eMedicine Specialties > Neurology > Neuro-imaging
Neuroimaging in Epilepsy Surgery
Updated: Sep 29, 2006
Introduction
The imaging of epilepsy has vastly changed in the last 15 years. Prior imaging with computed axial tomography (CT) infrequently revealed the pathologic substrate for epilepsy. Although early low field strength MRI increased the diagnostic yield, it could only identify obvious pathology such has neoplasms, encephalomalacia, and vascular malformations. The advent of high-resolution magnetic resonance imaging with a dedicated epilepsy protocol has significantly increased the frequency that a pathologic substrate for epilepsy is identified. This has had a dramatic clinical impact on the evaluation and management of epilepsy since MRI findings can assist with classification, determine prognosis for remission, predict long-term intractability to antiepileptic medications, and identify potential surgical candidates.
The International League Against Epilepsy Guidelines for Neuro-imaging in the Epilepsy Patient (1997) recommends a dedicated epilepsy protocol MRI for all patients with a new-onset seizure or newly diagnosed epilepsy in a nonemergent setting. Unfortunately, in most clinical practice settings, a nonepilepsy protocol MRI is performed. While a routine MRI will exclude ominous structural substrates that require urgent treatment in themselves such as high-grade gliomas and arteriovenous malformations, subtle structural substrates such as hippocampal sclerosis and malformations of cortical development (MCDs) will be missed. Identification of these substrates has long-term therapeutic and prognostic implications for remission versus intractability.
Approximately one third of patients with partial-onset epilepsy are medically intractable defined as failing at least two first-line antiepileptic medications. The chance of being considered an epilepsy surgery candidate is greatly enhanced when a structural substrate is found on MRI. In a recent cross-sectional study of 495 patients in an epilepsy clinic, 51% had an abnormal scan (CT or MRI). In those patients with focal epilepsy who had a standard MRI, an abnormality was found in 49%, and in those who had an epilepsy protocol MRI an abnormality was found in 72%. An earlier study by Li et al (1995) found similar findings.
Another study compared standard MRI interpreted by nonexpert neuroradiologists to the interpretation by an expert neuroradiologist. The same patients then had a follow-up epilepsy protocol MRI interpreted by an expert neuroradiologist. The nonexpert reports of standard MRI identified 39% as abnormal, the expert reports of the same MRI identified 50% as abnormal. Epilepsy protocol MRI interpreted by an expert identified a focal lesion in 85% of those with a normal standard MRI. The most commonly missed finding was hippocampal sclerosis followed by tumor and other findings. Less commonly missed findings included vascular malformations and malformations of cortical development. The ILAE Commission on Neuroimaging also recommends an epilepsy protocol MRI in all patients with intractable epilepsy.
A study by Kwan and Brodie in 2000 found that, in patients with newly diagnosed epilepsy, only 47% became seizure free with the first AED and only 14% with the second AED. Further AED trials achieved seizure freedom in only 5% of patients. This study showed that intractable epilepsy is more common than generally believed, and can be identified early. However, referral for epilepsy surgery is delayed in half of intractable epilepsy patients who are referred greater than 10 years from the onset of intractability. This further underscores the importance of epilepsy protocol MR imaging which has a high yield for identifying a structural substrate that would enable these patients to be referred early for an epilepsy surgery evaluation thus, providing them with an optimal chance of attaining seizure freedom.
Epilepsy protocol MRI
Conventional MRI imaging is inadequate for epilepsy patients since many of the findings are subtle and easily missed. Routine MR imaging consists of a short scan time, 3-to 5-mm thick slices with an interslice gap of 2-3 mm. These studies do not include SPGR or MPRAGE T1-weighted images that enhance gray white matter differentiation, which is crucial to analyze the cortical architecture. These images are also not acquired in a coronal oblique plane perpendicular to the long axis of the hippocampus, which is particularly important when evaluating temporal lobe epilepsy (TLE).
Epilepsy protocol MR imaging, on the other hand, includes the entire brain from nasion to inion, T1-weighted MPRAGE or SPGR images 1.5-mm slice thickness with no intervening gap obtained in the coronal oblique plane (if temporal lobe epilepsy is suspected). These images are acquired as a 3-D volume thereby allowing postprocessing to correct for head misalignment and for reformatting images into multiple planes to confirm a subtle malformation of cortical development. An epilepsy protocol MRI also includes coronal and axial FLAIR sequences with 3-mm slice thickness and 0- to 1-mm interslice gap. A conventional thin slice (3 mm) T2-weighted axial and coronal sequence is also obtained. Gadolinium is not required, unless a tumor or a vascular malformation is identified and in some neurocutaneous syndromes such as Sturge-Weber syndrome in order to visual leptomeningeal angiomatosis.
Functional neuroimaging techniques, including positron emission tomography (PET), single-photon emission computerized tomography (SPECT), magnetic resonance spectroscopy (MRS), magnetic source imaging (MSI), and functional MRI (fMRI) are not universally available, but can be very helpful in the localization of the epileptogenic zone and for mapping functional areas of the brain such as language and motor function. The clinical utility of each imaging modality will be reviewed, with an emphasis on MRI structural imaging.
For excellent patient education resources, visit eMedicine's Brain and Nervous System Center. Also, see eMedicine's patient education article Epilepsy.
MR Imaging Of The Temporal Lobe
Hippocampal sclerosis
Hippocampal sclerosis (HS) is characterized by neuronal loss and gliosis. HS is the most common pathologic substrate of surgically treated epilepsy in adults seen in 67% of patients. In patients with newly diagnosed epilepsy, it has been reported in 1.5-3% of adults. When evaluating the medial temporal structures (hippocampus, amygdala, entorhinal cortex, and parahippocampal gyrus), one should evaluate the size, signal, shape, and dual pathology (SSSD). The typical MRI findings of HS include atrophy of the hippocampus on T1-weighted SPGR that is typically seen in 90-95% of cases. The atrophy is most prominent in the hippocampal body.
On FLAIR imaging, increased signal is observed within the hippocampus. FLAIR is ideally suited to identify signal changes within the hippocampus since gliotic changes have increased water content appearing as increased signal on T2 weighted MRI. The FLAIR sequence nulls the increased signal intensity of the CSF in the temporal horn of the lateral ventricle and the choroidal fissure that can dwarf the increased signal in the hippocampus on a conventional thin-slice T2-weighted spin echo image. One must also be aware that the baseline signal of the hippocampus on FLAIR MRI is greater than of the cortex, and can be mistakenly interpreted as bilateral HS. In these cases, thin-slice coronal T2-weighted images should be reviewed for confirmation.
Hippocampal atrophy and increased signal are not always seen together in the same patient. For example, some patients have increased FLAIR signal or T2 signal without accompanying atrophy. The increased T2 signal is felt to reflect gliosis rather than neuronal cell loss. Occasionally, secondary findings of hippocampal sclerosis will be observed such as (1) enlargement of the ipsilateral temporal horn, (2) thinning of the fornix, (3) mamillary body atrophy, (4) loss of normal interdigitations of the hippocampal head, and (5) atrophy of the collateral white matter between the hippocampus and the collateral sulcus.
