, especially if we have additional information not available to the radiologist. And, of course, ordering the wrong protocol. If we order the wrong protocol, we’re limiting our ability to evaluate the structures in question. And, of course, incidental findings. Incidental findings can be a problem. Don’t over-image, use imaging where it’s appropriate. All right. Let’s try to apply these principles to actual cases. These are real cases that I’ve seen. This is a 28-year-old woman who presents with decreased vision in both vie eyes. She has pain with eye movements that started two weeks ago. And four days after that, noticed decreased vision in both eyes with sensitivity to light and difficulty seeing colors. I want you to pay attention here because we’re gonna ask a question to the audience, where does this localize? As we’re presenting the case, think about where does this problem come from? And the vision continued to decline in both eyes over the days. And the MRI of the brain and orbits was read as normal bit radiologist. The parable saw an optometrist who found hazy optic disc margins and diagnosed nonorganic visual loss. There was visual acuity, pretty clear. Diminished color vision, sluggish pupillary reactions in both eyes. And was able to count fingers in all quadrants and had some mild disc edema in each eye. Think about where does in localize? Here are the visual field test results. So, we see some generalized depression in the right eye, and the left eye, central depression that extends into the mid-periphery as well. And on OCT, see some evidence of disc edema that’s fairly mild in both eyes. Get ready to respond here. Where do you think this patient’s vision problem localizes? Is this bilateral retina? Bilateral optic nerves, bilateral optic tracts, bilateral occipital lobes or bilateral visual association cortex. That poll is up there, and leave that up for a little bit and allow everyone to answer. Okay. And most respondents responded bilateral optic nerves which is the correct answer. That’s going to focus when we’re ordering the neuroimaging and looking at it, we’re gonna focus on that area, on that location. Let’s do that. Here is T1 axial post-contrast images on the top left, and the bottom left, T1 coronal post-contrast. It is subtle, but there is some signal of the optic nerve. Right here and here. What it should follow look like is the brain here, kind of dark, no contrast at all. Here it kind of lights up a little bit. Even though it’s subtle, it’s there. This was missed by the radiologist. Probably because it was bilateral. It looked the same. With the clinical history, this is optic neuritis, a nerve abnormality. Seems to be in both eyes. The T2 coronal images show increased signal of the optic nerve, sign of edema or prior damage. Since there was no prior damage, fits with edema, acute optic neuritis. And notice how dark it is, same darkness as the frontal white matter. Here it’s clearly brighter, so you know that’s abnormal. And fortunately, she was undiagnosed with non-organic vision loss, and treated for opt ecosystem vision loss. She was left with a little bit of optic atrophy and did well over all. And she was grateful we were able to localize and review the images ourselves. And case two, 62-year-old man with decreased left eye vision. Noticed decrease left eye vision three years ago, had optic disc edema in the left eye. And MRI brain with contrast read as showing posterior left optic nerve thickening with consistent with optic neuritis, how it was read. There’s a left affect pupillary defect, and constriction on visual field testing with mild optic disc edema in the left eye. Thinking like a neuro-ophthalmologist and we want to localize first. This is obvious, this is a left optic nerve problem. Think about what can this be? Does this fit left optic neuritis? Going on for three years with optic disk edema over three years? No, it wouldn’t. We have a reason to suspect that the radiologist read might not be entirely accurate because we can take a history and exam from the patient. So, we get this T1 coronal post-contrast image, showing the enlarged optic nerve sheathe contrast here that’s light. This is contrast-enhanced. And coronal image, showing brightness in the T2 weighted sequence. Things like edema can do that. And here on the T1 axial, we see the area, it’s very well demarcated. And now thinking about this doesn’t fit clinically with optic neuritis. What can be there for three years and cause a little bit of disc edema and have this focal area of enlargement that’s been there for a few years as well? Think about it, an optic nerve sheathe meningioma. That’s treated differently an optic neuritis. Part of this is we could review the images on our own and armed with the history and examination. Now case number three, we have a 70-year-old woman with bitemporal visual field loss. And this vision loss