INTRODUCTION

Neuroimaging in epilepsy is undergoing a renaissance. Gone are the days of just “making sure there wasn’t anything big or acute.” Thin slice MRI, fMRI, diffusion tensor imaging, SPECT/Perfusion, PET, and magnetoencephalography are just some of the modalities neurologists have at their disposal, and many exams yield complementary information. In addition, academic neuroimaging departments are developing specialized epilepsy protocols in conjunction with neurologists, taking into account their specific patient population. The complex milieu of diagnostic tests, multidisciplinary approaches, and subsequent medical decision making cannot be undertaken without a thorough understanding of what the images are trying to tell us. It would be prudent to know what diagnostic tests are offered at your facility, and to know how and why they may be different than at another facility or outpatient center.

THE IMAGING ARSENAL

MRI

The most important imaging modality in epilepsy is MRI, which is the study of choice for epilepsy patients. While MRI was initially used to rule out tumors, hemorrhage, and large lesions, modern-day MRI can detect subtle lesions such as mesial temporal sclerosis and cortical dysplasia. MRI sensitivity is increased when suspicion of the anatomic lesion (via clinical suspicion and EEG findings) are available to the radiologist. Most pediatric academic centers will image a patient with epilepsy a different way than a patient with a headache. The primary difference is thinner slices through the temporal lobes, in the hopes of catching mesial temporal sclerosis or hippocampal atrophy. Another important sequence that may be added for epilepsy patients are those that improve detection of polymicrogyria and heterotopic foci.

While a complete list of all the specialized MRI sequences that can be used and have been used to image epilepsy patients would be exhaustive, it is imperative to have a basic understanding of what is possible so that you can assist the radiologist in getting the best exam for the patient.

T1

The Basics: T1 is used primarily to evaluate the anatomy of the brain. After myelination, all the gray matter fibers are gray and all the white matter fibers are white, an important differentiating feature from all other sequences. Fat, blood, melanin, and protein are bright on T1 and fluid is dark.

Advanced Concepts: High-resolution (2 mm thick slices or less) T1-weighted imaging can be helpful in gray–white matter differentiation. These would be highly useful in patients suspected of having gray matter heterotopia, polymicrogyria, and schizencephaly. Subtle areas of cortical dysplasia may only be identified after EEG findings are correlated with high-resolution T1 sequences. Finally, high-resolution T1 images are often fused with MEG, fMRI, PET, and SPECT images, yielding a link between anatomic and physiologic information.

T2

The Basics: T2 is primarily used to look for any signs of brain edema. In myelinated patients, gray matter is white and white matter is dark. Both fat and fluid are bright on T2-weighted imaging. In addition, flowing blood vessels appear as dark flow voids on T2-weighted sequences, and gross anatomic information about the major vessels of the circle of Willis could be obtained without the need of an MRA.

Advanced Concepts: High-resolution thin section T2-weighted imaging, usually abbreviated CISS, FIESTA, or SPACE, depending on the MRI vendor, are ideal for looking at intraventricular lesions, cysts, or any lesion with mass effect. Unfortunately, these sequences lose a lot of soft tissue contrast, and would not be ideal for evaluating a lesion within the brain parenchyma such as cortical dysplasia.

T2 FLAIR

The Basics: FLAIR stands for fluid attenuated inversion recovery, and means exactly what the acronym stands for. In FLAIR, a fluid-sensitive sequence such as T2 is “attenuated and inverted” in such a way as to make all the CSF signal dark. What this means is that all the T2 hyperintense fluid that was close in signal to CSF is now “dark” and any fluid that is not CSF signal will remain bright. For example, enlarged perivascular spaces contain normal CSF fluid, and will be bright on T2 but dark on T2 FLAIR (so called “FLAIR suppression”). On the other hand, infection and infarction would not contain normal CSF fluid and would be bright on T2 FLAIR (nonsuppression on FLAIR). Similarly, hippocampal post-ictal edema would show up as a bright signal on T2 FLAIR sequences.

FLAIR usually refers to T2 FLAIR. There is also a T1 FLAIR sequence that has nothing to do with the previous conversation.

Advanced Concepts: Volumetric T2 FLAIR imaging is a relatively new concept of being able to visualize T2 FLAIR signal abnormalities in any plane. This can be particularly useful in surgical planning for resection of cortical tubers in TS patients and in identifying small demyelinating lesions.

T2 STIR

The Basics: STIR stands for short Tau inversion recovery. This is a T2-weighted sequence where the fat signal is attenuated, without any inversion to fluid. As a result, all fluid is bright and everything else is dark. This sequence is very sensitive, but not specific, as prominent perivascular spaces can have a bright signal.

Advanced Concepts: In this author’s experience, thin slice STIR images through the hippocampi are the most sensitive in the detection of both hippocampal edema and hippocampal volume loss. They can almost be too sensitive, as subtle abnormalities related to normal developmental asymmetry could be mistaken for pathology. If an abnormal STIR signal is detected, these should be correlated to other sequences, especially T2 FLAIR.