High-resolution MRI imaging is 80-90% sensitive for identifying hippocampal sclerosis by qualitative interpretation. Bilateral hippocampal atrophy is seen in 10-20% of patients, but it may be difficult to visualize unless quantitative measures are done. In children with intractable epilepsy, HS is also seen on MR imaging in children. For example, Hs is observed in 21% of children with newly diagnosed TLE and up to 57% of children with intractable TLE. More common findings in children with intractable TLE include malformations of cortical development and developmental tumors.
Entorhinal cortex atrophy
While quantitative assessment of the entorhinal cortex (ERC) is ideal, ERC size can be evaluated by visual assessment. ERC atrophy is often seen concomitantly with hippocampal atrophy, but it can also be seen independently. It is characterized by thinning of the collateral white matter of the ERC.
Hippocampal sclerosis and dual pathology
Approximately 15-20% of patients with HS have dual pathology, that is, another pathologic substrate within or outside the temporal lobe in addition to HS. Dual pathology is more commonly observed with malformations of cortical development and developmental lesions such as porencephaly. It is less commonly seen with vascular malformations (7%) and neoplasms (2%). With vascular malformations, HS is more likely to be present if the lesion is in close proximity to the hippocampus.
Pearls and pitfalls in the evaluation of hippocampal sclerosis
One must also be vigilant not to attribute all cases of increased signal within the hippocampus to HS. While increased signal on FLAIR indicative of HS is not always accompanied by hippocampal atrophy, the hippocampus should never be abnormally enlarged since this indicates a low-grade neoplasm, hamartoma, or hippocampal dysplasia and not HS.
Amygdala atrophy and dysplasia
The amygdala should be evaluated for symmetry. Occasionally, the amygdala may be reduced in size along with hippocampal atrophy. However, this is difficult to observe visually. Amygdala atrophy accompanying HS has been identified with volumetric analysis in 12% of pathologically confirmed patients with HS. Another study identified ipsilateral amygdala atrophy by volumetry accompanied HS in 20% of patients. However, they also found that amygdala atrophy was also frequently seen contralateral to HS in 15%.
Sometimes, the amygdala is pathologically enlarged consistent with amygdala dysplasia, hamartoma, or a low-grade neoplasm. This enlargement may or not be associated with increased signal on FLAIR MR imaging. Usually, in the setting of a low-grade glioma or a developmental tumor, accompanying T2 signal changes are present. Amygdalar enlargement is not associated with hippocampal atrophy and is seen in "imaging negative" TLE. However, sometimes amygdala enlargement is accompanied by hippocampal enlargement and likely represents hamartomatous enlargement of the hippocampus and the amygdala.
Viewing the images in the axial plane may also increase the yield of identifying enlargement of the amygdala. One group recommended that abnormal amygdala enlargement may be visually assessed by evaluating for prominent amygdala gray matter extending medially, anterior to the sylvian fissure.
MR Imaging Of Malformations Of Cortical Development And Neoplasms
Classification of malformations of cortical development
Malformations of cortical development (MCD) are classified into 3 different categories based on the stage the MCD occurred: (1) cortical dysplasia (neuronal and glial proliferation or apoptosis <10 wk), (2) heterotopias (abnormal neuronal migration 10-20 wk), (3) polymicrogyria (abnormal late cortical migration and organization >20 wk). Some degree of cortical organization also occurs postnatally. For a detailed developmental and genetic classification of MCD, please refer to Barkovich et al, 2005.
Imaging of malformations of cortical development
Malformations due to abnormal neuronal and glial proliferation or apoptosis
Microlissencephaly and microcephaly with simplified gyral pattern has been described in patients with profound congenital microcephaly (head circumference that is more than 3 standard deviations below normal at birth). These conditions are felt to result from abnormally decreased cellular proliferation or pathologically increased apoptosis. The sulcal pattern is similar to the sulcal pattern in normal patients; however, there are too few sulci. If the cortex is of normal thickness (3 mm), a diagnosis of microcephaly with simplified gyral pattern can be made. If the cortex is abnormally thick, then a diagnosis of microlissencephaly is made.
Hemimegalencephaly is characterized by abnormal enlargement of a lobe, multiple lobes, or a hemisphere due to marked dysplasia. This condition can be an isolated finding or can be seen with neurocutaneous syndromes such as (1) epidermal nevus syndrome, (2) hypomelanosis of Ito, (3) neurofibromatosis type 1, (4) Klippel-Trenaunay-Weber syndrome, and (5) tuberous sclerosis. Imaging findings include large clumps of grey matter that extend from the pial surface to the ventricle. The affected lobe or hemisphere is enlarged, the white matter shows increased signal on T2-weighted images, and usually the lateral ventricle is enlarged in the dysplastic region.
Focal cortical dysplasia with balloon cells is also known as Taylor-type cortical dysplasia. Theses lesions are nearly identical to those seen in Tuberous Sclerosis. Usually, a small focal region is involved. Sometimes, the cortical dysplasia extends from the pia to the ventricle (transmantle dysplasia). MR imaging findings include (1) cortical thickening that is observed on at least 3 or more contiguous slices; (2) blurring of the grey-white matter junction; (3) increased signal of the underlying white matter on T2-weighted images; and (4) often times, a linear, curvilinear, or funnel-shaped abnormal signal intensity is seen extending from the cortical white matter junction to the surface of the lateral ventricle.
Malformation due to abnormal neuronal migration
Classic lissencephaly (smooth brain) is defined as reduced sulcation, and the cortical surface shows reduced number and depth of sulci. In some cases, there is complete absence of sulcation, whereas in others, there is a reduction. This condition is due to a mutation of either the LIS1 gene at chromosome 17p13.3 or the DCX gene at Xq22. Both mutations are believed to cause lissencephaly by interfering with translocation of migrating neurons as they advance along radial glial cells. With the LIS1 mutation, the abnormality is most severe in the parieto-occipital lobes, and with the DCX mutation, the gyral abnormality is most severe in the middle and anterior half of the frontal lobes.
Subcortical band heterotopia (SBH) represents a milder form of the LIS1 or DCX mutation. On MRI imaging, either a thick or a thin band of cortex is seen in the white matter underlying and running parallel to the normal-appearing cortex. If the SBH is mainly frontal, the DCX mutation is the likely cause, whereas if it is observed in the parieto-occipital region, the LIS1 mutation is more likely. Other mutations have also been described, and the reader is referred to a recent review by Barkovich and colleagues (2005). SBH can easily be missed when it consists of a thin band of cortex. Clinically, some of these patients appear to have localization-related epilepsy from another region. However, surgical outcome in this group is poor since the SBH is a marker of diffuse epileptogenicity.
Heterotopia refers to a collection of neurons in an abnormal location. They can be located in the subcortical white matter, subependymal region (periventricular), or underlying the normal-appearing cortex in a laminar pattern (SBH). They are round to ovoid nodules that consist of both neuronal and glial cells. On MRI imaging, they are grey matter isointense on both T1 weighted and T2 weighted imaging. Subcortical heterotopias can range from single to multiple grey matter nodules that may extend from the ventricular wall to the cortical mantle. Occasionally, there may be dimpling of the cortex overlying the region of the heterotopia. They are usually unilateral but can be bilateral in which case they are associated with cognitive delay.