started six months ago, and an MRI of the brain was done at the time and was read as normal. So, the patient was diagnosed with migraine, but the symptoms seemed to have progressed or at least didn’t really go away. So, visual field testing showed this. And so here we see a bitemporal hemianopsia that’s pretty dense. And then on the OCT, I took a snapshot from the ganglion cell analysis, which remember, these are the cell bodies of the axons that are forming the nerve fibers in the opt ecosystem nerve. If I draw a imagery line through the center of the macula, the fovea, we see that there’s nasal thinning in the left eye more than the right eye here. When we see that pattern, there’s just nasal thinning and you have the temporal field defects, that localizes somewhere very, very specific. And you guys may be aware of where that location is. And to not give it away here, I’m just gonna ask us: Based on your own thoughts about where that localization is, where does this problem come from, what would be the most appropriate imaging study to order in this case? So the audience response question is up. And please select the most appropriate answers. So, it would be a CT of the head, a CT of the orbits, an MRI of the brain, a MRI of the orbits or a MRI of the sella/pituitary? All right. And most-answered sella and pituitary. That would be appropriate. We notice that this localizes to the cerebellum. In the brain, it might get some of the area that we need, but certain a better protocol is MRI sella and pituitary. Remember, this MRI of the brain was read as normal. And see this circular structure right here that’s dark. So, it’s the same density as the CSF. And here we are some other views. Here’s a coronal T2 image. What I’ll point out on the normal image over here of it, this is a T2 coronal that’s normal, notice the optic chiasm right here. Look at its shape. And then notice this one. This optic chiasm doesn’t have the same shape, does it? It’s bowed upward. And it’s bowed upward in the same region that the circle. And you can see the circle that’s bright and it’s the same as the CSF. There’s something there and presses on the chiasm. And the T2 coronal flair, and you can see the dark circular thing that seems to be disforming the optic chiasm. You can see how the radiologist misses it, it looks like the CSF. We know where the problem localizes and is expected. This was a pituitary cyst that was compressing the chiasm, and because we were able to read the images independently and know that the MRI was not normal, this patient was able to get surgery and had improvement in her vision afterwards. The surgery took care of the cyst, drained it and took it out. It was a great outcome that was not possible if we just trusted the radiology read. Case number four. A 83-year-old woman with a left homonymous hemianopia. Noticed about six months. Her husband tried testing visual fields and found it. And there was no microvascular changes and no acute abnormalities. Here is the visual field. This is a pretty obvious homonymous hemianopias are on the left side of the brain. This is a problem in the right retro chiasmal. It could be the occipital lobe, any of the locations. But there should be something there. The MRI was read as normal. Do we agree with that? Read it ourselves. Here is a T2 FLAIR axial image. As we compare the occipital lobe on the patient’s right, which is the left side of the image, and the patient’s left, which is the right side of the image, we notice there’s a difference between the two. Paired structures, we look for symmetry, and it’s not symmetric. This is normal on the left, this is not normal on the right. See the dark space here, we also see some bright signal around it. This is damage to the brain that’s happened from something and in this case, it happened to be in an ischemic stroke, and we’re catching it months later. Doesn’t have the characteristics of an acute ischemic stroke, but it’s left behind brain damage that was detectable. The MRI was read as normal by the radiologist. We knew where to look and knew localization. With the images, we know that’s not normal. There’s some damage that’s there. And now case number five. So we have a 58-year-old man with a history of hypertension. And he is eating dinner as a restaurant and his right eyelid started switching. And around the same time, he noticed an aching pain around his right eye. The next morning his right eyelid of the drooping and had pain radiating to the right cheek and ear. So, on examination, we find that he has anisocoria. To find out which is the abnormal pupil, we examined him in the light and the dark. And in the light, he’s holding his upper lid up here. We can see the eye better. We see that the right pupil is smaller than the left pupil. And then in the dark, the left pupil is smaller than the right pupil, but greater in the dark. That tells the