| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
Neuroradiology |
1 From the MR Unit, Department of Veterans Affairs Medical Center (P.V., N.S., G.B.M., M.W.W.) and Departments of Neurology (K.D.L.), Radiology (M.W.W.), Pharmaceutical Chemistry (G.B.M.), and Medicine (M.W.W.), University of California, San Francisco. Received October 12, 2001; revision requested January 8, 2002; revision received April 17; accepted May 24. K.D.L., G.B.M., and M.W.W. supported by National Institutes of Health grant ROI-NS31966. P.V. supported by DFG Forschungsstipendium VE 190/1-1. Address correspondence to P.V., Department of Clinical Research, Magnetic Resonance Spectroscopy and Methodology, MR-Center 1, University and Inselspital, Inselheimmatte, CH-3010 Bern, Switzerland (e-mail: peter.vermathen@insel.ch).
 |
ABSTRACT |
|---|
 TOP ABSTRACT INTRODUCTIONMATERIALS AND METHODS RESULTS DISCUSSIONREFERENCES |
|---|
MATERIALS AND METHODS: Eleven patients with mesial TLE and 13 control subjects were examined with multisection MR spectroscopic imaging. Three MR spectroscopic imaging sections were acquired. Thirteen brain regions in each hemisphere and the midbrain were analyzed in each patient, and the NAA to creatine-phosphocreatine (Cr) plus choline-containing compounds (Ch) (NAA/[Cr + Ch]) ratios were determined. In addition, hemispheric and whole-brain values were calculated and statistically analyzed.
RESULTS: The NAA/(Cr + Ch) ratio in the ipsilateral hippocampus was significantly reduced, compared with that in the contralateral hippocampus (P < .002) and compared with that in control subjects (P < .03), confirming findings in previous studies. In patients, whole-brain NAA/(Cr + Ch) ratio outside the hippocampus was significantly lower than that in control subjects (P < .002). For the ipsilateral hemisphere in patients, NAA/(Cr + Ch) ratio was significantly lower than that in control subjects (P < .0002). Comparisons between individual brain regions revealed trends toward lower NAA/(Cr + Ch) ratios in many areas of the ipsilateral and, to a lesser extent, the contralateral hemisphere outside the hippocampus and temporal lobe, suggesting diffuse impairment.
CONCLUSION: Results suggest that repeated seizure activity damages neurons outside of the seizure focus.
© RSNA, 2002
Index terms: Brain, metabolism, 1341.99 • Brain, MR, 1341.121411, 1341.12145 • Epilepsy, 1341.83 • Hippocampus, 1341.83 • Magnetic resonance (MR), spectroscopy, 1341.12145
|
INTRODUCTION |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODS RESULTS DISCUSSIONREFERENCES |
|---|
Several observations suggest that mesial TLE may affect brain regions outside of the medial temporal lobe. In a neuropathologic study in patients with TLE, Margerison and Corsellis (10) observed hypoxic damage that scattered far beyond the confines of a single temporal lobe. In addition, extrahippocampal volume abnormalities were observed at MR imaging performed in patients with TLE (11-13). The volume deficits were bilateral and did not lateralize the epileptogenic hemisphere. PET study findings showed metabolic abnormalities in regions remote from the focus (14-18). Diffuse metabolic changes in the contralateral hemisphere in patients with mesial TLE were observed in a phosphorus 31 (31P) MR spectroscopic imaging study (19). Finally, we (2,3) and others (7,9,20,21) reported reductions of NAA concentrations in the contralateral hippocampus in patients with mesial TLE. These reductions of NAA concentrations in the contralateral hippocampus could be due to bilateral medial temporal sclerosis or repeated seizure activity that might cause neuronal loss or injury, as reflected by a decreased NAA concentration. In addition, several groups (20–23) have reported that the contralateral hippocampal NAA concentration increased after surgery for epilepsy. Altogether these findings suggest that repeated seizures might produce temporary metabolic changes or damage to neurons outside the focus and that this might result in reversible NAA concentration decreases. If repeated seizure activity is responsible for reduced contralateral NAA concentrations, then NAA concentrations might also be reduced in other regions outside the hippocampus. This concept might help explain the general cognitive impairment observed in patients with mesial TLE (24,25), which is not well understood.