Periventricular nodular heterotopias (PVN) are most commonly seen in the lateral ventricles, with the trigones and the frontal horns the most frequent location. They are round or oval and can lie in the wall of the ventricle and protrude into the ventricular space or sometimes are present in the periventricular white matter. In mild cases, only a few nodules are seen, whereas, in more severe cases, a continuous layer can be seen lining the ventricular wall. These nodules are grey matter isointense on all MR imaging sequences. While many cases are sporadic, some cases are due to a mutation of the FLNA gene that codes for filamin 1, which is involved in the migration of the neuron from the germinal zone onto the radial glial fibers, and is located on band Xq28.
Identification of PVN is clinically important in that seizure-free outcome after anteromedial temporal is poor in patients with PVN with clinical and EEG features of temporal lobe epilepsy. The reason for this is that these nodules are epileptogenic, and recent small series have demonstrated that a good surgical outcome can be obtained when the PVN are unilateral, seizures localize to the nodule(s) and are resected.
Malformations secondary to abnormal late migration and organization
Polymicrogyria (PMG) consists of an excessive number of small gyri with shallow sulci. PMG can be (1) focal and limited, (2) focal unilateral and extensive, (3) bilateral and symmetrical, (4) bilateral and asymmetrical, (5) multifocal, or (6) diffuse. On MR imaging, numerous small gyri with shallow sulci are seen. Sometimes, the gyri appear cortex has an irregular, bumpy inner and outer cortical surface with broad gyri and shallow sulci. The cortical-subcortical junction is often irregular.
Schizencephaly consists of a CSF cleft extending from the subarachnoid space to the lateral ventricle. The wall of the cleft is line with dysplastic cortex and/or polymicrogyria. The lips of the cleft can be closed (type 1) or open (type 2). The most common location is posterior peri-sylvian cortex. Two thirds of cases are unilateral, and one third are bilateral. Schizencephaly is commonly associated with polymicrogyria, optic nerve hypoplasia, and absence of the septum pellucidum. Although both PMG and schizencephaly are strikingly abnormal on MR imaging, they often contain primary motor, sensory, or visual function.
Focal cortical dysplasia (FCD) without balloon cells consists of a focal abnormality of cortical lamination of the cerebral cortex and the underlying white matter with abnormal cortical neurons but without the presence of balloon cells or cells extending from the pial surface to the ventricular surface. MR imaging findings include cortical thickening on at least 3 or more contiguous slices, blurring of the grey white matter junction, and focal cortical thinning with volume loss of the underlying white matter. Increased signal on T2-weighted imaging may be seen; however, the signal does not extend to the ventricular surface.
An occasional helpful finding is the "cleft-dimple" complex where the size of the subarachnoid space overlying the FCD is enlarged and the cortex appears to "buckle" away from the subarachnoid space. The neuroimager must be aware that normal cortex may appear thickened when sliced obliquely cross a sulcus. In this regard, it is imperative to make sure that the findings seen on numerous contiguous slices are still present when reformatted into multiple planes and correlated with clinical, EEG, and functional imaging data. One must also be aware that the mildest forms of FCD characterized by dyslamination alone may not show any abnormality on even high-resolution MR imaging.
MRI imaging of neoplasms associated with epilepsy
Neoplasms are the structural substrate in 3-4% of patients with epilepsy in the general population. In patients with intractable epilepsy treated with epilepsy surgery, neoplasms account for approximately 20% of cases. Neoplasms associated with chronic epilepsy are usually located in the cortex and are not usually associated with mass effect or vasogenic edema. The temporal lobe is the most common location (68%). Neoplasms found in patients with chronic epilepsy include (1) low-grade astrocytic tumors, (2) oligodendroglioma, (3) gangliogliomas, (4) dysembryoplastic neuroepithelial tumor (DNET), and (5) a pleomorphic xanthoastrocytoma (PXA). On MR imaging, most neoplasms are hypointense on T1-weighted images and hyperintense on T2-weighted images. However, it is often difficult to distinguish these neoplasms, unless specific imaging characteristics for each tumor are present on MR imaging (see below).
Astrocytomas, fibrillary subtype, (WHO grade 2) are usually ill-defined infiltrative tumors that usually do not enhance with gadolinium. Pilocytic astrocytomas, on the other hand, are well defined though not encapsulated, and a mural nodule is seen after gadolinium enhancement. Oligodendrogliomas are usually peripherally located and may appear cortically based with gyriform calcifications and adjacent changes in the calvaria. They are commonly seen in the frontal or temporal lobe. Gadolinium enhancement is variable.
Gangliogliomas are most commonly seen in the temporal lobe of patients younger than 30 years. They are mixed solid and cystic lesions that are cortically based with minimal or no mass effect. Calcification is often present. Gadolinium enhancement is variable. The finding of calcification and cystic changes in a cortically based lesion raises the possibility of this neoplasm. These lesions can be associated with concomitant cortical dysplasia.
Dysembryoplastic neuroepithelial tumors (DNET) are benign low-grade, multicystic, and multinodular cortical-based tumors mostly seen in children and young adults. A cortically based hypointense nodule can be seen on T1-weighted images. Calvarial remodeling may be seen. Cortical dysplasia can be seen in 20-30% of these tumors. Gadolinium enhancement is variable. If a cortically based, multicystic tumor is seen on MR imaging, a DNET should be considered.
Pleomorphic xanthoastrocytoma (PXA) are superficially located tumors adjacent to the leptomeninges with an enhancing mural nodule. Leptomeningeal involvement is characteristic of this tumor. Local recurrence and malignant transformation can occur with this tumor in 50% of patients.
Positron Emission Tomography
Fluoro-deoxyglucose positron emission tomography
Fluoro-deoxyglucose positron emission tomography (FDG-PET) reveals interictal hypometabolism of the epileptogenic temporal lobe in more than 85% of cases. This zone of hypometabolism is much larger than the ictal-onset zone defined electrophysiologically and the epileptogenic region defined pathologically. This test is more sensitive when an asymmetry index is calculated comparing the quantitative metabolism of each temporal lobe and prevents misinterpretation due to partial volume averaging artifact. The degree of hypometabolism does not correlate with the degree of cell loss or the degree of hippocampal atrophy identified by MRI. In patients with TLE, unilateral hippocampal atrophy, and concordant EEG data, FDG-PET provides redundant data. However, it may provide additional information in patients whose MRI and EEG findings are discordant and in patients whose MRI findings are normal.
Visual analysis of FDG-PET is less sensitive in frontal lobe epilepsy, with fewer than 50% of cases showing localized abnormalities. In these cases, quantitative normalized analysis may improve sensitivity of this test. Newer techniques, such as statistical parametric mapping (SPM) and 3-D stereotactic surface projection (3-DSSP) images, may be more sensitive than conventional FDG-PET analysis. With the SPM technique, the subject's PET scan is subtracted on a pixel-by-pixel basis from a normal database of control subjects. This technique may provide localizing information in patients with either extratemporal epilepsy or TLE with a normal MRI. Co-registration with MRI may improve the sensitivity and specificity of FDG-PET by correcting for partial volume effects.
11 C-flumazenil PET
Preliminary evidence suggests that radioligand PET scans with the benzodiazepine antagonist11 C-flumazenil (FMZ) may have greater sensitivity in identifying the epileptogenic region than FDG-PET. FMZ labels central GABA receptors. Early studies at the University of Michigan showed a reduction in FMZ binding in the temporal lobe of patients with intractable TLE, which is more restricted than the region of hypometabolism seen with FDG-PET. This reduction in FMZ binding correlates with neuron loss in the hippocampus.