smaller pupil is abnormal. That makes sense, there’s ptosis and mitosis on that side. You may know where this is going. Now after the administration of apraclonidine eye drops, we find that it’s reversed, now the right pupil is larger than the right pupil. We knows that an oculosympathetic palsy, Horner’s syndrome. Results from damage to a very long pathway in the nervous system. And to know how to order neuroimaging for these patients, understand a little bit about the anatomy. This oculosympathetic path aye is a 3-neuron chain. And a lesion anywhere along that 3-neuron chain can cause a Horner syndrome. Starts in the hypothalamus, descends through the brain stem and the cervical spinal cord. Exits where the cervical and thoracic spinal cord meet, and ascends in the cervical ganglion, the chain, and the second synapse. The first synapse was here in the spinal cord. And then the fibers climb on the internal carotid artery into orbit. That’s where it localizes. We don’t know where in the pathway, but we need to look at the pathway with imaging. Remember, we have our history and exam, confirms a right Horner syndrome. And the localization is somewhere along that pathway that we just talked about on the right side. And then figure out what studies to order for this. So, now — so an acute painful Horner syndrome should prompt immediate imaging to exclude which of the following? A cavernous sinus thrombosis, a carotid artery dissection, pituitary adenoma, giant cell arteritis, or an orbital mass? All right. And most answered a carotid artery dissection, which is correct. So, any time you have a painful Horner syndrome, always think about carotid artery dissection as a can’t miss diagnosis. And here is what this patient’s dissection looked like. And this is an MRA of the neck. We see a hematoma within the blood vessel wall that’s partially compressing and occluding the lumen. This is part of the stroke — I’m sorry, not a stroke. But this is the Horner’s syndrome was a result of this dissection. And here is a coronal view. Here is the left internal carotid artery I’m following with the cursor here. Notice how wide opening and smooth, and follow the vessel on the right side and notice how ratty it is, how thin and kind of irregular, and becomes normal right here again. That section right here, that’s all been — that’s an arterial dissection. And that increases the risk of stroke, and that’s why it’s important to find. So, Horner syndrome imaging, what do we actually order? An MRI of the brain, we’ll cover that part from the hypothalamus down through the brain stem and cover some of the portions where the pathway then enters the eye socket. An MRA or CTA of the head and neck is helpful if carotid dissection within the past month is possible. I say within the past month, because 82% of strokes happen within the first week, and virtually all of them happen within four weeks. If we know that this Horner syndrome is there for three or four months, I’m not worried about dissection, it would have healed anyway. And MRI with cervical spine with contrast looking at the cervical cord and through the neck. These three imaging studies are helpful in patient with Horner syndrome to adequately look at the area that was localized. And one more case, a 24-year-old woman presents with one month of headaches and transient dimming with vision in both eyes for seconds when bending over, coughing, or sneezing, gained 40 pounds in the past three months and noticed a whooshing sound when laying down. And gets an MRI and has findings associated with idiopathic intracranial hypertension. The most sensitive sign for IIH is transverse sinus stenosis, which we see here. See the narrowing of the distal sinuses. It’s 84% sensitive and 95% for IIH in pooled meta analyzes. And you can see the protrusion of the optic nerve head in the globe. You can see increased CSF around the optic nerve. An empty sella turcica– cerebellar tonsillar ectopia, this part of the cerebellum protrudes below the foramen magnum. That’s associated with IIH. And the posterior scleral flattening, and other signs as well, but these are some of them. However, not everyone with radiographic signs of IIH actually has IIH. So, one thing I want to emphasize here, is there was a study done — of these signs of IIH, and it tends — it seems that radiologists have really caught on to these signs and reporting them frequently. Do we need to be worried about IIH when a radiologist reports the radiographic features? The study was great, looked at 300 patients with an MRI for any reason. And a masked neuroradiologist was asked to look for different signs of radiographic hypertension. And fundus photographs to screen for papilledema and one in three had an empty sella. That’s not specific for intracranial hypertension, it’s very, very common. And only five of the patients had papilledema, two had a brain tumor, one had