Findings in an MR spectroscopic imaging study of Li et al (26) indicated a reduced NAA to creatine-phosphocreatine (Cr) ratio in frontocentroparietal regions in patients with TLE. The overall purpose of our study was to determine whether MR spectroscopic imaging can be used to detect metabolic changes, especially decreased NAA concentrations in different areas outside the medial temporal lobe in patients with mesial TLE, that are consistent with neuropathologic findings of extratemporal neuronal impairment (10). The specific goals were to verify previous findings of reduced NAA concentrations in the ipsilateral hippocampus at multisection MR spectroscopic imaging in patients with mesial TLE; to determine whether NAA concentrations are reduced in other brain regions outside the hippocampus; and if such changes exist, to determine whether NAA concentration changes are widespread or localized and whether they are mainly unilateral or bilateral.
|
MATERIALS AND METHODS |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODS RESULTS DISCUSSIONREFERENCES |
|---|
Only patients whose ictal recordings demonstrated either localized voltage attenuation or rhythmic sharp activity that preceded or coincided with the clinical onset of the seizure were included. All patients underwent separate independent MR imaging, and findings were interpreted by a neuroradiologist. MR imaging results were consistent with electroencephalographic results in nine of 11 patients and were classified as not lateralized in two patients. In three of 11 patients, no hippocampal atrophy was detected. The seizure frequency was between one and 30 seizures per month (mean, 7 seizures ± 8). The duration of epilepsy ranged from 15 to 41 years (mean, 29 years ± 9). Six of the 11 patients underwent surgery for seizures, and all six patients were seizure free more than 1 year since surgery.
The patients were compared with 13 healthy control subjects who were identified as healthy on the basis of their own statements and a review of their medical histories. Potential control subjects with medical conditions known to affect the central nervous system, with a history of spells that might suggest prior seizures, or with a history of any involvement of the central nervous system were excluded. Control subjects included five men and eight women. The age range was 19-46 years (mean age, 32.5 years ± 7.9).
Measurements
Proton multisection MR spectroscopic imaging spectra were acquired with a 1.5-T unit (Magnetom Vision; Siemens, Erlangen, Germany) by using a standard circularly polarized head coil. After scout imaging in coronal and sagittal orientations, T1-weighted spin-echo MR images (repetition time msec/echo time msec, 500/14; section thickness, 3 mm) were obtained with angulation parallel to the long axis of the hippocampus for localization in the transverse plane. In addition, an oblique transverse T2-weighted spin-echo data set (2,550/20, 80; section thickness, 3 mm) was acquired with angulation along the optical nerve.
After global shimming, three 15-mm sections were measured. One section was acquired with angulation parallel to the T1-weighted images and included both hippocampi (Fig 1). The other two sections were acquired with alignment parallel to the T2-weighted images; the lower section just included the corpus callosum and the other was slightly above the corpus callosum (Fig 1). Section-selective inversion pulses (inversion time, 170 msec) were used to reduce contamination of the MR spectra with lipid resonances. The equivalent of 36 x 36 phase-encoding steps over a circular k-space region was used (28), and this yielded a nominal voxel size of 8 x 8 x 15 mm3. The data were acquired with a repetition time of 1,800 msec and an echo time of 140 msec, and the measurement time was 35 minutes.
|
Figure 1 displays the regions for multisection MR spectroscopic imaging voxel selection. The selection of individual voxels and processing of the spectra were performed with blinding to the seizure lateralization findings (one author [P.V.] was responsible for voxel selection; another [G.B.M.] was responsible for blinding). The voxel selection was performed by using anatomic landmarks marked on a template. Although the possibility of slight voxel positioning variations was associated with this procedure, the blinded analysis ensured an unbiased processing.
A total of 27 single voxels from 13 different brain regions in each hemisphere and one from the midbrain were analyzed in each patient by one of us (P.V.). The regions were chosen before any results were known, and they were chosen with the main focus on regions (except the midbrain) that were most likely to enhance the lateralization ability of MR spectroscopic imaging. The midbrain, in addition, was analyzed to investigate possible nonlateralizing metabolic changes. As a threshold for inclusion of spectra, we used a signal-to-noise ratio of at least 4 for NAA signal intensities and a ratio of less than 1.2 and greater than 0.8 between the fitted line width of NAA and the mean line width of Cr and Ch.
The regions were chosen to include predominantly gray matter, although white matter was included to some degree in several regions. The spectral quality in the most anterior part of the lateral temporal lobe was generally inadequate for spectral quantitation, and this region, therefore, was excluded from the analysis. Brain regions for voxel selection included the left and right hippocampi, the midbrain and the occipital regions in the section that included the hippocampus, and the spectra from the left and right frontal and parietal regions in the two upper sections (Fig 1).