Newer techniques, using SPM and an MRI-based method for partial-volume effect correction, have shown that the reduction in FMZ binding is greater than what would be expected from volume loss alone. This finding suggests that, in addition to neuronal loss, GABA binding in the epileptogenic hippocampus is reduced. FMZ-PET also shows enhanced sensitivity in patients with malformations of cortical development (MCD). More recent studies have shown either increases or decreases in benzodiazepine receptor density in regions of MCD. However, surgical outcome in patients with a localized abnormality on FMZ-PET and normal MRI findings is not yet known.
Alpha methyl L-tryptophan and serotonin receptor PET imaging
In view of basic science, evidence that serotonin plays a role in epilepsy, PET imaging with serotonin precursors or serotonin agonists has been recently used with the hope of improving the detection of the epileptogenic zone. For example, reduced concentrations of brain serotonin are found in the brains of the genetically epilepsy-prone rat (GEPR). In addition, treatment with agents that facilitate serotonergic transmission inhibit seizures in many animal models of epilepsy. Reduction of brain serotonin concentrations, on the other hand, increases seizure susceptibility in animal models of epilepsy. Furthermore, in human epileptic brain tissue resected for the treatment of epilepsy, increased serotonin was found.
Alpha methyl L-tryptophan PET (AMT-PET), like L-tryptophan, is a serotonin precursor that can help measure brain serotonin synthesis rates. Like tryptophan, AMT is metabolized into serotonin, but unlike tryptophan, it is not converted into protein. AMT is converted to alpha methyl serotonin, but unlike serotonin, it is not metabolized by monoamine oxidase. Chugani and colleagues used AMT in patients with tuberous sclerosis and found reduced AMT uptake in cortical tubers as compared to normal cortex. However, epileptogenic tubers confirmed by ictal onset region demonstrated increased uptake.
Another study by Fedi and colleagues evaluated patients with either cortical dysplasia on MRI or a normal MRI. Increased AMT uptake was identified in 60% of patients with cortical dysplasia and 30% of patients with normal MRI. Juhasz et al reported similar findings. AMT has also been studied in patients who failed epilepsy surgery and was able to identify increased AMT uptake in residual epileptogenic cortex as identified by intracranial EEG. However, it could only identify the epileptogenic region in 43% of patients.
Serotonin 5-HT1A receptor binding has also been studied with the PET ligands (18F)FCWAY and (11C)WAY. Toczek et al found reduced 5-HT1A binding in the medial and lateral temporal regions ipsilateral to the epileptogenic temporal lobe. Savic and colleagues reported similar findings, but they also reported reduced binding in limbic connections such as the cingulate cortex and the insula.
Single-photon Emission Computerized Tomography
Despite great progress in structural neuroimaging, in most specialized epilepsy centers, the epileptogenic zone remains unlocalized by MRI scanning in approximately 20-50% of patients with medically intractable epilepsy. This problem has stimulated efforts to develop functional neuroimaging techniques that can demonstrate transient physiologic disturbances, not just static structural ones.
Single-photon emission computerized tomography (SPECT) scanning, after the administration of technetium Tc 99m hexamethylpropyleneamine oxime (99m Tc-HMPAO, Ceretec) or (99m Tc-ECD, Neurolite), is a readily available and relatively inexpensive method of measuring regional cerebral blood flow. These radiotracers are taken up rapidly by the brain during the first pass and, after entering neural cells, are converted rapidly to hydrophilic compounds that are trapped intracellularly and are stable for several hours. Thus, SPECT scanning can provide a semiquantitative image of cerebral blood flow 30-60 seconds after intravenous injection.
SPECT is performed during the ictal period to help delineate the epileptogenic zone. It is particularly helpful in patients with normal MRI findings, as well as in patients with abnormal MRI findings and a nonlocalizing EEG. Since seizures are associated with increased glucose metabolism (metabolism is closely coupled to cerebral blood flow), ictal SPECT scans show increased perfusion in the region of seizure onset. However, obtaining a true ictal injection is important, particularly for extratemporal lobe seizures, since with late injections, the areas of increased perfusion may represent seizure spread rather than seizure onset.
Obtaining a snapshot of cerebral blood flow during the epileptic seizure ("ictal SPECT") and comparing that to the results of an injection when the patient is free of seizures ("interictal SPECT") is relatively convenient. An area of increased blood flow during the seizure that demonstrates decreased blood flow during the interictal period is more likely to be the site of seizure onset and correlates highly (approximately 90% sensitivity) with MRI abnormalities.
In TLE, ictal SPECT has 90% sensitivity in localizing seizures, with good interobserver reliability. Ictal increased perfusion is seen in both the medial and the lateral temporal lobe. In the immediate postictal period (60 seconds), hyperperfusion of the medial temporal lobe with hypoperfusion of the lateral temporal lobe are noted. In the late postictal period (up to 20 minutes postictally), perfusion in both the medial and lateral temporal lobes may be decreased.
From a practical point of view, however, SPECT scanning adds little useful information in patients who have lesions detected by MRI and localizing or lateralized EEG findings. Ictal SPECT is not helpful in localizing seizures in patients with bilaterally independent temporal lobe seizures, since the procedure samples only one seizure at a time. Moreover, false lateralization with ictal SPECT may occur if the seizure ceases in the temporal lobe of origin while continuing in the contralateral temporal lobe at the time of tracer injection. For extratemporal lobe seizure, such as frontal and parietal lobe seizures, ictal SPECT has sensitivity as high as 90% in localizing seizures if ictal injection occurs shortly after ictal onset (ie, within 20 seconds).
Subtraction ictal SPECT co-registered to MRI
The sensitivity of ictal SPECT is increased significantly when ictal and interictal images are subtracted. This subtracted image is then superimposed on high-resolution MRI, which further increases the sensitivity and specificity of the interpretation. Surgical outcomes in patients whose seizure focus is localized with this technique are under study. More recently, postictal subtraction SPECT co-registered to MRI has been studied as a method of localizing the epileptogenic zone. Newer methods include statistical parametric mapping where a control database of inter-ictal SPECT scans are subtracted from the patient's ictal SPECT scan and a z score is generated. This subtraction image is subsequently co-registered to MRI.
Magnetic Source Imaging
Magnetoencephalography (MEG) detects the magnetic fields produced by the electrical currents of neuronal activity. Unlike the electrical currents of neuronal activity, which are extracellular, magnetic fields are produced by the intracellular currents of apical dendrites, which are recorded from the scalp by MEG. Unlike conventional EEG that detects radially oriented electrical activity that is attenuated in strength and spatially distorted by tissues between brain and scalp surface, magnetic fields are minimally affected by intervening tissue layers. Furthermore, MEG measures a subset of neuronal activity that is tangential to the scalp.
These magnetic dipoles generated by MEG are then superimposed on structural MR images creating magnetic source imaging (MSI). Numerous studies have shown that this technique is helpful in patients with neocortical epilepsies to map interictal epileptiform activity, which in conjunction with other noninvasive structural and imaging data, guide intracranial subdural grid placement to improve surgical outcome.
A large series of 455 patients showed that MSI identified the lobe to be treated in 89% of patients. In all extratemporal cases, MSI correctly identified the correct lobe. One might argue that MSI provides redundant data. However, in this study, MSI provided additional information about the epileptogenic zone in 35%, and it provided crucial information for surgical decision making in 11%.