an encephalocele. And if you had hypertension, more likely to have papilledema, but if there was only one sign of IH, the risk was low, only 3%. Signs of intracranial hypertension does not automatically mean you have that condition. I recognize that many of you may in settings where neuroimaging resources are limited. What do you do in those cases? One is to reassess the need for the scan. Do you really need an imaging study here? And I would prioritize the clinical exam and the localization, and figure out from there, how important is it that I get a scan in this patient? And may save neuroimaging for cases that need it the most, visual field defects, afferent pupillary defects that don’t have an examination, abnormal optic disc findings, a pathology that’s observed. And maybe have access to different modality, maybe X-ray but not MRI, or maybe CT scan but not MRI. And think about what modalities you have and use those. CT is still helpful to exclude large intracranial masses and evaluate for orbital trauma. Just because you can’t get an MRI doesn’t mean you can’t do important imaging. And think about the observation of the clinical symptoms be a alternative? If someone is not worsening and the deficits are mild, with maybe it’s okay to observe even without neuroimaging. Those are things to think about. And also are the deficits chronic and nonprogressive? If so, maybe don’t need imaging. Here are ten neuroimaging pearls that I’ll display on the screen real quickly. Many of these we have gone over and this just reiterates them and puts them in one slide for you. I’ll especially emphasize two and one, localization. Know where to look based on the history and exam. And always review the images yourself, don’t assume that the radiologist picks up all the abnormalities there and here in conclusion, a smarter approach to neuroimaging. That’s again focusing on the history and exam. Based on the history and exam, localizing where the problem is. And then selecting the best imaging study for that localization and that disease that you may be suspecting. And then, of course, review the images yourself and the report from the radiologist. And with those, you’re well-equipped to use neuroimaging to help your patients and to be a partner with your radiologist instead of relying on them completely. Always do them in that order. All right. And I’ll entertain any questions now. So, to do that, I’m gonna pull up the — the Q&A. And if you have any questions that you would like to ask, feel free to chime in, in the Q&A there. So, one question is that has come up is how long after brain tumor surgery should papilledema progress? That’s a good question. If the intracranial pressure — it depends on how severe it is. And there’s cases that take weeks to improve, and others that take a few days with more mild edema. Depends on the severity, but days to weeks depending on that. And then in case one, is this MS-related or MOG or NOM spectrum disorder. This was one of those. This was a case of idiopathic optic neuritis I showed in case one. Didn’t have a MOG or MS diagnosis with that. And question came up of regarding the pediatric age group, do the same rules apply? In terms of the neuroimaging rules and the framework that we talked about? Absolutely. All those same things apply. So comment here about the NHS in the UK advocates for imaging only atypical optic neuritis, not all cases. Wow, I completely disagree with that. And that is because it — the MRI of the brain, when someone has acute optic neuritis, it the best way to look for evidence of multiple sclerosis. If you miss multiple sclerosis, you’re not treating it, and that means that a person is accumulating more disability. I would find that unacceptable. Especially typical optic neuritis will behave that way. Associated with MS in many cases. I’ve had many cases where we’ve diagnosed optic neuritis, no idea they were going to have multiple sclerosis, we would have missed it without the brain MRI. Great question, when do you image cranial nerve palsies, third, fourth, or sixth nerve palsies? All of them, if it’s not acute, if it’s been there for a long time, progressive, static, or improving. If it’s improving, you might not need — if it’s very clearly improving, might not need to image. Static or progressive, it — progressive, it’s mandatory to image. If it’s static or not changing, then you kind of make your own clinical judgment there. But in any acute fourth, third, or sixth nerve palsy, it’s appropriate to image them with the MR I and make sure there’s not aneurysm or a tumor or some other pathology. What is your standard workup for idiopathic optic neuritis? If it’s a typical case, get an MRI of the brain and orbits, and confirm the — and that’s all I do for a typical case. If it’s atypical optic neuritis, then ordering antiacquaporin antibodies and anti-MOG antibodies, and may look at additional