Diffusion-Weighted Imaging

The Basics: Diffusion-weighted imaging (DWI) involves echo planar images that can detect the motion of water. While a complete discussion of these sequences is beyond the scope of this text, suffice it to say that two series of images are created, a diffusion-weighted image (usually called B1000) and an ADC Map. Signal that is bright on the DWI and dark on ADC Map usually represents an acute process, and is termed restricted diffusion. This pattern can be seen in post-ictal edema, infarcts, infections, and highly cellular tumors. The pattern of restricted diffusion often helps differentiate between these etiologies.

Advanced Concepts: DWI is also helpful in detecting small areas of heterotopic gray matter. Thin slice DWI is extremely helpful in confirming subtle abnormalities and ruling out artifact. Since these are short sequences generally not lasting more than 2 minutes, repeat sequences can be considered. At our institution, we will sometimes obtain a routine 5 m sequence as well as a 3 mm thin slice DWI in another plane (ie, coronal). This is not recommended for every seizure patient, and is usually reserved for seizures in the neonatal setting or in patients who have known post-ictal edema on recent imaging.

SWI

The Basics: Susceptibility-weighted images (SWI) are most sensitive in detecting blood products and calcifications, which appear as areas of dark signal.

Advanced Concepts: Consider adding SWI sequences in patients with tuberous sclerosis, as calcified tubers are commonly associated with epileptogenic foci. These can also be added to patients with a prior history of hemorrhage. Outside of these two clinical scenarios, SWI has a limited role in epilepsy patients and is not a part of the routine epilepsy protocol at our institution.

Contrast-Enhanced T1

The Basics: Contrast-enhanced MRI refers to the intravenous administration of gadolinium, which shows up as a bright signal on T1-weighted images. Since gadolinium does not cross the blood–brain barrier, enhancement (bright signal) of the brain parenchyma indicates a break in this barrier. Contrast-enhanced MRI is usually obtained in patients when there is concern for infection or tumor.

Advanced Concepts: Outside of seizures due to suspected infection, vasculitis, or tumors, there is little role for contrast-enhanced MRI for seizure patients. Actively demyelinating lesions that can occur in MS and ADEM patients demonstrate incomplete enhancement and are often used when a known MS patient has new, acute symptoms.

Constructing an Epilepsy MRI Protocol

An epilepsy MRI protocol should at the very least include thin slice fluid-sensitive sequences through the hippocampal formations. At our institutions, we have two epilepsy protocols, one being a high-resolution seizure MRI protocol and the second being a routine seizure protocol (Table 74.1). The high-resolution MRI protocol can be used for the initial evaluation of seizures and in complicated cases, while the routine seizure protocol is often used in follow-up evaluations or in patients with unconfirmed seizures.

Computed Tomography

CT imaging plays an important role in the imaging of patients presenting to the emergency room with acute or new onset of seizures. Since CT images can be obtained rapidly, sedation is usually not needed. They are particularly useful when seizures are suspected to be secondary to hemorrhage, especially in the post-trauma setting. Hydrocephalus and brain herniation secondary to mass effect are other indications in which CT can be helpful, especially in the acute setting. In the outpatient setting, CT can be used to detect areas of mineralization, such as calcified tubers in tuberous sclerosis patients.

Table 74.1

As CT images are performed using ionizing radiation, which is sometimes 100 times greater than a chest radiograph, caution must be used when ordering CT studies. This is especially true in nonemergent settings, outpatients, or in patients with chronic hydrocephalus who have had multiple prior exams. CT is ideally performed without contrast, and contrast-enhanced CT have limited utilization in seizure patients.

Nuclear Medicine

Two nuclear medicine techniques are used in the imaging of seizure patients: single photon emission tomography (SPECT) and positron emission tomography (PET).

Single Photon Emission Tomography (SPECT)

SPECT perfusion scans measure cerebral blood flow. Usually chemicals such as 99mTc-exametazime (HMPAO) or 99mTc-bicisate are labeled with a radiotracer such as Technetium-99m (99mTc) and injected. This can be done while the patient is seizure-free, and called interictal SPECT (II-SPECT). In controlled settings, this can be administered during an active seizure and is called an ictal-SPECT (I-SPECT). Since epileptogenic foci usually do not have metabolically active tissue, these usually appear as dark (hypoperfused) regions on II-SPECT images. Conversely, seizure foci have increased metabolic demand during a seizure and will therefore be bright (hyperperfused) on I-SPECT images. Both II-SPECT and I-SPECT images can be fused together and subtracted to accentuate foci that demonstrate only subtle hypoperfusion and/or hyperfusion. This can be further fused to a structural image, such as a high-resolution T1-weighted sequence, giving anatomic details about the exact location of an epileptogenic focus.

Care should be taken in evaluating fused II-SPECT and I-SPECT images, as this requires significant postprocessing, especially when further fused with MRI, as this may introduce artifacts and result in false-positives or false-negatives. Also, I-SPECT agents have to be administered as soon as possible after seizure onset, usually within minutes. Therefore, I-SPECT studies are exclusively reserved for patients undergoing extensive seizure workup in an inpatient setting.