Thus, MSI is a promising modality for seizure localization in that it can confirm the epileptogenic zone along with other functional imaging data, aid in the identification of a subtle cortical abnormality on MRI, and provide localizing information not obtainable from other imaging modalities. In this regard, it can either obviate the need for invasive monitoring in cases with a structural lesion without localizing or lateralizing ictal EEG data or guide intracranial subdural electrode placement to improve localization of the epileptogenic zone and improve seizure-free outcome.
The disadvantage to MSI is that it is limited to a few centers; it is performed in the outpatient setting in the United States where AEDs cannot always be tapered or discontinued safely; and recording time is limited, which reduces the chance of obtaining sufficient interictal data with the exception of patients who have frequent inter-ictal activity. Furthermore, although ictal MSI has been recorded and is highly localizing, the chance of capturing a seizure during a study is small. Some centers partially taper AEDs and/or give clonidine to enhance the yield of identifying interictal epileptiform activity.
MRS, Functional MRI, And Novel Structural MRI Technique
Proton magnetic resonance spectroscopic imaging
This technique is based on the principle that N -acetyl aspartate (NAA) is found primarily within neurons and precursor cells; a reduction in NAA usually is regarded as indicating loss or dysfunction of neurons. Creatinine (CR) and choline (Cho) are present in much higher concentrations in glia than in neurons. Patients with TLE have reductions in the NAA/(Cho + CR) ratio. This reduction has been shown to correlate with the presence of hippocampal sclerosis and to correctly lateralize the side of seizure onset in 97% of patients. About 20-40% of patients have bilateral metabolic disturbances, and preliminary evidence suggests that this finding is associated with a higher probability of surgical failure.
Recent developments in multivoxel Proton magnetic resonance spectroscopic imaging may be of further value.
Whether this technique is useful in patients with no lesions on MRI, particularly those with
extratemporal epilepsy, is less clear. MRS also has been used to measure lactate levels postictally, although its clinical utility has yet to be established.
Functional MRI
Functional MRI (fMRI) evaluates cerebral blood flow by looking at the difference between venous oxyhemoglobin and deoxyhemoglobin; this is called the blood oxygen level–dependent (BOLD) contrast technique. During cortical activation, cerebral blood flow to the eloquent cortex increases focally as a response to the stimulus, but oxygen extraction changes little. This causes a relative increased concentration of oxyhemoglobin and a relatively decreased concentration of deoxyhemoglobin draining the activated cortex. Deoxyhemoglobin is paramagnetic; it exerts magnetic susceptibility effects on local tissue, which are detected by T2-weighted imaging as decreased signal intensity. Oxyhemoglobin, on the other hand, is diamagnetic and has little effect on T2-weighted images.
Thus, cortical activation results in a relative decrease of the lowered signal intensity produced by the decreased concentration of deoxyhemoglobin, which leads to a relative increase in signal in the activated cortex relative to contiguous cortex. fMRI has been used to map language, motor function, and interictal spikes. It also may be useful for seizure localization and has successfully mapped simple partial seizures. However, capturing seizures with fMRI is difficult, because seizures are unpredictable and complex partial seizures usually are associated with movement that obscures the fMRI image.
Novel structural MRI techniques
Surface-coil MRI and 3-D surface rendering may increase the yield in identifying focal areas of cortical thickening. The use of multichannel phased array head coils is preferred over conventional quadrature coils. Other techniques, such as T1-weighted and T2-weighted inversion recovery, also may increase the sensitivity to identify subtle cortical malformations. 3T-phased array (PA) MRI can further increase the signal to noise ratio 6-8 fold as compared to nonphase array coil 1.5T MRI. A recent study found improved lesion detection with 3T PA MRI in patients with intractable epilepsy.
Preliminary data suggest that 3-D preoperative maps of hippocampi can help predict surgical outcome. However, future studies are needed to determine whether this will be an independent predictor of surgical outcome. In novel techniques such as voxel-based morphometry (VBM), a whole brain gray matter voxel-based comparison is made of the patient and the control group. A z score map is then generated for the patient. This method demonstrates enhanced sensitivity in identifying subtle gray matter abnormalities and for identifying additional areas of gray matter abnormalities in patients with focal cortical dysplasia.
Summary And Conclusions
The care of patients with epilepsy, as that of most other patients with neurological diseases, has been revolutionized by developments in neuroimaging over the past 15 years. This has led to a far more accurate diagnosis of the pathologic substrate of epilepsy, which is essential for accurate classification, determination of prognosis, and surgical candidacy. Structural MR imaging has greatly reduced the need for invasive EEG evaluation of patients with intractable epilepsy and has therefore reduced morbidity. Although general consensus exists among neurologists specializing in epilepsy as to when and what type of neuroimaging studies should be performed in patients with epilepsy, these views have not yet been accepted completely by general neurologic and medical practitioners despite the recommendations of the International League Against Epilepsy (1997).
Evaluation of a first seizure
The author's practice is that all patients presenting with a first seizure in adulthood be evaluated with high-resolution epilepsy protocol MRI. In one study of 300 consecutive patients presenting with a first seizure, an epileptogenic lesion was identified by MR imaging in 14%. In another study, MR imaging detected etiologically relevant structural abnormalities in 12.7%. Thus, even in patients with a single seizure, a significant number of patients had an abnormal MRI that correlated with their epilepsy.
In intractable epilepsy, on the other hand, MR imaging identifies the pathologic substrate in 82-86% of patients. Children presenting with focal seizures also should be evaluated by MRI scanning. Children or adults with clinically evident idiopathic generalized epilepsies (eg, childhood absence epilepsy, juvenile myoclonic epilepsy) probably can forgo MRI scanning, although in clinical practice being confident of the diagnosis at the time of presentation is often difficult. The presence of MRI
abnormalities in patients with new-onset epilepsy is predictive of seizure recurrence and also predicts lack of seizure control with medical therapy.
Most importantly, MRI scanning in this setting will detect abnormalities, such as brain tumors, arteriovenous malformations, or cryptogenic infarctions that may require further diagnostic and therapeutic interventions to prevent neurological deterioration. CT scanning is used widely in patients presenting with a first seizure, usually because it is more readily available than MRI in most emergency departments. However, patients should be referred for an epilepsy protocol MRI scanning as an outpatient.
Evaluation of medically intractable epilepsy
All patients with medically intractable partial epilepsy (defined as continued seizures despite therapy with therapeutic doses of appropriate antiepileptic drugs) should be referred for MRI scanning using an epilepsy protocol. A significant number of these patients will have hippocampal sclerosis, malformations of cortical development, developmental or glial tumors, or other abnormalities such as vascular malformations and encephalomalacia, which are predictive of continued failure of medical therapy.
Further, if ictal EEG recordings establish that these lesions correlate with the site of seizure onset, surgical resection of the lesion and associated epileptogenic cortex is often curative. Thus, patients with intractable epilepsy defined as having failed at least two antiepileptic drugs with persistent disabling seizures and MRI lesions should be put on a "fast track" for referral to a specialized epilepsy center for consideration of surgical management. If the an epilepsy protocol MRI is normal, further trials of medical therapy may be indicated, although persistence of medically intractable seizures for 2 years or more is also an indication for referral to a specialized epilepsy center, regardless of MRI findings. These patients can also be surgical candidates if seizures are well localized and functional imaging data (see below) are concordant with the seizure semiology and EEG findings.