testing, it depends on the clinical question and course. And how to differentiate — thrombosis on the basis of MRV. Sometimes that’s difficult. Sometimes you can see, is it — is the question — is the sinus in question, is it uniformly narrow throughout the course? If so, maybe that’s more of a hypoplastic issue. Also, if you can get your hands on prior imaging studies that might answer that question, that can be helpful. Also see if it’s possibly a fully occlusive thrombus, do they have symptoms associated with that? Papilledema, I think something that makes me think about venous thrombus — and what is the sensitivity of DWI images of retinoblastoma? I don’t know the answer. That’s a Google search and see if someone has published on that. Suspected retrobulbar hemorrhage, which imaging modality? CT or MRI. So, a CT of the orbits will be great for looking are a acute blood. If this is a recent retrobulbar hemorrhage, that can be helpful. And the blood will show up as bright. And so that is maybe the most reasonable study to do there. If it’s — if it’s maybe older blood or if it’s — if for some reason the CT scan was done and didn’t show it, you might think about MRI and see if it can show it. But acute blood, CT is often a good bet for acute blood. So, precautions before sending patients for MRI, there’s a screening form that patients need to complete to make sure they don’t have a metallic object in their body and other criteria to make sure that the MRI is safe for them. That’s the only precautions is that checklist. What are the biggest red flags in a neuro-ophthalmic examination that should prompt urgent imaging? That was one of the slides in the webinar. I’ll have you refer back in the webinar and that will help — that will help answer that. There we go. And then when are you happy to call it idiopathic optic neuritis? So I’m happy to call it idiopathic optic neuritis if the MRI doesn’t show any signs of another disease like multiple sclerosis or anything else. If it’s a typical cause of optic neuritis that behaves the way it should. Painful eye movement gets better over time, mild to moderate visual impairment — all the typical optic neuritis kind of features. And does pediatric optic neuritis need repeat MRI with recurrent optic neuritis episodes? And if it’s isolated in idiopathic can we diagnosis clinically and start ONTT? If a repeat MRI helps you manage the patient, sometimes you don’t if there’s reoccurrence or not without an MRI. It can be reasonable, especially in kids. It’s a bigger deal to get an MRI in kids. If they’re really small children, may need sedation. That’s more risk. Balance that. Am I going to get information that’s really going to help me here. Yeah. And then when would a CT scan be better than MRI for neuroimaging study? Have you refer back to the slide that answers that. So, resources for learning how to interpret neuro images. There is some websites that are dedicated to that. I know there’s a — one called Radiopaedia. And that can be really helpful for looking at examples of certain disease processes with MRIs and CT scans. So, that would be a good resource. There are probably others as well. But there’s really no substitute for you looking — at any patient where you have ordered a neuroimaging study or one has been already performed and you’re examining the patient for something relevant to that study, always look at that study yourself and there’s a huge amount of learning in that. So, in a patient age 40 with hypertension and diabetes, should we do an MRI if the patient has disc edema but no signs of hypertension or disability retinopathy? Depends on the case. What you’re suspecting. If it’s possible the patient has a condition where the MRI would be helpful, compressive optic neuropathy or optic neuritis or venous sinus thrombosis. anything you can think of that imaging studies help, do it. What method do you use to evaluate color vision in your presentation? To with evaluate color vision, I use are the Ishihara color plates. There’s a number of other ones that you can use, but those are the ones that I use. Okay. And then question in Spanish here. So, in retinoblastoma — so the question is about calcifications in retinoblastoma. Yeah, that one I’ll have to defer to someone else on that. I’m not as familiar with that. I don’t see a lot of retinoblastoma patients in my clinic. All right. In the case of any benign IC — I’m not sure what ICSOL is, so I won’t be able to answer that. Okay. What imaging features help distinguish demyelinating optic neuritis from ischemic, inflammatory, or infiltrative optic neuropathies? So ischemic optic neuropathies shouldn’t contrast enhance, optic neuritis which should know enhancement. And inflammatory, that’s same thing as optic neuritis, may not be ed idiopathic. But if it’s true an inflammatory process, expect abnormal contrast enhancement. And expect that with