3-Tesla MRI and Temporal Lobe Epilepsy

3-Tesla MRI and Temporal Lobe Epilepsy

3-Tesla MRI and Temporal Lobe Epilepsy Juan Alvarez-Linera Prado, MD Magnetic resonance imagine (MRI) is the essential tool for the diagnosis and eval...

9MB Sizes 2 Downloads 48 Views

3-Tesla MRI and Temporal Lobe Epilepsy Juan Alvarez-Linera Prado, MD Magnetic resonance imagine (MRI) is the essential tool for the diagnosis and evaluation of epileptic patients due to the possibility of it identifying the epileptogenic substrate and the decisive value for the selection of surgical candidates in cases of refractory epilepsy. Because the main role of MRI in epilepsy relates to the detection of a structural injury, the quality of the image clearly determines the results. In structural studies, it can contribute to detecting small hippocampal alterations that may be missed with conventional techniques, and it is a more reliable tool for detecting small focal cortical dysplasias (FCD), which constitute the most frequent cause of dual injury. 3-Tesla MRI also provides relevant information in functional studies. Its advantages rely on the rise in signal-to-noise ratio and its higher T2 contrast, and also, in the case of MR spectroscopy, on its higher chemical shift, which in general increases both the quality and the rapidity of acquisition. Finally, epilepsy is one of the main indications for obtaining cerebral activation maps, usually using the blood oxygen level dependent (BOLD) technique, which is one of the techniques that most benefits from the magnetic field increase. Semin Ultrasound CT MRI 28:451-461 © 2007 Elsevier Inc. All rights reserved.

M

agnetic resonance imaging (MRI) has entailed one of the most significant advances in the management of epileptic patients due to the fact that it has made it possible to identify the epileptogenic substrate with increased effectiveness. Today, MRI is an essential part of the evaluation of epilepsy and its use is recommended for all patients, with the sole exception of those with idiopathic generalized epilepsy and benign childhood epilepsy. MRI provides relevant information for all stages of managing these patients. In the initial evaluation, MRI is one of the criteria used in syndromic classification. In patients with refractory epilepsy, MRI is decisive for the selection of surgical candidates and also contributes to treatment planning. Because the main role of MRI in epilepsy relates to the detection of a structural injury, the quality of the image clearly determines the result. This has been set out in the literature as technical advances in terms of both hardware and software have been implemented. For example, the sensitivity of MRI in patients with mesial temporal sclerosis (MTS) in the 1980s did not exceed 50%; the application of high resolution sequences with fast techniques (T2-fast spin echo [FSE], 3D-T1) and later the application of FLAIR sequences entailed a very significant increase in sensitivity,1 which in the 1990s surpassed 90%. With regard to the magnetic field, the difference in the result between 0.5-T and

Chief of Division of Neuroradiology, Diagnostic Imaging Service, Ruber International Hospital, Madrid, Spain. Address reprint requests to: Juan Alvarez-Linera Prado, La Masó 38, Madrid 28034, Spain. E-mail: [email protected]

0887-2171/07/$-see front matter © 2007 Elsevier Inc. All rights reserved. doi:10.1053/j.sult.2007.09.007

1.5-T magnets is also significant, and today an epilepsy study in fields lower than 1.5 T would not be conceived. MRI has a high value prognosis, because on the one hand the detection of a structural injury is related to a significant degree to a poorer medical control of the seizures, and on the other hand when MRI shows a structural injury the surgical result is clearly better. Even in cases of seizures that are initially labeled as generalized, the detection of an injury in MRI leads to undertaking a reevaluation of the seizures, which in some cases turn out to be partial seizures with fast secondary generalization. This fact obviously alters the way the patient is managed, which underscores the high value of injury detection in the evaluation of epileptic patients and emphasizes the importance of obtaining images of the utmost quality.2 In addition to advances in software, the application of more efficient coils, such as multiple channels coils, when the signal-to-noise ratio (SNR) is raised increases the capacity to detect small injuries.3 The standard quality in MRI has usually been obtained with 1.5-T magnets, and higher field magnets (3 T or above) have been restricted to use in research and especially for working with functional techniques or magnetic resonance spectroscopy (MRS). Nevertheless, 3-T magnets are increasingly accessible and in the present day they are beginning to become available in most reference hospitals. It has been shown that not only do they provide relevant information in functional studies, but they also allow higher quality structural images to be obtained, which in some cases may change the way the patient is managed.4 In addition to the structural study, diffusion, perfusion and spectroscopy 451