Functional neuroimaging (ie, SPECT, PET, MRS, MSI) should be reserved for use by specialized epilepsy centers in the evaluation of the subset of patients with medically intractable epilepsy. These studies are helpful in the following cases: (1) abnormal MRI with nonlocalizing EEG or discordant clinical semiology and/or EEG findings, (2) multifocal MRI (ie, tuberous sclerosis), (3) normal MRI with lateralized or localized EEG findings, and (4) normal MRI with nonlocalizing EEG findings.
In summary, high-resolution dedicated epilepsy protocol imaging has revolutionized the evaluation, classification, and management of epilepsy. It has allowed the determination of the structural substrate for epilepsy in many cases and helps to determine prognosis for remission or intractability. Furthermore, structural neuroimaging and functional neuroimaging have enabled a greater number of medically intractable epilepsy patients to become surgery candidates with an increased chance for seizure freedom and improved quality of life.
Keywords
medically refractory epilepsy, focal epilepsy, neuroimaging in epilepsy surgery
More on Neuroimaging in Epilepsy Surgery |
| References |
References
Aghakhani Y, Kinay D, Gotman J, et al. The role of periventricular nodular heterotopia in epileptogenesis. Brain. Mar 2005;128(Pt 3):641-51. [Medline].
Barkovich AJ, Kuzniecky RI, Jackson GD, et al. A developmental and genetic classification for malformations of cortical development. Neurology. Dec 27 2005;65(12):1873-87. [Medline].
Barkovich AJ, Raybaud CA. Malformations of cortical development. Neuroimaging Clin N Am. Aug 2004;14(3):401-23. [Medline].
Barkovich AJ. Malformations of Cortical Development. Magnetic Resonance Imaging in Epilepsy: Neuroimaging Techniques. 2nd ed. 2005:221-248.
Bastos AC, Comeau RM, Andermann F, et al. Diagnosis of subtle focal dysplastic lesions: curvilinear reformatting from three-dimensional magnetic resonance imaging. Ann Neurol. Jul 1999;46(1):88-94. [Medline].
Berg AT, Langfitt J, Shinnar S, et al. How long does it take for partial epilepsy to become intractable?. Neurology. Jan 28 2003;60(2):186-90. [Medline].
Bernasconi A, Martinez V, Rosa-Neto P, et al. Surgical resection for intractable epilepsy in "double cortex" syndrome yields inadequate results. Epilepsia. Sep 2001;42(9):1124-9. [Medline].
Bernasconi N, Bernasconi A, Caramanos Z, et al. Entorhinal cortex atrophy in epilepsy patients exhibiting normal hippocampal volumes. Neurology. May 22 2001;56(10):1335-9. [Medline].
Bernasconi N, Kinay D, Andermann F. Analysis of shape and positioning of the hippocampal formation: an MRI study in patients with partial epilepsy and healthy controls. Brain. Oct 2005;128(Pt 10):2442-52. [Medline].
Bocti C, Robitaille Y, Diadori P, et al. The pathological basis of temporal lobe epilepsy in childhood. Neurology. Jan 28 2003;60(2):191-5. [Medline].
Bower SP, Vogrin SJ, Morris K, et al. Amygdala volumetry in "imaging-negative" temporal lobe epilepsy. J Neurol Neurosurg Psychiatry. Sep 2003;74(9):1245-9. [Medline].
Brinkmann BH, O''Brien TJ, Mullan BP, et al. Subtraction ictal SPECT coregistered to MRI for seizure focus localization in partial epilepsy. Mayo Clin Proc. Jun 2000;75(6):615-24. [Medline].
Bronen RA, Fulbright RK, Spencer DD, et al. Refractory epilepsy: comparison of MR imaging, CT, and histopathologic findings in 117 patients. Radiology. Oct 1996;201(1):97-105. [Medline].
Bronen RA, Knowlton R, Garwood M. High Resolution Imaging of Epilepsy. Epilepsia. 2002;43(suppl):11-18.
Bronen RA, Fulbright RK, Kim JH, et al. A systematic approach for interpreting MR images of the seizure patient. AJR Am J Roentgenol. Jul 1997;169(1):241-7. [Medline].
Bronen RA, Spencer DD, Fulbright RK, et al. Cerebrospinal fluid cleft with cortical dimple: MR imaging marker for focal cortical dysgenesis. Radiology. Mar 2000;214(3):657-63. [Medline].
Bronen RA, Cheung G, Charles JT, et al. Imaging findings in hippocampal sclerosis: correlation with pathology. AJNR Am J Neuroradiol. Sep-Oct 1991;12(5):933-40. [Medline].
Bronen RA, Fulbright RK, Spencer DD, et al. MR characteristics of neoplasms and vascular malformations associated with epilepsy. Magn Reson Imaging. 1995;13(8):1153-62. [Medline].
Burdette DE, Sakurai SY, Henry TR, et al. Temporal lobe central benzodiazepine binding in unilateral mesial temporal lobe epilepsy. Neurology. May 1995;45(5):934-41. [Medline].
Burneo JG, Kuzniecky RI, Bebin M, Knowlton RC. Cortical reorganization in malformations of cortical development: a magnetoencephalographic study. Neurology. Nov 23 2004;63(10):1818-24. [Medline].
Cendes F, Cook MJ, Watson C, et al. Frequency and characteristics of dual pathology in patients with lesional epilepsy. Neurology. Nov 1995;45(11):2058-64. [Medline].
Chan S, Chin SS, Nordli DR, et al. Prospective magnetic resonance imaging identification of focal cortical dysplasia, including the non-balloon cell subtype. Ann Neurol. Nov 1998;44(5):749-57. [Medline].
Chugani DC, Muzik O, Chakraborty P, et al. Human brain serotonin synthesis capacity measured in vivo with alpha-[C-11]methyl-L-tryptophan. Synapse. Jan 1998;28(1):33-43. [Medline].
Chugani HT, Chugani DC. Imaging of serotonin mechanisms in epilepsy. Epilepsy Curr. Nov-Dec 2005;5(6):201-6. [Medline].
Colliot O, Bernasconi N, Khalili N, et al. Individual voxel-based analysis of gray matter in focal cortical dysplasia. Neuroimage. Jan 1 2006;29(1):162-71. [Medline].
Commission on Neuroimaging of the International League Against Epilepsy. Recommendations for neuroimaging of patients with epilepsy. Epilepsia. Nov 1997;38(11):1255-6. [Medline].
Detre JA, Sirven JI, Alsop DC, et al. Localization of subclinical ictal activity by functional magnetic resonance imaging: correlation with invasive monitoring. Ann Neurol. Oct 1995;38(4):618-24. [Medline].
Drzezga A, Arnold S, Minoshima S, et al. 18F-FDG PET studies in patients with extratemporal and temporal epilepsy: evaluation of an observer-independent analysis. J Nucl Med. May 1999;40(5):737-46. [Medline].
Ebersole JS. Magnetoencephalography/magnetic source imaging in the assessment of patients with epilepsy. Epilepsia. 1997;38 Suppl 4:S1-5. [Medline].
Eliashiv DS, Elsas SM, Squires K, et al. Ictal magnetic source imaging as a localizing tool in partial epilepsy. Neurology. Nov 26 2002;59(10):1600-10. [Medline].