infiltrative optic neuropathies, those can be hard to distinguish based on the imaging caricaturists. Question about how to choose between MRA and CTA? That can be a challenge. Sometimes it depends on the availability of each in your institution. The institution where I am now, we tend to get CT angiograms. They tend to be, I think a bit easier in terms of — they’re faster, give us the quality that we need and sometimes can be less expensive than MRA as well. Either one can give you the information you need in many cases. Okay. And then a question: Do you think that in all cases we can find the localization based on exam? No, you can’t do Noemi in all cases. Do the best you can. Localize the best you can, and if you can’t, you did the best with what information you have. Okay. So, when to scan in new onset sequence and children without signs of nerve palsies. Depends on the neurologic disorder. Signs or symptoms that make you worry about a neurologic disorder or that would be — or that makes this an atypical case of strabismus-related business musculoskeletal in a child, consider imaging. Do a lumbar puncture on all patients? No, I don’t. If it’s an atypical case, it’s worth condition. And otherwise I typically do not have a lumbar puncture performed. For thyroid disease, is it okay to order all the time or in some cases? You can do either one. A CT scan of the orbits can tell you about enlargement of the extra ocular muscles. And that’s one of the things you’re looking for with active thyroid eye disease. That can be a CT scan or an MRI. Either one will answer that. Are you in favor of making the favor of IIH without papilledema, if it helps the patient, potentially. Patients without papilledema cannot lose vision from IIH. That’s good. But if it’s helpful in management of the symptoms, it’s sometimes helpful for a patient to have a label to their symptoms. Okay. What is the difference between typical and atypical optic neuritis? Typical optic neuritis has a typical course, decreased central and color vision, pain with eye movements that may be present for a couple days before the onset of vision loss. And even this the vision loss tends to improve over a course of weeks to months, and eventually usually recovers quite well and does not reoccur, at least in the short-term. Those would be typical features. Atypical is anything that deviates from that. Bilateral, super severe vision loss, significant optic disc edema with their optic neuritis. Early reoccurrence of the vision loss or like a dispense to the vision loss, all atypical feature that would require further evaluation. Indications for ultrasound imaging. So, I didn’t really go over that. But sometimes that can be used in especially in babies and neonates to look for different abnormalities in the brain. And actually can be used in utero a lot as well. There’s no ionizing radiation. So, that can be an indication for it. Also, ultrasound to look for optic nerve Drusen can be helpful. Also, ultrasound can be a quick way to look for even extra ocular muscle enlargement or nerve sheathe fluid that’s excessive. There’s a number of things that ultrasound can be useful for. Certainly an option if you don’t have other imaging modalities available. And then would you consider a plain CT scan with orbital cuts enough to assess for pediatric patients with vertical tropia? Trends on the pathology you’re looking for. If the pathology is something that could be missed by that, then maybe not. So, yeah, it’s hard to give a blanket statement without a specific case. I know I see cases, for example, where something like that wouldn’t be caught on a CT scan, so something like a — a cranial nerve Schwannoma, those aren’t super-common in the pediatric realm, but other things that are very subtle but might not be picked up by CT scan. It depends on the clinical history and the actual patient. Okay. So, for case of WEBINO, wall-eyed bilateral internuclear ophthalmoplegia — preferred imaging. Brain with contrast. That’s the imaging there were. A motor vehicle accident six months ago, now diplopia, what could be the cause and possible lesion localization? The six nerve palsy can be falsely localizing. May not be a lesion along the sixth nerve. Concern about high intracranial pressure. Think about that. And any lesion along the course of the sixth nerve? With the head trauma, it could be a traumatic sixth nerve palsy. In that case, it would be the peripheral nerve that was damaged. Orbital cellulitis, CT scan or MRI, scan of the orbits can look for orbital cellulitis. Prompt imaging modality for neuritis, best imaging modality is the MRI of the orbits. If it’s acute optic neuritis, get an MRI of the brain at the same time to screen for multiple sclerosis. And bitemporal — normal MRI with contrast. What to order more? If it’s bitemporal, think of the localization at the chiasm. First review the MRI