452 techniques are increasingly used as additional information to the structural images, although their routine use has yet to be clearly defined. In this regard, the use of 3 T provides evident advantages due to the rise in SNR and its higher T2 contrast, and also, in the case of MRS, due to its higher chemical shift, which in general increases both the quality and the rapidity of acquisition, which may contribute to it being used to a greater extent in the clinical environment. Finally, epilepsy is one of the main indications for obtaining cerebral activation maps, usually using the BOLD technique, which is one of the techniques that most benefits from the magnetic field increase. Of all the disadvantages of 3-T magnets, perhaps the one that today represents the greatest drawback to its application in clinical routine is the increased cost, since epilepsy is a major disease and therefore the indication of a large number of 3-T MRI studies could have a significant impact on health care expenditure.5 When MRI was first being developed, the cost of the diagnostic process using MRI as opposed to computed tomography (CT) was studied. The results indicated that despite the greater cost of MRI, the final cost using MRI was lower than with CT,6 because other more invasive tests required after a negative CT are avoided. The situation is similar when one considers 0.5-T MRI compared with 1.5-T MRI, which although less expensive is not recommended due to its poor yield; even 1.5-T MRI with routine protocol is not considered suitable for examining epileptic patients, especially in cases of refractory epilepsy.7 Therefore, it is probable that as 3-T MRI becomes more affordable and accessible, 3-T instruments are more likely to be used for examining epileptic patients.8,9 However, more experience is needed to determine the role of 3-T MRI in epilepsy with regard to both structural and functional studies.

3-T MRI: Advantages and Disadvantages The main result of the increase in the magnetic field is a rise in the SNR. Theoretically, the SNR has a linear relation with the field, which is why, in principle, the SNR at 3 T would have to be double the 1.5-T level.10 Nevertheless, the relationship between the SNR in clinical imaging and the field is very complex and depends on many factors that include sequence type used, the efficiency of the coil and the radiofrequency (RF) system, as well as the noise derived from the field increase, its homogeneity, and the dielectric effect.11,12 Comparative studies show variable results, from a 25% increase in hydrogen spectroscopy sequences13 to 100% in BOLD technique cases.14 However, with the technology currently available, the rise in SNR is considered to be approximately 70%.15 Taking into account that the acquisition time is proportional to the square root of the SNR, 3 times less time is required to obtain an image with SNR similar to 1.5 T in a 3-T magnet. Increased rapidity may be used to obtain the same image in less time, but it mainly allows higher resolution images to be obtained using the same acquisition time. The increase of the magnetic field causes an increase of T1 and a (discrete) decrease of T2 and above all T2*, and this