Calculation of NAA/(Cr + Ch) Ratio
NAA/(Cr + Ch) ratios were calculated for all selected voxel positions. Two different analyses were performed.
Analysis 1.—In each subject (patients and control subjects), a single mean NAA/(Cr + Ch) ratio (whole-brain ratio) and in all patients an ipsilateral and contralateral hemispheric mean value were calculated from the ratios of all voxels outside the hippocampus. The NAA/(Cr + Ch) ratios of the different brain regions were first normalized by weighting with the mean value of control subjects at the same position. This procedure accounts for inherent regional NAA/(Cr + Ch) ratio differences when the whole-brain and hemispheric mean values were calculated and helps to avoid calculation of different values because of missing values in some regions. The normalized ratios from all selected voxels outside the hippocampus (ratios from right and left sides of 11 different brain regions) were averaged to calculate the whole-brain ratio in each subject. In addition, in each patient, the ipsilateral and contralateral hemispheric mean values were calculated by averaging all NAA/(Cr + Ch) ratios of one hemisphere outside the hippocampus.
Analysis 2.—A mean NAA/(Cr + Ch) ratio for each position in the brain was calculated for each group (ie, the ipsilateral side in patients, the contralateral side in patients, and the control subjects) by averaging all ratios of an individual voxel position without normalization.
Statistical Analysis
Statistical analysis of the MR spectroscopic imaging data was performed by using one-way analysis of variance with software (SAS; SAS Institute, Cary, NC) to compare the NAA/(Cr + Ch) ratios in all three groups, that is, the ipsilateral hemisphere in patients, the contralateral hemisphere in patients, and the control subjects. For comparisons of pairs of groups, one-tailed paired (ipsilateral side in patients vs contralateral side in patients) and unpaired (patients vs control subjects) t tests were performed. The use of a one-tailed instead of a two-tailed t test was justified on the basis of previously reported NAA/(Cr + Ch) ratio reductions in patients with TLE. A P value of less than .05, after correction for multiple comparisons according to Bonferroni (SAS/SAT User’s Guide, release 6.03 edition, 1988), was considered to indicate a statistically significant difference for comparisons of whole-brain and hemispheric mean values between patients and control subjects.
In contrast, comparisons for all different individual brain regions among patients (ipsilateral side and contralateral side) and control subjects were presented without corrections for multiple comparisons to possibly reveal subtle differences despite numerous comparisons. However, these findings are presented as trends only. For the comparisons of spectra from the hippocampal region, no corrections for multiple comparisons were necessary, because this analysis was based on a different hypothesis. Pearson product moment correlation coefficients were used to test correlations. All data were presented as the mean ± 1 SD.
No correlation analyses between outcome after surgery and MR spectroscopic imaging results could be performed, because all patients who underwent surgery were seizure free. Furthermore, no correlation analyses between neuropsychiatric data and MR spectroscopic imaging results were performed, because neuropsychiatric data were available for only seven of the 11 patients, and of these seven patients, four had similar NAA/(Cr + Ch) ratios.
|
RESULTS |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODS RESULTS DISCUSSIONREFERENCES |
|---|
|
|
|
|
|
In two of three patients without hippocampal atrophy, the hippocampal NAA/(Cr + Ch) ratio clearly indicated a lower ipsilateral ratio than a contralateral ratio; in the third patient, no contralateral hippocampal value could be obtained. However, the hemispheric NAA/(Cr + Ch) ratios were similar to those in patients with hippocampal atrophy (Table 1).
