Diagnostic Imaging
January 2000

MR and PET assist in workup of patients with epilepsy

Imaging’s importance steadily increases in the assessment, monitoring, and treatment of seizure disorders

By Linda J. Bagley, M.D., and Kim M. Cecil, Ph.D.

Epilepsy is a psychosocially devastating and sometimes life-threatening disorder that affects about 0.5% to 1% of the U.S. population.1 While advances have been made in medical therapy, between 5% and 20% of epilepsy cases remain medically intractable.2,3 Surgery is appropriate for some patients, including lesional resections, temporal lobectomies, callosotomies, hemispherectomies, and sub-pial transections.2 With the increasing use of surgical management, the role of imaging for localization of the seizure focus has increased in importance.

Relevant improvements in MR technology include FLAIR and volumetric sequences, surface coil imaging, perfusion imaging, and functional MRI (fMRI). MR spectroscopy can be used to identify areas of neuronal loss and/or metabolic abnormality, and nuclear medicine studies, including ictal and interictal SPECT and PET, can demonstrate areas of hypo- and/or hyperperfusion and metabolism.1,2,4

Conventional MRI of patients with epilepsy is tailored to detect common etiologies of seizure disorders: mesial temporal sclerosis (MTS), cortical dysplasia, migrational abnormalities, tumors, and vascular malformations. Particular attention is paid to the temporal lobe, which is the most common site of seizure onset in patients with partial epilepsy.2 Typical imaging protocols include high-resolution coronal fast spin-echo T2-weighted images, coronal FLAIR sequence, and coronal T1-weighted volumetric acquisition. Mesial temporal sclerosis, the most common cause of intractable partial complex seizures, is manifest by atrophy of the hippocampus, amygdala, and/or parahippocampal gyrus. Signal abnormality and/or architectural distortion may also be present (Figure 1).

Quantitative volumetric measurements of the hippocampus improve the detection of MTS on MR from 80%-90% to 90%-100%.2,4 The epilepsy imaging protocol also includes an axial T2-weighted gradient-echo sequence (TR 750/TE 40/10 flip angle, 2 NEX) to improve detection of otherwise occult vascular malformations, such as cavernous angiomas. Gadolinium-DTPA contrast medium is administered when there is a suspicion of neoplasia. Tumors presenting with medically intractable epilepsy are typically slow-growing and benign.2

Surface coil imaging has been used to improve detection of dysplasias and gyral abnormalities in patients with neocortical epilepsy. Multiple phased-array surface coils centered over the suspected eleptogenic area provide high-resolution images with high signal-to-noise ratios. Grant et al compared head and surface coil imaging in 25 patients with medically refractory partial neocortical epilepsy. The imaging protocols used included a coronal volumetric gradient-echo series and in most cases a fast spin-echo or spin-echo high-resolution T2-weighted sequence. They found that the surface coil images improved detection and definition of focal cortical lesions in 64% of the patients. Additional foci of dysplasia were seen, polymicrogyria was distinguished from pachygyria, and areas of T2 signal abnormality were more confidently classified as neoplastic or non-neoplastic.5

Multislice continuous arterial spin-labeled perfusion MRI has also been used in the study of patients with temporal lobe epilepsy. In preliminary work, Wolf et al found interictal asymmetries in perfusion of the medial temporal lobes in these patients. The authors also reported a trend toward correlation between the magnitude of perfusional asymmetry and seizure-free outcome of surgery.6

Most candidates for epilepsy surgery undergo preoperative Wada testing—cerebral angiography and intracarotid injections of amobarbitol followed by neuropsychological testing—in an effort to lateralize memory and language function and thereby minimize postoperative deficits.4 The role of functional MRI has been increasing in an attempt to replace this invasive procedure. Functional MRI employs a blood oxygen level-dependent (BOLD) technique. Deoxyhemoglobin has paramagnetic effects, and its presence thereby results in T2 signal loss. Oxygenated hemoglobin has minimal paramagnetic effect. Levels of blood oxygenation are increased in areas of cortical activation. Therefore, T2 signal loss due to the presence of deoxyhemoglobin is less pronounced in these activated areas. The BOLD technique has been used to localize visual and motor cortex and language centers. Efforts to improve localization of memory with fMRI continue.1

MR Spectroscopy

Proton MR spectroscopy complements conventional MRI in understanding disorders of the brain, including epilepsy. Proton MRS provides metabolic information by determining the presence and levels of select neurochemicals. In patients with suspected mesial temporal sclerosis, small regions of interest (voxels) are placed over the medial temporal lobes. Reduced levels of N-acetyl aspartate, found primarily in axons and neurons, have been reported in the temporal lobe containing the seizure focus. The myoinositol resonance, seen on short-echo MRS, has been associated with areas of gliosis and may be elevated in the temporal lobe ipsilateral to the seizure focus (Figure 2). Phosphorus spectroscopy can identify reduced levels of phosphocreatine and inorganic phosphate in the affected temporal lobe.