J.A.-L. Prado affects both the SNR and the contrast. The spin-echo technique (SE) at 3 T obtains less T1 contrast, but this type of sequence is not in fact recommended in epilepsy protocols due to the low contrast between gray matter (GM) and white matter (WM). The T1 images used in an epilepsy protocol are obtained using gradient-echo techniques (GE) in 3D acquisition, usually with an inversion recovery (IR) preparation pulse to increase the GM/WM contrast (3D-IRprep-GE). Obtaining an isotropic 3D-IRprep-GE block enables reconstruction in any plane and offers the possibility of carrying out volumetry. The GM/WM contrast with 3D-IRprep-GE technique is similar at 1.5 T and 3 T, although at 3 T the SNR is 50% to 70% higher. The result is that the quality of the image in T1 with this technique is clearly superior at 3 T, which facilitates the detection of small cortical injuries and contributes to improving the results of the volumetry.16 The field increase produces some T2 shortening, which tends to increase the contrast. The rise in SNR produced by the field increase fully exceeds the discrete signal loss that occurs due to a shorter T2 (Fig. 1). In this way, it is possible to detect injuries at 3 T that go unnoticed at 1.5 T,9,17 and this could be improved still more through the use of new double-inversion FLAIR sequences.18 Improved injury detection may change patient management,4,19 which should be valued in terms of cost/benefit not only due to the fact that more expensive tests are avoided but because the risk these entail is also avoided. The influence of the magnetic field on the spectroscopy is determined by the rise in SNR and the increase in chemical shift, which maintains a linear proportion with the field. The increase in chemical shift has two main results. First, it allows more efficient water suppression, which gives rise to a flatter basal line and therefore to a more robust spectra with lower variability (Fig. 2). Second, increasing the separation between the peaks of the different metabolites makes it easier to quantify and identify metabolites that cannot be solved with 1.5-T magnets such as the “Glx” complex, which contains glutamate and glutamine.20 Nevertheless, it should be taken into account that T2 shortening also has two important effects. It produces higher signal loss in the long echo time (TE) sequences. This means that if sequences with very long TE are used, such as 288 msec, which has traditionally been used at 1.5 T, the SNR gained from the field increase is lost for the most part as a result of such long TE, which is why the final result is not superior at 3 T. Nevertheless, the main reason for the use of long TE has been to obtain a flatter basal line due to the low suppression of the water peak that takes place at 1.5 T and the increased difficulty of obtaining a quality spectrum with short TE PRESS sequences, due to the eddy currents and shimming problems. Current technical advances make it possible to work with short TE without too many problems. Bearing in mind that the suppression of the water is superior at 3 T, it is not useful to use sequences with TE higher than 136 msec. Although there is a scarcity of clinical experience, in general at 3 T short TE sequences tend to be used, because they maintain a quality spectrum and improved SNR and provide information on more metabolites. When only Nacetyl-aspartate (NAA), creatine (Cr), and choline (Cho) are valued, it is possible to use sequences with intermediate TE of

3-Tesla MRI and temporal lobe epilepsy

453

Figure 1 Normal hippocampi on 1.5 T (A) and 3 T (B). 3-T images have more SNR and more contrast between gray and white matter showing the internal hippocampal structure more clearly.

between 60 and 100 msec. T2 shortening, even with short TE sequences, causes changes in the values of some metabolites that have a shorter T2, such as myoinositol, which also shows a double peak at 3.6 ppm with TE of 35 msec, making it more difficult to quantify and raising the possibility of using sequences with shorter TE than usual (25-30 msec). All this makes it necessary to define the normal values at 3 T. Despite these disadvantages, it should be taken into account that in epileptic patients, changes in the metabolites are usually small (around 15% in relation to the normal spectra), which is why the main objective is to obtain robust spectra with little variability that allow the normal and pathological spectra to be distinguished in a reliable fashion. This is better obtained using 3-T magnets. The signal increase along with the increased T2 effect obtained with 3 T is ideal for improving the quality of the diffusion and perfusion sequences, although it is less used in epilepsy. Perfusion studies have demonstrated that they can detect areas of decreased regional blood volume (RBV) related to the epileptogenic focus.21 However, studies of interictal perfusion with single-photon emission computed tomography (SPECT) have not demonstrated that they contribute more information than the structural MRI, and it is not practical to conduct ictal studies in perfusion MRI except in special cases. 3-Tesla magnets enable arterial spin labeling (ASL) sequences to develop better,

and although the outlook is promising (because it is not necessary to use IV contrast, and it can therefore be repeated without problems and in addition the flow can be quantified),22 more experience is required for it to be applied on a routine basis. With regard to the diffusion studies, the field increase allows sequences with higher b value to be carried out, which can increase sensitivity to detect subtle alterations because the T2 contribution decreases, and also makes it possible to study changes in compartments with slower molecular movement, due to the fact that the apparent diffusion coefficient (ADC) has at least a biexponential relation with the b value.23 By means of diffusion tensor imaging (DTI) it is possible to obtain fractional anisotropy (FA) maps that in many cases detect alterations in the WM that are not visible with structural techniques. Some studies have indicated the presence of alterations in mean diffusivity and in the FA in the WM of patients with focal epilepsy and normal MRI, which is why it is possible that the use of hi-res DTI sequences, both with regard to the voxel size and the number of directions, which is possible with higher field magnets, may allow the improved detection of injuries that are not visible in conventional MRI, especially certain focal cortical dysplasias (FCD) (Fig. 3). BOLD sequences probably obtain the most advantage from the field increase, since they relate a linear increase