Comparison of NAA/(Cr + Ch) Ratios in Individual Brain Regions
The reduced NAA/(Cr + Ch) ratios observed in the whole brain outside the hippocampus could be due to widespread or focal NAA concentration losses. To detect focal NAA concentration loss, we compared the ratios of individual voxel positions among patients (voxels obtained from the ipsilateral side and the contralateral side) and control subjects (Table 2). Mean ipsilateral NAA/(Cr + Ch) ratios in patients were lower in 10 of 11 different brain regions outside the hippocampus compared with the control mean for these regions. In five of 11 brain regions outside the hippocampus (ie, the anterior frontal, posterior frontal, posterior parietal, and lateral occipital lobes and the sensory cortex), these ipsilateral NAA/(Cr + Ch) ratios were reduced compared with those of control subjects, with a P value less than .05 (calculated with a one-tailed t test without correction for multiple comparisons). In addition, the NAA/(Cr + Ch) ratio in the midbrain was lower in patients than it was in control subjects. In 10 of 11 brain regions outside the hippocampus, the mean ipsilateral NAA/(Cr + Ch) ratio also was lower compared with that in the corresponding contralateral regions, and the difference between ipsilateral and contralateral ratios was less than .05 in the anterior frontal lobe (P < .03) and the sensory cortex (P < .03), with a one-tailed paired t test without correction for multiple comparisons (Table 2). The contralateral NAA/(Cr + Ch) ratio was lower than the ratio for control subjects in nine of 11 regions outside the hippocampus, with that in two of 11 regions (ie, the central occipital lobe and the parietal cortex) being reduced compared with that in controls, with a P value less than .05 (one-tailed paired t test without correction for multiple comparisons).
Because the characteristics of the NAA/(Cr + Ch) ratio reductions in patients with mesial TLE outside the medial temporal lobe were widespread and nonfocal and because the reductions tended to be bilateral rather than unilateral, the differences between ipsilateral and contralateral ratios of individual voxels were small. Therefore, although most individual brain regions outside the hippocampus demonstrated a slightly lower ipsilateral NAA/(Cr + Ch) ratio compared with the contralateral side, the results did not assist in seizure focus lateralization and did not improve the lateralization obtained with spectra from hippocampal regions alone.
Correlations between NAA/(Cr + Ch) Ratios in the Hippocampus with Those in Other Brain Regions
We correlated the NAA/(Cr + Ch) ratios in ipsilateral hippocampus with ipsilateral ratios in individual brain regions outside the medial temporal lobe, as well as with the whole-brain and hemispheric ratios. The NAA/(Cr + Ch) ratio in the central frontal lobe and the ipsilateral hippocampus correlated positively (r = 0.76, P < .05, without correction for multiple comparisons), whereas NAA/(Cr + Ch) ratios in other regions and in the whole brain and the ipsilateral hemisphere were not significantly correlated with the hippocampal values.
|
DISCUSSION |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODS RESULTS DISCUSSIONREFERENCES |
|---|
In patients, the ipsilateral hippocampal NAA/(Cr + Ch) ratio was reduced, compared with that in contralateral hippocampus and control subjects. When results in both hippocampi could be obtained, all patients were lateralized correctly on the basis of the lower ipsilateral NAA/(Cr + Ch) ratio, compared with the contralateral ratio. The reduced hippocampal NAA/(Cr + Ch) ratio could be due to neuron loss and gliosis or neuronal dysfunction. These findings obtained with multisection MR spectroscopic imaging confirm findings in previous studies (1–9), which solely measured hippocampal tissue and/or parts of the temporal lobe. In contrast to findings in previous studies (2,3,7,9,20,24), however, the contralateral hippocampal NAA/(Cr + Ch) ratio in the hippocampal body was not different from the ratio in control subjects, whereas in the anterior hippocampus, there was a trend of lower contralateral NAA/(Cr + Ch) ratios compared with the ratios in control subjects. This could be due to the relatively small sample size, or it may reflect the finding of anteroposterior hippocampal differences in the ratios between patients and control subjects (29).
In an MR spectroscopic imaging study in regard to patients with TLE before and after surgery, Cendes et al (21) determined that NAA/Cr ratio loss is widespread in the ipsilateral temporal lobe in patients with TLE and extends beyond the regions of abnormality, as usually observed by using conventional MR imaging. Similarly, Li et al (26) demonstrated reduced NAA/Cr ratio in frontocentroparietal regions in patients with TLE. The findings of this study imply that, in addition, the NAA concentration is reduced in many other areas outside the hippocampal region and the temporal lobe. The largest NAA/(Cr + Ch) ratio reductions were observed in the ipsilateral hemisphere in patients with mesial TLE, in addition to smaller NAA/(Cr + Ch) ratio losses in different regions of the contralateral hemisphere when the ratios were compared with those in control subjects. These results suggest a diffuse impairment of the brain in patients with mesial TLE.