Positron Emission Tomography (PET)

PET has also been used in some centers to evaluate epilepsy patients. The most commonly used radiotracer is 2-deoxy-2[¹⁸F] fluoro-d-glucose. Other radiotracers may be used in research protocols. While SPECT looks at cerebral blood flow, FDG-PET evaluates glucose metabolism. Since PET radiotracers are more expensive and have a shorter half-life than SPECT radiotracers, ictal PET scans are not done and images are usually acquired only in the inter-ictal setting. Since PET images have poor anatomic detail similar to SPECT, these images are also usually fused with high-resolution MRI sequences (usually thin slice T1) to improve anatomic localization.

Advanced Techniques (MRS, fMRI, MEG)

Functional MRI (fMRI) and diffusion tensor imaging (DTI) can provide additional information including localization of language centers and white matter tracts, respectively. Advanced imaging techniques including magnetic resonance spectroscopy, functional MRI, magnetoencephalography, and diffusion tensor imaging are discussed in the text that follows.

MR Spectroscopy

MR spectroscopy should be considered in patients with suspected syndromic disorders such as Canavan’s disease and mitochondrial disorders. In Canavan’s disease, a high NAA peak is virtually diagnostic of the disorder and can sometimes be present before white matter T2 flair hyperintensities are evident. In mitochondrial disorders, a typical lactate peak at 1.3 ppm is indicative of the increased lactic acid due to mitochondrial dysfunction and anaerobic glycolysis.

Functional MRI

Functional MRI is usually performed in patients with refractory epilepsy in which surgical resection is being considered. A confirmed epileptogenic focus is known prior to the exam, and fMRI helps in determining if the epileptogenic focus is adjacent to or a vital part of functioning. This is particularly important in identifying language centers prior to temporal lobe resections, and in tuberous sclerosis patients prior to resection of an epileptogenic tuber.

In the pediatric setting, it is often not possible to obtain a functional MRI as these patients require sedation. A passive fMRI protocol where sedated patients’ limbs are moved by the radiologist while images are being acquired can be used to identify vital motor centers and has been successfully implemented at our institution.

Magnetoencephalography

Magnetoencephalography (MEG) can be used to identify potential seizure foci. Action potentials within neurons are similar to a wire generating an electric current, and MEG can measure slight variations to this current. Magnetic information overlayed upon structural images, typically thin slice T1 sequences, can provide information about differences in voltage and hence cortical activity. This can be especially helpful in identifying epileptogenic cortex, which can be occult even in high-resolution MRI.

Diffusion Tensor Imaging and Functional Anisotropy

Diffusion tensor images (DTI) can provide information on the direction of white matter tracts, which is physiologic information that can’t be obtained on traditional imaging. The key to understanding DTI imaging is the concept of anisotropy, which is a measure of the direction of water flow. Since water will preferentially flow along the long axis of an axon, the directionality of that water flow is synonymous with the directionality of the white matter tracts. This “functional anisotropy” can be further processed to create detailed maps of a patient’s white matter tracts, termed DTI-tractography. Care should be taken when evaluating these DTI images, as white matter tracts are determined by complex mathematical algorithms that rely on probability assumptions. Hence, while DTI provides excellent physiologic information, anatomic localization and size characterization may not be accurate.

There are many clinical applications of DTI imaging, including evaluation of the patient post corpus callosotomy, and decussation of white matter tracts in suspected cases of Joubert syndrome. DTI also has an important role in the management of intracranial and spinal tumors.

Diffusion tensor imaging can be obtained in lieu of diffusion weighted sequences, as they can provide similar information about restricted diffusion in addition to information about the integrity of white matter tracts. They are also echo planar sequences and take as much time as traditional diffusion weighted images (Figure 74.1).

Miscellaneous Techniques

Both angiography and CT angiography have a limited role in the evaluation of seizure patients, especially since the widespread development and utilization of MR-angiography (MRA). MRA of the brain is typically obtained as a 3D-time of flight sequence, the exact details of which are beyond the scope of this chapter. While good anatomic detail of the major vessels comprising the circle of Willis can be obtained on noncontrast MRA, contrast is needed to evaluate smaller branch vessels. MRA is typically considered for patients in whom an underlying vasculitis or vascular malformation such as AVM is suspected. They can also be used in patients with neurofibromatosis with suspicion of Moya-Moya disease. Cerebral angiography was the basis for development of the intracarotid sodium amorbarbital (ISAP) test, better known as the Wada test. This is often used in seizure patients to confirm lateralization of language and memory and can be complementary to FMRI.

Radiographs have limited usefulness in the seizure setting, but have been used for evaluation of shunt catheters and depth electrodes. They are particularly useful in evaluating the placement of grid electrodes as the post operative cross sectional images can have significant artifact.

FIGURE 74.1 Corpus callosotomy A and B. (A) FA image post corpus callosotomy demonstrating no decussation of white matter fibers along the corpus callosum. (B) Corresponding high-resolution T1 image.

DISEASES WITH SPECIFIC IMAGING CONSIDERATIONS

Stay updated, free articles. Join our Telegram channel

Join