J.A.-L. Prado

454

of magnetic susceptibility to the signal increase, which is the basis of the contrast in this technique. Higher SNR makes it possible to work with a smaller sized voxel, which is why there is less partial volume with “inactive” parenchyma, and in addition there is less contribution from larger vessels such as the veins, with changes remaining more restricted at the capillary level (Fig. 4). In this way, the signal difference between the phases of activity and rest are up to 110% higher with 3-T images,14 which increases sensitivity to detect small areas of activation or areas that show weaker changes not visible with 1.5-T magnets. One of the disadvantages of the field increase is that it is more difficult to examine areas near the bone or the air of the paranasal sinuses, such as the mesial temporal region. However, the use of parallel imaging (PI) sequences10 (ASSET, SENSE) significantly reduces image distortion. The use of PI, along with an adjustment of TE and an increase of the spatial resolution, makes it possible to properly examine these areas.

3-T MRI and Temporal Lobe Epilepsy

Figure 2 Small focal lesion in the lentiform nucleus (A). 1.5-T multivoxel (1 cc) spectroscopy (A) showing noisy basal line and poor resolution of metabolites. On 3-T multivoxel with same parameters (B) the basal line is less noisy and metabolites are clearly depicted, showing a small decrease in NAA/Cr on the lesion. (Color version of figure is available online.)

Temporal lobe epilepsy (TLE) is the most frequent form of epilepsy in adults, and around 70% are part of the medial temporary sclerosis syndrome (MTS). MRI is a very sensitive technique for detecting MTS, and although it is difficult to estimate sensitivity and specificity, since the data almost always come from surgical series, where intervention is very infrequent on the hippocampus with normal MRI, most of the current series show a higher than 90% sensitivity and specificity for 1.5-T MRI. It therefore seems difficult for 3-T MRI to be able to provide further relevant information. Cases of TLE with negative MRI usually show poorer surgical results, and frequently the pathological anatomy findings are far from specific. Nowadays there is a trend toward considering TLE with negative MRI cases as being different from the MTS syndrome,24 not just because of the different prognosis but because they show neocortical alterations in both functional studies and hydrogen spectroscopy.25 It would therefore be worthwhile to ensure in the most reliable manner that the MRI does not show any signs of hippocampal pathology, which is why it may be useful to attempt to increase the sensitivity of the MRI. There are no published series that compare 1.5-T MRI and 3-T MRI, although isolated cases of negative 1.5-T MRI with positive 3-T MRI have been published.26 Many of the patients that attend reference centers provide studies of 1.5-T with negative reports in which the image quality is unsuitable27 or which show a few inconspicuous alterations that had gone unnoticed.7 Frequently, for FCD cases and certain MTE cases, once the study at 3 T has been conducted and an alteration is observed, a review of the 1.5-T MRI reveals that an alteration already existed but was so subtle that it prevented a reliable diagnosis. In fact, studies that have compared the routine diagnosis conducted on a study of 1.5 T with the reading of the same study by an expert have demonstrated that the second reading increased the diagnosis by a significant degree, even in the case of MTE.7

3-Tesla MRI and temporal lobe epilepsy

455

Figure 3 Right mesial frontal focal cortical dysplasia. There is a subtle cortical thickening and loss of cortical differentiation (A and B) (not visible on 1.5 T). DTI FA map shows decrease of FA in subcortical white matter (C). On fluoro-2-deoxy-D-glucose positron emission tomography (D) there is a decrease of glucose metabolism in the same place. (Color version of figure is available online.)