PET and SPECT provide physiologic information about the eleptogenic brain.1,2,4 When used with the radioisotope F-18 fluorodeoxyglucose, PET can measure glucose metabolism in neurons. Interictally, the temporal lobe ipsilateral to the seizure focus is hypometabolic. The sensitivity of FDG PET for identifying the abnormal temporal lobe in patients with partial epilepsy has been reported to be 70% to 91%.1 The sensitivity of this study for localizing the seizure focus is reduced in extra-temporal epilepsy, however. SPECT is performed with the radiotracer technetium-99m HMPAO, which is distributed in the brain in proportion to regional blood flow. Interictally, the sensitivity of SPECT is much lower than that of PET. However, when the isotope is injected during a seizure, increased blood flow typically results in increased uptake of the radiotracer, and sensitivity of the study for lateralizing the seizure focus has been reported to exceed 95%.1

Neurodegenerative Disease

The term neurodegenerative disorders encompasses a broad spectrum of disease processes, most of them idiopathic, which result in progressive disintegration of a portion or portions of the nervous system. Clinical presentations are variable but include dementia, movement disorders/weakness, visual loss, and language dysfunction. The imaging findings of these neurodegenerative diseases are often very nonspecific. Atrophy (focal or diffuse), mineral deposition, and/or abnormal signal intensity in the basal ganglia and/or various motor tracts may be seen in a large number of degenerative conditions, including Alzheimer’s disease, Parkinson’s disease, Huntington’s chorea, motor neuron diseases, such as amyotrophic lateral sclerosis (ALS), and metabolic disorders, such as Wilson’s disease and Leigh’s disease.

Volumetric studies, both CT and MRI, have been used in the evaluation of these degenerative disorders, including Alzheimer’s disease. Lower total brain volumes, higher total cerebrospinal fluid volumes, and accelerated changes in these values have been reported in Alzheimer’s patients. Specifically, reductions in the amount of gray matter, most notably in the temporal lobes, basal ganglia, thalami, and hippocampi, have been observed.7 Focal atrophic regions have also been identified in the spinocerebellar degenerative syndromes and in certain other neurodegenerative diseases.

As the conventional imaging findings of the neurodegenerative disorders are frequently nonspecific, magnetization transfer imaging (MTI) and MRS have been increasingly used in the initial diagnostic evaluation and in the continued monitoring of patients with these diseases. MTI is based on the principle that protons bound in macromolecular structures exhibit T1 relaxation coupling with protons in the aqueous phase. An off-resonance saturation pulse can be applied to selectively saturate the bound protons. Subsequent exchange of longitudinal magnetization with the free water protons leads to a resultant reduction in signal intensity detected from these free protons.8,9 The magnetization transfer ratio (MTR) provides a quantitative index of this MT effect10 and may be viewed as a quantitative measure of the structural integrity of tissues. In this way, MTI may provide data additional to that obtained with conventional MRI. Preliminary studies have revealed MTI to be sensitive for the detection of various degenerative processes, including demyelination and Wallerian degeneration.10-12

MTR abnormalities have been shown to be more widespread than T2 signal abnormalities in multiple sclerosis.12-14 Lexa’s study of Wallerian degeneration in the feline visual pathway revealed alterations in MTR values prior to the appearance of signal abnormalities on T2-weighted images, when histologic changes were evident by electron microscopy.11 Similarly, Kato et al reported statistically significant reductions in MTR values in the corticospinal tracts of ALS patients, when T2-weighted images often failed to disclose any abnormality.15

Proton MRS has been extensively used in the study of metabolic diseases. Regions of the brain, often within the basal ganglia, can be interrogated with MRS; typical voxel sizes are on the order of 4 to 8 cc. This noninvasive measure can provide within minutes information on the elevation of lactic acid, the most noted MRS feature of metabolic and mitochondrial disorders.16 Alterations in the levels of other metabolite resonances, such as NAA, creatine, and choline can also be demonstrated, depending on the disease.