images that was done. And if that’s not an abnormality, get an MRI of the sella with contrast and see if that shows an abnormality. OCT can be helpful as well to confirm is there optic nerve damage? Especially is there nasal ganglion cell loss which could also localize to the chiasm. So, how to write requisition for lesions involving the course of the facial nerve? You can just say which facial nerve you’re suspecting the abnormalities in and please thoroughly image the course of the left facial nerve. You can say something like that, and that would be enough for the radiologist to understand. When ordering a CT scan for an orbital tumor in the case, thinking of bone erosion and malignant tumor. Yes, if there’s a concern for bone erosion, a CT scan is probably gonna be the best modality for looking for that. So, yeah, that would be good to do. Is FL AIR the same as T2? It’s not the same as T2. The difference is T2 FLAIR has the difference subtracted out. And the T2 does not. There are advantages to each of those, those are a little different. Use of steroids in optic neuritis, outside of the discussion today. And so question about someone who had bilateral disc edema due to oral tetracycline, and showed a hypoplastic transverse sinus. Yeah, hypoplastic transverse sinus is not abnormal. There wouldn’t be a need for any intervention or further evaluation of that. You would to want treat the disc edema. And if they have high intracranial pressure, treat them as such. A question about spinal MRI in cases of bilateral optic neuropathy such as in the case presented. Yeah, if you suspect or you’re looking for evidence of multiple sclerosis in someone with optic neuritis, may be possible to do a spinal MRI. I don’t do that on everyone, I’m getting MRI brain, orbits, and cervical and thoracic spine. And certainly if they have any findings on their examination that could point to spinal cord pathology, or anything in their history that could suggest they may have had an abnormality in the past related to the cervical or thoracic spine, do it. Or MRI of the brain, close to meeting the diagnostic criteria for MS, but not quite, if there’s a lesion in the cervical or thoracic spine where they meet the criteria, that’s where I get the spinal MRI. Patients with a reported normal MRI, what’s the systemic approach to if it’s the imaging modality or the protocol or the analysis itself? That’s what whole webinar was about, go back through the recording and remember to focus on history and examination, localization, getting right imaging study or reviewing the localized area in the imaging study that was already done. And then review the images yourself. And you may everyone call up the radiologist and explain the localization, too, and sometimes you can review that together. So question about book for neuroradiology. I’m not sure. I haven’t looked too much into books. I’m not sure what those good ones would be. Okay. Sorry, my Q&A is scrolling here without me. I’m trying to catch up here. So, okay. So, is it correct to sigh anisocoria in the dark is sympathetic and the light is parasympathetic? Not always, it’s not always have a neurologic cause. If it is, then you’re correct. Disadvantages of MRI, the cost, sometimes finding incidental findings. Sometimes patients not able to tolerate the duration of the scan and being claustrophobic in a small space. Those would be some potential disadvantages of MRI. MRI signs of retrobulbar inflammation and pseudotumor, find contrast enhancement in the orbits. That will be one of the main signs. Okay. Diabetic patient with massive renal impairment, comes with signs and symptoms of neurological deficit like a cranial nerve palsy, CT or MRA. Maybe, depends on what cranial nerve palsy. Cranial nerve palsy 3, if there’s that and it’s acute, that should always prompt CTA or MRA, particularly if it’s pupil-involving. So, yeah, with the others, with cranial 4 or 6, might not need vascular imaging, depends on the case. Is it mandatory to do imaging in all cases of disc edema? All cases of bilateral disc edema do need imaging. The reason is you have to exclude an intracranial mass lesion. If that hasn’t been done, do it in all cases of bilateral disc edema. Yeah. If you suspect NAAION, if you’re pretty certain about that diagnosis, don’t necessarily need neuroimaging in those cases. Okay. And can a pseudoaneurysm of the middle meningeal artery cause a partial third nerve palsy post-trauma? No. I don’t see how that can happen. That artery does not travel next to the third nerve. However, if there’s a CC fistula, then that can change things. There would still have to be something right next to the third cranial nerve, though. Let’s see… okay. How to differentiate between red/green color Blindness — that’s a general neuro-ophthalmology question. We’ll stick to the neuroimaging ones here. Any tips on how to learn to interpret