The 3-T MRI may be useful when doubts exist concerning the possibility of an injury, but if the 1.5-T MRI protocol is correct, it probably provides less information in the case of patients with TLE and normal MRI than in patients with FCD, although more experience is required to determine the impact of the 3-T MRI on the MTE. Furthermore, it is important

to bear in mind the possibility of a dual pathology that occurs in up to 15% of adult cases and in up to 67% of pediatric series.28 The pathology related to MTS is almost always FCD, generally type I, which makes image quality very important, because the way patients with dual pathology are managed may differ from MTE patients. Whereas in the case of MTE

J.A.-L. Prado

456

Figure 4 Brain activation with BOLD technique using low resolution (3 ⫻ 3 ⫻ 5 cm; A and B) and high resolution (1 ⫻ 1 ⫻ 3 cm; C and D). Signal changes are more pronounced on high resolution and more coincident with cortical locations. Moreover, there are more activated pixels with the high resolution technique. (Color version of figure is available online.)

the sensitivity of the 1.5-T MRI is very high, this does not apply for malformations of cortical development (MCD), which is why it may be useful to conduct an 3-T MRI, especially in children, if the clinical examination is not typical or

when the 1.5-T MRI raises doubts as to the possibility of dual pathology. Multiple studies have demonstrated hydrogen spectroscopy’s ability to lateralize the focus in patients with temporary epilepsy, but they are rarely patients with patho-

3-Tesla MRI and temporal lobe epilepsy logical MRI. If the MRI displays signs of MTE and the clinic and the EEG are congruent, then other tests are not generally necessary, which is why spectroscopy in the MTE is more focused on cases that show doubtful or negative structural MRI. There are few studies concerning the lateralization capacity of the spectroscopy focus when the MRI is negative.29 Some studies have shown much less frequent differences when the MRI is negative (27% compared with 80%),30 and more recent studies do not show significant differences in the hippocampus of patients with negative echo train length (ETL) and MRI.31 It is noteworthy that other studies conducted using the multivoxel technique demonstrate extrahippocampal and even extratemporal alterations,32 and a recent study using simple voxel technique on 14 subjects with negative MRI showed a decrease in the NAA/Cho coefficient in the lateral temporary region matching alterations in the magneto-encephalography, with no significant changes being observed in the hippocampus.33 These findings are in agreement with studies conducted with PET and DTI, which indicate the existence of neocortical alterations in patients with normal MRI and ETL. All this suggests that many of the patients with negative MRI and TLE probably constitute a different group from the ETM, with a different postsurgical prognosis34 and for which more extensive surgical treatment than the amygdalo-hippocampectomy may have to be envisaged. All this further emphasizes the importance of ensuring a structural MRI of the utmost quality (Fig. 5), where the 3-T MRI may increase in sensitivity to detect subtle alterations in the hippocampus that are hardly visible with 1.5-T MRI (Fig. 6). On the other hand, the most recent studies conducted with single-voxel technique use small volumes to only measure the hippocampus, and some authors point out the advantage that the use of 3-T MRI to obtain a small volume with a higher quality spectrum with improved peak definition can entail (Fig. 7). When the results of the single-voxel spectroscopy limited to the hippocampus (approximately 3 cc) are compared with the multivoxel spectroscopy with a resolution of 2 mL and 1 mL in patients with typical MRI of MTE, one observes that as the spatial resolution increases, the differences between the hippocampus with MRI alterations and the contralateral also increase, which is why 3-T MRI will probably constitute a very useful tool in the study of ETM. Diffusion studies have also been used to identify changes in the hippocampus of patients with TLE, since the loss of the inner structure of the hippocampus produces a discrete increase of the ADC.35-37 Nevertheless, in patients with normal MRI, other authors indicate that the changes are not significant.38 It is probable that the alteration in the diffusion is lower in these patients and it is therefore not possible to detect subtle changes. Comparing the differences in the ADC of patients with typical MRI findings of MTE using b values of 1000 and 3000 confirms that the differences are clearly greater when a higher b value is used (Fig. 8). The main difficulty with using high b values is the severe decrease of the SNR, which is why 3-T MRI can be very useful. Other studies have detected alterations in the FA of the white matter of the temporal lobe of patients with ETM and normal MRI, using

457

Figure 5 High resolution T2 sagittal slices of normal hippocampus (A) and MTS (B). Hippocampal head of MTS shows volume decrease (more pronounced on C-4) and hyperintensity.

DTI.39 Again, the advantages of using 3-T MRI are clear with regard to DTI, which is why although as yet no experience in MTE and normal MRI cases has been gained the use of higher

Figure 6 Left mesial temporal sclerosis (MTS), barely visible on 1.5 T (A) and clearly visible on 3 T.