In certain disease processes, changes may be noted on MRS prior to the appearance of signal abnormalities on conventional MRI. Chan et al reported a significant difference in NAA/creatine ratios obtained in motor cortex between patients with motor neuron diseases and controls. Abnormally reduced NAA/creatine ratios were found in 79% of the patients with ALS and in 67% of the patients with primary lateral sclerosis (PLS); the typical conventional imaging findings of increased signal intensity in the corticospinal tracts and central sulcus enlargement were seen in only 43% of the ALS patients and 24% of the PLS patients.17 In Parkinson’s disease, elevated lactate levels have been noted, while significant declines in NAA levels have not. This spectroscopic result may aid in differentiating Parkinson’s disease from other causes of dementia.18

MRS can also be useful in assessing the progression of a disease process. In Alzheimer’s disease, elevations of myoinositol levels and reductions in NAA levels have been observed.7 The myoinositol resonance, seen on short-echo MRS, has been associated with areas of gliosis as well as osmolytic dysfunction, depending on the disorder. Serial evaluations of patients with neurodegenerative disorders can indicate neuronal loss, demyelination, and gliosis.19

Newer and more quantitative methods, such as volumetric studies, MTI, and proton MRS, may complement MRI in the noninvasive diagnostic assessment and therapeutic monitoring of neurodegenerative disorders and metabolic diseases.

Dr. Bagley is an assistant professor of radiology at the University of Pennsylvania in Philadelphia. Dr. Cecil is a research associate and MRI spectroscopist at Children’s Hospital Medical Center in Cincinnati.


  1. Fisher RS, Stein A, Karis J. Epilepsy for the neuroradiologist. AJNR 1997;18:851-863
  2. Jack CR. Epilepsy: surgery and imaging. Radiology 1993;189:635-646.
  3. Andermann F. Brain structure and epilepsy: the impact of modern imaging. AJNR 1997;18:302-306.
  4. Arroyo S, Freeman JM. Epilepsy surgery in children: state of the art. Advances in Pediatrics 1994;41:53-81.
  5. Grant PE, Barkovich AJ, Wald LL, et al. High-resolution surface-coil MR of cortical lesions in medically refractory epilepsy: a prospective study. AJNR 1997;18:291-301.
  6. Wolf RL, Alsop DC, French JA, et al. Detection of mesial temporal lobe hypoperfusion in patients with temporal lobe epilepsy using multislice arterial spin labeled perfusion MRI. ISMRM (abstract) Philadelphia 1999;371.
  7. Lexa FJ, Trojanowski JQ, Braffman BH, Atlas SW. The aging brain and neurodegenerative disorders. In: Atlas SW, ed. Magnetic resonance imaging of the brain and spine. Philadelphia: Lippincott-Raven, 1996:803-870.
  8. Wolff SD, Balaban RS. Magnetization transfer contrast (MTC) and tissue water proton relaxation in vivo. Magn Reson in Med 1989;10:135-144.
  9. Wolff SD, Eng J, Balaban RS. Magnetization transfer contrast: method for improving contrast in gradient-recalled-echo images. Radiology 1991;179:133-137.
  10. Dousset V, Grossman RI, Ramer KN. Experimental allergic encephalomyelitis and multiple sclerosis: lesion characterization with magnetization transfer imaging. Radiology 1992;182:483-491.
  11. Lexa FJ, Grossman RI, Rosenquist AC. Wallerian degeneration in feline visual system: characterization with magnetization rate with histopathological correlation. AJNR 1994;15:201-212.
  12. Loevner LA, Grossman RI, Lexa FJ, Kessler D. Microscopic disease in normal-appearing white matter on conventional MRI in patients with multiple sclerosis: assessment with utilizing magnetization transfer measurements. Radiology 1995; 196:511-515.
  13. Van Buchem MA, McGowan JC, Kolson D, et al. Quantitative volumetric magnetization transfer analysis in multiple sclerosis: estimation of macroscopic and microscopic disease burden. Magn Reson in Med 1996;36:632-636.
  14. Van Buchem MA, Udupa JK, McGowan JC, et al. Global volumetric estimation of disease burden in multiple sclerosis based on magnetization transfer imaging. AJNR 1997;18:1287-1290.
  15. Kato Y, Matsumura K, Kinosada Y, et al. Detection of pyramidal tract lesions in amyotrophic lateral sclerosis with magnetization transfer measurements. AJNR 1997;18: 1541-1547.
  16. Wang Z, Zimmerman RA. Proton MR spectroscopy of pediatric brain metabolic disorders. In: Castillo M, ed. Proton MR spectroscopy of the brain, vol. 8. Philadelphia: W.B. Saunders, 1998:781-807.
  17. Chan S, Shungu DC, Douglas-Akinwande A, et al. Motor neuron diseases: comparison of single voxel proton MR spectroscopy of the motor cortex with MR imaging of the brain. Radiology 1999;212:763-769.
  18. Bowen BC, Block RE, Sanchez-Ramos J. Proton MR spectroscopy of the brain in 14 patients with Parkinson disease. AJNR 1995;16:61-68.
  19. Tzika AA, Ball WS, Vigneron DB, et al. Clinical proton MR spectroscopy of disease in childhood. AJNR 1993;14:1267-81.

Acknowledgment: The authors wish to thank Lisa Desidario, R.T., Delight Roberts, and Jacqueline French, M.D., for their assistance.