These findings are consistent with those in a neuropathologic study (10) in patients with TLE. In that study, it was determined that although the greatest abnormalities were observed in the medial temporal area, these abnormalities were not confined to the temporal region. Instead, the brain was usually bilaterally affected in a number of different areas, including the cerebral cortex, the subcortex, and the cerebellum. Similarly, in an MR imaging study (11) in patients with TLE extrahippocampal volume abnormalities were observed bilaterally. Furthermore, in a 31P MR spectroscopic imaging study in patients with TLE, van der Grond et al (19) detected diffuse metabolite changes in the contralateral hemisphere besides ipsilateral abnormalities. Unexpectedly, the contralateral changes were in the opposite direction, compared with the ipsilateral changes. In addition, findings in PET studies (15–17) in patients with TLE revealed widespread impairment that was evidenced by hypometabolism of regions outside the temporal lobe.
The NAA/(Cr + Ch) ratios in most brain regions in the ipsilateral hemisphere were lower than the ratios in corresponding contralateral regions. However, these differences were statistically significant in only two brain regions outside the hippocampus, probably because the contralateral hemisphere was impaired also. Therefore, the current approach of selection of voxels from defined brain regions did not improve the seizure focus lateralization over use of only spectra from hippocampal regions or the temporal lobe. This result is consistent with an MR imaging finding that the pattern of extrahippocampal volume deficits did not lateralize the epileptogenic hemisphere because the deficits were present bilaterally (11). However, the trend of a reduced ipsilateral NAA/(Cr + Ch) ratio compared with a contralateral ratio in individual brain regions and a trend of reduced ipsilateral compared with contralateral hemispheric values indicates that the analysis of all acquired voxels improves the statistical power and may therefore assist in seizure focus lateralization.
The results clearly indicate that the multisection sequence is capable of detection of NAA/(Cr + Ch) ratio reductions in many brain areas. The ability to access most parts of the brain may help in seizure focus localization, especially in patients with neocortical epilepsy in whom the focus is outside the hippocampus.
Findings in previous studies from our laboratory (2,3) and in those of others (3,7,9) have suggested that the NAA concentration in the contralateral hippocampus is reduced in patients with mesial TLE. Furthermore, findings in studies (21,23) concerning patients with mesial TLE before and after surgery indicate that an initially reduced NAA concentration in the contralateral hippocampus recovers after surgery. A possible explanation for this recovery is that neuron impairment on the contralateral side is secondary to ipsilateral seizure activity rather than due to bilateral hippocampal sclerosis (23). The finding of a widespread reduced NAA/(Cr + Ch) ratio outside the hippocampus and the temporal lobe supports this hypothesis. The results suggest that besides the direct impairment of the ipsilateral hippocampus, ipsilateral epileptiform discharges lead to a diffuse impairment of the ipsilateral and, to a smaller extent, the contralateral hemisphere, and this impairment probably represents a neuronal dysfunction rather than a neuronal loss.
This concept is in contrast to findings in an earlier MR spectroscopic imaging study (3) of neocortical epilepsy with extratemporal seizure origin in which no reduction of hippocampal NAA concentration was established. Mechanistic differences in seizure propagation between patients with neocortical epilepsy and those with mesial TLE could be responsible for the different findings. Alternatively, the diffuse metabolic changes in patients with mesial TLE could be due to long-term effects of the drugs the patients were receiving. However, the previously obtained finding of an unchanged hippocampal NAA concentration occurred in patients with neocortical epilepsy who were receiving anticonvulsants in dosages similar to those that the patients with mesial TLE were receiving. This suggests that the treatment is not responsible for the detected changes. Further studies are underway to determine whether the preoperative extratemporal impairment improves after surgery.
Similar to findings in PET studies indicating that there was no correlation between hypometabolism in different regions of the brain and the severity of hippocampal neuron loss (16), our findings indicated that there were no significant correlations between the NAA/(Cr + Ch) ratio from regions outside the hippocampus and the values from the hippocampal regions.
The current study has several limitations: First, while the spectra of most measured regions in the brain yielded excellent spectral quality, there were several regions that consistently did not produce acceptable spectral quality, according to the applied threshold. These regions included the anterior lateral temporal lobe (which was not included in the analysis) and anterior hippocampus and central frontal lobe (with only 50% and 30% of the spectra, respectively, within the quality threshold). These regions are subject to susceptibility gradients from the air-to-tissue interface of the sinuses. A simultaneous measurement together with that of other brain regions yielding high spectral quality is therefore not possible without the use of special techniques (eg, acquisition of multiple spin echoes).