neuroimaging scans yourself? Yeah, practice. Try to find the visual pathways, especially in patients with abnormal findings. Really make sure that you always look at the images yourself and try to see what the radiologist is telling you about. Those would probably be some of the best ways and use resources online, things like Radiopaedia, may be courses or neuroradiology online. Those are things that you can try. So, difference between mucormycosis and AML on an MRI? What sequence required for orbital involvement? MRI is helpful for the soft tissue. Mucor often erodes bone, and go through bony sinuses. A CT scan of the sinuses or of the head can sometimes be helpful in clarifying that part of it. And mucor tends to progress, especially without treatment, very quickly over several days. Anyone are with risk factors, think about mucor, especially with that time course. How do you decide a patient needs more imaging versus better imaging versus reconsideration of the clinical diagnosis? Yeah, that’s — that’s one of the magic mysteries of the universe. That’s tough. You got to, again, go back to the basics. History and exam, localization, get the right study of the right area and review it yourself. Those things — that’s the approach that I use. Sometimes radiologists suggest further studies, how do you approach that? Well, if it’s something that’s relevant for me to do, then I will sometimes do these studies, depending on whether I think they’re helpful or not, and then if it’s something outside of my purview, then I would suggest the patient’s primary care provider determine whether further imaging is necessary if recommended by the radiologist. For a roof or orbital floor fracture — what sequence? A CT scan of the orbits would be the best way to look at orbital fractures. If you could standardize one MRI protocol for every patient with unexplained optic neuropathy, what sequences would be mandatory or why? Orbital protocol has the relevant sequences. We don’t really center to specify the sequences, they come with it. It’s gonna include T1 with fat suppression, include T2 sequences. There should be plenty of coronal images as well to look through. So, those are gonna be some of the basic ones. But yeah, that’s done automatically. Do I do an MRI for every case of acute cranial nerve palsy? If it’s crane yay nerves three, four or six, that’s ideal. If you don’t have the resources to do that, you can restrict to patients where they have — they don’t have vascular risk factors or it’s progressive or some other feature to it. Certainly pupil-involving third nerve palsy, that’s emergent and needs CTA or MRA to exclude aneurysm. Preferred neuroimaging sequence for cases of suspected NAAION, I don’t do imaging for NAAION. Okay. And we’ll answer just a couple more questions here. And let’s see. I think we’re almost through them all. To check for optic nerve compression and thyroid eye disease, is CT or should we order MRI? You could do either. The clinical examine is more important. Is there optic neuropathy on the clinical exam? That should tell you. What radio logical sign is considered for specific increased ICP? MRI signs for retrobulbar — contrast enhancement of the orbital fat or muscles or both. When clinical localization and neuroimaging disagree, which do you trust first? Clinical localization. What do you use to resolve the discrepancy? Great question. I have to look and see what are the absolute truths and sometimes I have to revise my own diagnosis. Sometimes I’m the one that’s wrong. Be open to that possibility. Do we need to order MRI brain in all patients with neurofibromatosis? Signs of optic neuropathy and so forth, may be necessary. So, FLAIR and STIR are specifically for T2, not T1 sequence. I think there can be one that’s beyond our discussion today. But the sequences that I showed that were FLAIR and STIR, those were T2-based. Would you image for NAION? If I think that’s the diagnose, no. Why don’t you use orbits dual echo sequence? I don’t know what that is, I’ll have to defer that question. Patients with bilateral disc edema, MRI showing optic nerve enhancement up to the chiasm, can we see any features on MRI to rule out infections like TB or lumbar puncture is compulsory? You can’t really rule out TB with a scan. You can suspect it, you can’t rule it in. Screen for signs of it, maybe abnormal contrast enhancement basal meninges, signs of tuberculosis. Can’t rule it out with an MRI scan. If you suspect it, test for it, if an LP is necessary, do that. If you suspect that a person has diabetes or hypertension, undergo lab tests or an MRI? Just having diabetes or hypertension is not an indication for an MRI. That was a lot of questions, everybody. Thank you all so much for your participation and for the interest. Hopefully you found the principles today here useful. We wish you the best in evaluation of your patients. Thank you.