J.A.-L. Prado

458

Figure 7 Left MTS. Multivoxel spectroscopy showing no significant asymmetry on 1.5 T (A and B) an clearly asymmetry on 3 T (B and C). (Color version of figure is available online.)

resolution and higher b values can probably provide more information. There are few studies with perfusion techniques in ETM, and although a decrease of the CBV in the temporal lobe of patients with ETM has been detected,21 the difference is small, in a similar way to interictal SPECT findings. In addition, there are no studies of patients with normal MRI, which explains why the usefulness of the perfusion studies seems limited. Nevertheless, 3-T magnets make it possible to conduct more reliable studies by means of techniques that magnetically mark the blood flow (arterial spin labeling, or ASL), and in addition to not using IV contrast they do not show artifacts related to major vessels and allow the flow to be quantified. It is possible that hi-res ASL techniques can provide additional information.22 Functional magnetic resonance imaging (fMRI) using BOLD techniques is increasingly used in patients with TLE, mainly in presurgical evaluation, to determine the hemispheric dominance for the language. Numerous studies have demonstrated a very good correlation with the Wada test,40-42

the advantage being that it is a risk-free technique. The main limitations of the fMRI are the presence of artifacts, due mainly to movement and the need for the patient to cooperate well in order to perform the tasks. In language studies, the BOLD effect is much lower than in tasks in which primary areas, such as the motor area, are involved. Therefore, any type of artifact interferes with the reliability of the result to a much greater degree. When 3-T magnets are used, the difference between the activation areas and the surrounding parenchyma is above 100% in comparison with 1.5 T, and, in addition, the possibility of increasing the spatial resolution taking advantage of the increased SNR decreases the partial volume between voxels that are activated and those that are not, as well as decreasing the influence of the veins. It therefore seems clear that the use of 3-T magnets will be of benefit in studies related to hemispheric dominance for the language, although most of the studies have been conducted with 1.5 T. In addition to language studies, functional studies to evaluate the contribution of the hippocampus in memory

3-Tesla MRI and temporal lobe epilepsy

Figure 8 Right MTS with some hippocampal atrophy and T2 hyperintensity (A). Isotropic diffusion weighted imaging (B) and ADC map (C) with b value of 1000 and 3000 (D and E). There is a subtle hypointensity on right hippocampus (B), which is more evident on (D). There is a small ADC increase on both ADC maps, but it is more pronounced on (E).

459

J.A.-L. Prado

460 tasks, especially verbal memory, are of great interest, as this is another one of the tasks carried out in the Wada test, aimed at determining the postsurgical prognosis with regard to verbal memory loss. Recent works show a good correlation with both the Wada test and the postsurgical functional alterations,43,44 but all have been conducted by means of a comparison of groups and using region-of-interest (ROI) analysis, due to the fact that few differences have been observed. This is why they currently remain within the domain of research. Improvements in the technique are needed to attempt to find differences that make an individual analysis possible with a view to possible clinical application, and the use of higher resolution BOLD sequences with 3-T magnets will without doubt contribute to this.

Conclusions 3-Tesla MRI is being implemented very quickly as a clinical tool in neuroradiology. Its main current limitation is increased cost, and it is foreseeable that in the near future it will be available in most reference hospitals and therefore epilepsy units will have easy access to it. In structural studies, it can contribute to detecting small hippocampal alterations that may be missed with conventional techniques, and it is a more reliable tool for detecting small FCD, which constitute the most frequent cause of dual injury. Spectroscopy, diffusion, and ASL perfusion may provide additional information in patients with TLE and normal MRI, and the increase of the magnetic field is a determining factor for raising the yield of these techniques. Cerebral activation studies are an alternative to the Wada test, although memory studies are still in the research phase. However, there is every indication that their development will be given a new boost with the incorporation of 3-T magnets.

9.

10.

11.

12.

13.

14.

15. 16.

17.

18.

19.

20.