Engel J, Henry TR, Swartz BE. Positron emission tomography in frontal lobe epilepsy. Adv Neurol. 1995;66:223-38; discussion 238-41. [Medline].
Fedi M, Reutens DC, Andermann F, et al. alpha-[11C]-Methyl-L-tryptophan PET identifies the epileptogenic tuber and correlates with interictal spike frequency. Epilepsy Res. Jan 2003;52(3):203-13. [Medline].
Fedi M, Reutens D, Okazawa H, et al. Localizing value of alpha-methyl-L-tryptophan PET in intractable epilepsy of neocortical origin. Neurology. Nov 13 2001;57(9):1629-36. [Medline].
Fischer MJ, Scheler G, Stefan H. Utilization of magnetoencephalography results to obtain favourable outcomes in epilepsy surgery. Brain. Jan 2005;128(Pt 1):153-7. [Medline].
Grant PE, Barkovich AJ, Wald LL, et al. High-resolution surface-coil MR of cortical lesions in medically refractory epilepsy: a prospective study. AJNR Am J Neuroradiol. Feb 1997;18(2):291-301. [Medline].
Grattan-Smith JD, Harvey AS, Desmond PM, Chow CW. Hippocampal sclerosis in children with intractable temporal lobe epilepsy: detection with MR imaging. AJR Am J Roentgenol. Nov 1993;161(5):1045-8. [Medline].
Harvey AS, Grattan-Smith JD, Desmond PM, et al. Febrile seizures and hippocampal sclerosis: frequent and related findings in intractable temporal lobe epilepsy of childhood. Pediatr Neurol. Apr 1995;12(3):201-6. [Medline].
Ho SS, Berkovic SF, Berlangieri SU, et al. Comparison of ictal SPECT and interictal PET in the presurgical evaluation of temporal lobe epilepsy. Ann Neurol. Jun 1995;37(6):738-45. [Medline].
Jack CR, Sharbrough FW, Cascino GD, et al. Magnetic resonance image-based hippocampal volumetry: correlation with outcome after temporal lobectomy. Ann Neurol. Feb 1992;31(2):138-46. [Medline].
Jack CR, Rydberg CH, Krecke KN, et al. Mesial temporal sclerosis: diagnosis with fluid-attenuated inversion-recovery versus spin-echo MR imaging. Radiology. May 1996;199(2):367-73. [Medline].
Jack CR Jr. Magnetic resonance imaging in epilepsy. Mayo Clin Proc. Jul 1996;71(7):695-711. [Medline].
Jackson GD, Kuzniecky RI, Cascino GD. Hippocampal sclerosis without detectable hippocampal atrophy. Neurology. Jan 1994;44(1):42-6. [Medline].
Jackson GD, Connelly A, Cross JH, et al. Functional magnetic resonance imaging of focal seizures. Neurology. May 1994;44(5):850-6. [Medline].
Jackson GD, Breillmann RS, Kuzniecky RI. Temporal lobe epilepsy. In: Kuzniecky RI, Jackson GD, eds. Magnetic Resonance in Epilepsy. Neuroimaging Techniques. 2nd ed. Amstersam: Elsevier Academic Press; 2005:. 99-176.
Juhasz C, Chugani DC, Padhye UN, et al. Evaluation with alpha-[11C]methyl-L-tryptophan positron emission tomography for reoperation after failed epilepsy surgery. Epilepsia. Feb 2004;45(2):124-30. [Medline].
Juhasz C, Chugani DC, Muzik O, et al. Alpha-methyl-L-tryptophan PET detects epileptogenic cortex in children with intractable epilepsy. Neurology. Mar 25 2003;60(6):960-8. [Medline].
Kagawa K, Chugani DC, Asano E, et al. Epilepsy surgery outcome in children with tuberous sclerosis complex evaluated with alpha-[11C]methyl-L-tryptophan positron emission tomography (PET). J Child Neurol. May 2005;20(5):429-38. [Medline].
King MA, Newton MR, Jackson GD, et al. Epileptology of the first-seizure presentation: a clinical, electroencephalographic, and magnetic resonance imaging study of 300 consecutive patients. Lancet. Sep 26 1998;352(9133):1007-11. [Medline].
Knake S, Triantafyllou C, Wald LL, et al. 3T phased array MRI improves the presurgical evaluation in focal epilepsies: a prospective study. Neurology. Oct 11 2005;65(7):1026-31. [Medline].
Knowlton RC, Laxer KD, Aminoff MJ, et al. Magnetoencephalography in partial epilepsy: clinical yield and localization accuracy. Ann Neurol. Oct 1997;42(4):622-31. [Medline].
Knowlton RC, Laxer KD, Ende G, et al. Presurgical multimodality neuroimaging in electroencephalographic lateralized temporal lobe epilepsy. Ann Neurol. Dec 1997;42(6):829-37. [Medline].
Knowlton RC, Lawn ND, Mountz JM, Kuzniecky RI. Ictal SPECT analysis in epilepsy: subtraction and statistical parametric mapping techniques. Neurology. Jul 13 2004;63(1):10-5. [Medline].
Knowlton RC. The role of FDG-PET, ictal SPECT, and MEG in the epilepsy surgery evaluation. Epilepsy Behav. Feb 2006;8(1):91-101. [Medline].
Knowlton RC, Shih J. Magnetoencephalography in epilepsy. Epilepsia. 2004;45 Suppl 4:61-71. [Medline].
Koepp MJ, Richardson MP, Labbé C, et al. 11C-flumazenil PET, volumetric MRI, and quantitative pathology in mesial temporal lobe epilepsy. Neurology. Sep 1997;49(3):764-73. [Medline].
Koepp MJ, Hand KS, Labbé C, et al. In vivo [11C]flumazenil-PET correlates with ex vivo [3H]flumazenil autoradiography in hippocampal sclerosis. Ann Neurol. May 1998;43(5):618-26. [Medline].
Krakow K, Woermann FG, Symms MR, et al. EEG-triggered functional MRI of interictal epileptiform activity in patients with partial seizures. Brain. Sep 1999;122 ( Pt 9):1679-88. [Medline].
Kuzniecky R, Jackson GD. Magnetic Resonance Imaging in Epilepsy. New York: Raven Press;1995.
Kuzniecky R, Cascino GD, Palmini A, et al. Structural neuroimaging. In: Engel J, ed. Surgical Treatment of the Epilepsies. 2nd ed. New York: Raven Press;1993:197-209.
Kuzniecky R, Hugg JW, Hetherington H, et al. Relative utility of 1H spectroscopic imaging and hippocampal volumetry in the lateralization of mesial temporal lobe epilepsy. Neurology. Jul 1998;51(1):66-71. [Medline].
Kuzniecky R, Hugg J, Hetherington H, et al. Predictive value of 1H MRSI for outcome in temporal lobectomy. Neurology. Sep 11 1999;53(4):694-8. [Medline].
Kuzniecky RI, Bilir E, Gilliam F, et al. Multimodality MRI in mesial temporal sclerosis: relative sensitivity and specificity. Neurology. Sep 1997;49(3):774-8. [Medline].
Laxer KD, Garcia PA, Ng TC. Presurgical evaluation: 1H magnetic resonance spectroscopy. In: Wyllie E, ed. The Treatment of Epilepsy: Principles and Practice. 2nd ed. Baltimore: Williams & Wilkins;1996:1030-5.