Second, we did not determine absolute metabolite concentrations. Therefore, the detected changes in the NAA/(Cr + Ch) ratio could be due to NAA concentration reductions or to Cr and/or Ch increases. However, because in studies about epilepsy in which absolute values were determined, the findings generally indicated NAA concentration reductions rather than Cr and/or Ch increases, it is likely that the demonstrated changes in the NAA/(Cr + Ch) ratio are primarily due to a reduced NAA concentration. In either case, the finding of a reduced NAA/(Cr + Ch) ratio reflects some impairment of the brain.
Third, the changes could be the result of changes in relaxation times. Although we cannot exclude this possibility, it is unlikely, because very different relaxation time changes would be necessary for NAA and for Cr or Ch to account for changes in the NAA/(Cr + Ch) ratio.
Fourth, we did not perform segmentation of the MR imaging data, which would provide information on the amount of gray and white matter contributions to each voxel. Since metabolite concentrations are different in gray and white matter, the detected changes could therefore also be owing to differences in gray and white matter voxel composition between patients and control subjects. However, the visual selection of voxels in predominantly gray matter was performed with blinding to identification of patients and control subjects and to knowledge of the side of the seizure focus. It is therefore very unlikely that differences in gray and white matter voxel contributions are responsible for the detected changes. Nevertheless, correction for proportions of gray and white matter in voxels would probably enhance the results further.
In conclusion, findings at multisection MR spectroscopic imaging in patients with mesial TLE confirmed previous findings of reduced NAA concentration in ipsilateral hippocampal structures and simultaneously revealed reductions in the NAA/(Cr + Ch) ratio in widespread areas of the brain, and these findings suggested that repeated seizures that originate from the hippocampus lead to impairment of other brain regions. Contralateral structures were also affected. Therefore, the lateralization with spectral information from hippocampal areas alone could not be improved with the additional information. Further studies are underway to determine whether the analysis of all acquired voxels, because of greater statistical power, will provide an improved separation between the ipsilateral and contralateral sides in patients with mesial TLE.
|
FOOTNOTES |
|---|
Author contributions: Guarantors of integrity of entire study, K.D.L., M.W.W.; study concepts and design, all authors; literature research, P.V., K.D.L.; clinical studies, P.V., G.B.M., K.D.L.; data acquisition, P.V., G.B.M.; data analysis/interpretation, all authors; statistical analysis, P.V.; manuscript preparation, P.V.; manuscript definition of intellectual content, all authors; manuscript editing, P.V.; manuscript revision/review, K.D.L., N.S., G.B.M., M.W.W.; manuscript final version approval, all authors.
|
REFERENCES |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODS RESULTS DISCUSSIONREFERENCES |
|---|
This article has been cited by other articles:
|
|
 |
|
  K.K. Oguz, A. Kurne, A.O. Aksu, E. Karabulut, A. Serdaroglu, S. Teber, S. Haspolat, N. Senbil, S. Kurul, and B. Anlar Assessment of Citrullinated Myelin by 1H-MR Spectroscopy in Early-Onset Multiple Sclerosis AJNR Am. J. Neuroradiol., April 1, 2009; 30(4): 716 - 721. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 F. Riederer, R. Lanzenberger, M. Kaya, D. Prayer, W. Serles, and C. Baumgartner Network atrophy in temporal lobe epilepsy: A voxel-based morphometry study Neurology, August 5, 2008; 71(6): 419 - 425. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
H. P. Hetherington, R. I. Kuzniecky, K. Vives, O. Devinsky, S. Pacia, D. Luciano, B. Vasquez, S. Haut, D. D. Spencer, and J. W. Pan A subcortical network of dysfunction in TLE measured by magnetic resonance spectroscopy Neurology, December 11, 2007; 69(24): 2256 - 2265. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 J. F. A. Jansen, W. H. Backes, K. Nicolay, and M. E. Kooi 1H MR Spectroscopy of the Brain: Absolute Quantification of Metabolites. Radiology, August 1, 2006; 240(2): 318 - 332. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
R. S. Briellmann, G. D. Jackson, G. S. Pell, L. A. Mitchell, and D. F. Abbott Structural abnormalities remote from the seizure focus: A study using T2 relaxometry at 3 T Neurology, December 28, 2004; 63(12): 2303 - 2308. [Abstract] [Full Text] [PDF]
|
|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
| RADIOLOGY | RADIOGRAPHICS | RSNA JOURNALS ONLINE |