Li LM, Fish DR, Sisodiya SM, et al. High resolution magnetic resonance imaging in adults with partial or secondary generalised epilepsy attending a tertiary referral unit. J Neurol Neurosurg Psychiatry. Oct 1995;59(4):384-7. [Medline].
Li LM, Cendes F, Andermann F, et al. Surgical outcome in patients with epilepsy and dual pathology. Brain. May 1999;122 ( Pt 5):799-805. [Medline].
Li LM, Dubeau F, Andermann F, et al. Periventricular nodular heterotopia and intractable temporal lobe epilepsy: poor outcome after temporal lobe resection. Ann Neurol. May 1997;41(5):662-8. [Medline].
Lin JJ, Salamon N, Dutton RA, et al. Three-dimensional preoperative maps of hippocampal atrophy predict surgical outcomes in temporal lobe epilepsy. Neurology. Oct 11 2005;65(7):1094-7. [Medline].
Mamelak AN, Lopez N, Akhtari M. Magnetoencephalography-directed surgery in patients with neocortical epilepsy. J Neurosurg. Oct 2002;97(4):865-73. [Medline].
McNally KA, Paige AL, Varghese G, et al. Localizing value of ictal-interictal SPECT analyzed by SPM (ISAS). Epilepsia. Sep 2005;46(9):1450-64. [Medline].
Meencke HJ, Veith G. Hippocampal sclerosis in epilepsy. In: Luders H, ed. Epilepsy Surgery. New York: Raven Press;1991:705-15.
Minassian BA, Otsubo H, Weiss S, et al. Magnetoencephalographic localization in pediatric epilepsy surgery: comparison with invasive intracranial electroencephalography. Ann Neurol. Oct 1999;46(4):627-33. [Medline].
Moseley ME, Glover GH. Functional MR imaging. Capabilities and limitations. Neuroimaging Clin N Am. May 1995;5(2):161-91. [Medline].
O'Brien TJ, Newton MR, Cook MJ, et al. Hippocampal atrophy is not a major determinant of regional hypometabolism in temporal lobe epilepsy. Epilepsia. Jan 1997;38(1):74-80. [Medline].
O'Brien TJ, So EL, Mullan BP, et al. Subtraction SPECT co-registered to MRI improves postictal SPECT localization of seizure foci. Neurology. Jan 1 1999;52(1):137-46. [Medline].
Palmini A, Sam MC, Cascino GD. MRI in the evaluation of epilepsy surgery. In: Wyllie E, ed. The Treatment of Epilepsy: Principles and Practice. 2nd ed. Baltimore: Williams & Wilkins;1996:1006-16.
Raymond AA, Fish DR, Stevens JM, et al. Association of hippocampal sclerosis with cortical dysgenesis in patients with epilepsy. Neurology. Oct 1994;44(10):1841-5. [Medline].
Raymond AA, Fish DR, Sisodiya SM, et al. Abnormalities of gyration, heterotopias, tuberous sclerosis, focal cortical dysplasia, microdysgenesis, dysembryoplastic neuroepithelial tumour and dysgenesis of the archicortex in epilepsy. Clinical, EEG and neuroimaging features in 100 adult patients. Brain. Jun 1995;118 ( Pt 3):629-60. [Medline].
Richardson MP, Koepp MJ, Brooks DJ, et al. Benzodiazepine receptors in focal epilepsy with cortical dysgenesis: an 11C-flumazenil PET study. Ann Neurol. Aug 1996;40(2):188-98. [Medline].
Richardson MP, Koepp MJ, Brooks DJ, Duncan JS. 11C-flumazenil PET in neocortical epilepsy. Neurology. Aug 1998;51(2):485-92. [Medline].
Rivera PP, Willinsky RA, Porter PJ. Intracranial cavernous malformations. Neuroimaging Clin N Am. Feb 2003;13(1):27-40. [Medline].
Ryvlin P, Philippon B, Cinotti L, et al. Functional neuroimaging strategy in temporal lobe epilepsy: a comparative study of 18FDG-PET and 99mTc-HMPAO-SPECT. Ann Neurol. Jun 1992;31(6):650-6. [Medline].
Savic I, Thorell JO, Roland P. [11C]flumazenil positron emission tomography visualizes frontal epileptogenic regions. Epilepsia. Dec 1995;36(12):1225-32. [Medline].
Savic I, Lindstrom P, Gulyas B, et al. Limbic reductions of 5-HT1A receptor binding in human temporal lobe epilepsy. Neurology. Apr 27 2004;62(8):1343-51. [Medline].
Scott CA, Fish DR, Smith SJ, et al. Presurgical evaluation of patients with epilepsy and normal MRI: role of scalp video-EEG telemetry. J Neurol Neurosurg Psychiatry. Jan 1999;66(1):69-71. [Medline].
Semah F, Picot MC, Adam C, et al. Is the underlying cause of epilepsy a major prognostic factor for recurrence?. Neurology. Nov 1998;51(5):1256-62. [Medline].
Sisodiya SM, Stevens JM, Fish DR, et al. The demonstration of gyral abnormalities in patients with cryptogenic partial epilepsy using three-dimensional MRI. Arch Neurol. Jan 1996;53(1):28-34. [Medline].
Tassi L, Colombo N, Cossu M, et al. Electroclinical, MRI and neuropathological study of 10 patients with nodular heterotopia, with surgical outcomes. Brain. Feb 2005;128(Pt 2):321-37. [Medline].
Theodore WH. Functional neuroimaging. In: Wyllie E, ed. The Treatment of Epilepsy: Principles and Practice. 2nd ed. Baltimore: Williams & Wilkins;1996:1017-29.
Theodore WH, Sato S, Kufta CV, et al. FDG-positron emission tomography and invasive EEG: seizure focus detection and surgical outcome. Epilepsia. Jan 1997;38(1):81-6. [Medline].
Tilz C, Hummel C, Kettenmann B, Stefan H. Ictal onset localization of epileptic seizures by magnetoencephalography. Acta Neurol Scand. Oct 2002;106(4):190-5. [Medline].
Toczek MT, Carson RE, Lang L, et al. PET imaging of 5-HT1A receptor binding in patients with temporal lobe epilepsy. Neurology. Mar 11 2003;60(5):749-56. [Medline].
Van Paesschen W. Qualitative and quantitative imaging of the hippocampus in mesial temporal lobe epilepsy with hippocampal sclerosis. Neuroimaging Clin N Am. Aug 2004;14(3):373-400, vii. [Medline].
Vattipally VR, Bronen RA. MR imaging of epilepsy: strategies for successful interpretation. Neuroimaging Clin N Am. Aug 2004;14(3):349-72. [Medline].
Von Oertzen J, Urbach H, Jungbluth S, et al. Standard magnetic resonance imaging is inadequate for patients with refractory focal epilepsy. J Neurol Neurosurg Psychiatry. Dec 2002;73(6):643-7. [Medline].
Wieshmann UC. Clinical application of neuroimaging in epilepsy. J Neurol Neurosurg Psychiatry. Apr 2003;74(4):466-70. [Medline].
Zubal IG, Spencer SS, Imam K, et al. Difference images calculated from ictal and interictal technetium-99m-HMPAO SPECT scans of epilepsy. J Nucl Med. Apr 1995;36(4):684-9. [Medline].
Further Reading
Keywords
medically refractory epilepsy, focal epilepsy, neuroimaging in epilepsy surgery