References 1. Taillibert S, Oppenheim C, Baulac M, Dormont D, Marsault C, Cabanis EA, et al: Yield of fluid-attenuated inversion recovery in drug-resistant focal epilepsy with noninformative conventional magnetic resonance imaging. Eur Neurol 41(2):64-72, 1999 2. Duncan JS: Neuroimaging for epilepsy: quality and not just quantity is important. J Neurol Neurosurg Psychiatry 73(6):612-613, 2002 3. Grant PE, Barkovich AJ, Wald LL, Dillon WP, Laxer KD, Vigneron DB: High-resolution surface-coil MR of cortical lesions in medically refractory epilepsy: a prospective study. AJNR Am J Neuroradiol 18(2):291301, 1997 4. Knake S, Triantafyllou C, Wald LL, Wiggins G, Kirk GP, Larsson PG, et al: 3T phased array MRI improves the presurgical evaluation in focal epilepsies: a prospective study. Neurology 65(7):1026-1031, 2005 5. Griffiths PD, Coley SC, Connolly DJ, Hodgson T, Romanowski CA, Widjaja E, et al: MR imaging of patients with localisation-related seizures: initial experience at 3.0T and relevance to the NICE guidelines. Clin Radiol 60(10):1090-1099, 2005 6. Bronen RA, Fulbright RK, Spencer SS, Spencer DD, Kim JH, Lange RC: Economic impact of replacing CT with MR imaging for refractory epilepsy. Magn Reson Imaging 15(7):857-862, 1997 7. Von Oertzen J, Urbach H, Jungbluth S, Kurthen M, Reuber M, Fernandez G, et al: Standard magnetic resonance imaging is inadequate for patients with refractory focal epilepsy. J Neurol Neurosurg Psychiatry 73(6):643-647, 2002 8. Briellmann RS, Pell GS, Wellard RM, Mitchell LA, Abbott DF, Jackson

21.

22.

23.

24.

25.

26.

27.

28.

GD: MR imaging of epilepsy: state of the art at 1.5 T and potential of 3 T. Epileptic Disord 5(1):3-20, 2003 Regula S, Briellmann GSP, Wellard RM, Mitchell LA, Abbott DF, Jackson GD: MR imaging of epilepsy: state of the art at 1.5 T and potential of 3 T. Epileptic Disord 5:(1):3-20, 2003 Frayne R, Goodyear BG, Dickhoff P, Lauzon ML, Sevick RJ: Magnetic resonance imaging at 3.0 Tesla: challenges and advantages in clinical neurological imaging. Invest Radiol 38(7):385-402, 2003 Roschmann P: Radiofrequency penetration and absorption in the human body: limitations to high-field whole-body nuclear magnetic resonance imaging. Med Phys 14(6):\922-931, 1987 Ibrahim TS, Lee R, Baertlein BA, Robitaille PM: B1 field homogeneity and SAR calculations for the birdcage coil. Phys Med Biol 46(2): 609-619, 2001 Inglese M, Spindler M, Babb JS, Sunenshine P, Law M, Gonen O: Field, coil, and echo-time influence on sensitivity and reproducibility of brain proton MR spectroscopy. AJNR Am J Neuroradiol 27(3):684-688, 2006 Okada T, Yamada H, Ito H, Yonekura Y, Sadato N: Magnetic field strength increase yields significantly greater contrast-to-noise ratio increase: measured using BOLD contrast in the primary visual area. Acad Radiol 12(2):142-147, 2005 Vlaardingerbroek MT DBJ: Magnetic Resonance Imaging: Theory and Practice. New York, NY, Springer-Verlag, 1999 Wieshmann UC, Free SL, Stevens JM, Shorvon SD: Image contrast and hippocampal volumetric measurements. Magn Reson Imaging 16(1): 13-17, 1998 Wattjes MP, Lutterbey GG, Harzheim M, Gieseke J, Traber F, Klotz L, et al: Higher sensitivity in the detection of inflammatory brain lesions in patients with clinically isolated syndromes suggestive of multiple sclerosis using high field MRI: an intraindividual comparison of 1.5 T with 3.0 T. Eur Radiol 16(9):2067-2073, 2006 Wattjes MP, Lutterbey GG, Gieseke J, Traber F, Klotz L, Schmidt S, et al: Double inversion recovery brain imaging at 3T: diagnostic value in the detection of multiple sclerosis lesions. AJNR Am J Neuroradiol 28(1):54-59, 2007 Wattjes MP, Harzheim M, Kuhl CK, Gieseke J, Schmidt S, Klotz L, et al: Does high-field MR imaging have an influence on the classification of patients with clinically isolated syndromes according to current diagnostic MR imaging criteria for multiple sclerosis? AJNR Am J Neuroradiol 27(8):1794-1798, 2006 Di Costanzo A, Trojsi F, Tosetti M, Schirmer T, Lechner SM, Popolizio T, et al: Proton MR spectroscopy of the brain at 3 T: an update. Eur Radiol 2007 Wu RH, Bruening R, Noachtar S, Arnold S, Berchtenbreiter C, Bartenstein P, et al: MR measurement of regional relative cerebral blood volume in epilepsy. J Magn Reson Imaging 9(3):435-440, 1999 Liu HL, Kochunov P, Hou J, Pu Y, Mahankali S, Feng CM, et al: Perfusion-weighted imaging of interictal hypoperfusion in temporal lobe epilepsy using FAIR-HASTE: comparison with H(2)(15)O PET measurements. Magn Reson Med 45(3):431-435, 2001 Cendes F: October 11 Highlights and Commentary: Higher resolution MRI and image modeling for predicting surgical outcome in partial epilepsy. Neurology 65(7):975, 2005 Mueller SG, Laxer KD, Schuff N, Weiner MW: Voxel-based T2 relaxation rate measurements in temporal lobe epilepsy (TLE) with and without mesial temporal sclerosis. Epilepsia 48(2):220-228, 2007 Hammen T, Stefan H, Pauli E, Schafer I, Huk W, Tomandl B: 1H-MR spectroscopy: a promising method in distinguishing subgroups in temporal lobe epilepsy? J Neurol Sci 215(1-2):21-25, 2003 Sawaishi Y, Sasaki M, Yano T, Hirayama A, Akabane J, Takada G: A hippocampal lesion detected by high-field 3 tesla magnetic resonance imaging in a patient with temporal lobe epilepsy. Tohoku J Exp Med 205(3):287-291, 2005 McBride MC, Bronstein KS, Bennett B, Erba G, Pilcher W, Berg MJ: Failure of standard magnetic resonance imaging in patients with refractory temporal lobe epilepsy. Arch Neurol 55(3):346-348, 1998 Bocti C, Robitaille Y, Diadori P, Lortie A, Mercier C, Bouthillier A, et al:

3-Tesla MRI and temporal lobe epilepsy

29.

30.

31.

32.

33.

34.

35.

The pathological basis of temporal lobe epilepsy in childhood. Neurology 60(2):191-195, 2003 Suhy J, Laxer KD, Capizzano AA, Vermathen P, Matson GB, Barbaro NM, et al: 1H MRSI predicts surgical outcome in MRI-negative temporal lobe epilepsy. Neurology 58(5):821-823, 2002 Woermann FG, McLean MA, Bartlett PA, Parker GJ, Barker GJ, Duncan JS: Short echo time single-voxel 1H magnetic resonance spectroscopy in magnetic resonance imaging-negative temporal lobe epilepsy: different biochemical profile compared with hippocampal sclerosis. Ann Neurol 45(3):369-376, 1999 Mueller SG, Laxer KD, Suhy J, Lopez RC, Flenniken DL, Weiner MW: Spectroscopic metabolic abnormalities in mTLE with and without MRI evidence for mesial temporal sclerosis using hippocampal short-TE MRSI. Epilepsia 44(7):977-980, 2003 Vermathen P, Laxer KD, Schuff N, Matson GB, Weiner MW: Evidence of neuronal injury outside the medial temporal lobe in temporal lobe epilepsy: N-Acetylaspartate concentration reductions detected with multisection proton MR spectroscopic imaging—initial experience. Radiology 226(1):195-202, 2003 Shih JJ, Weisend MP, Lewine J, Sanders J, Dermon J, Lee R. Areas of interictal spiking are associated with metabolic dysfunction in MRInegative temporal lobe epilepsy. Epilepsia 45(3):223-229, 2004 Cohen-Gadol AA, Wilhelmi BG, Collignon F, White JB, Britton JW, Cambier DM, et al: Long-term outcome of epilepsy surgery among 399 patients with nonlesional seizure foci including mesial temporal lobe sclerosis. J Neurosurg 104(4):513-524, 2006 Duzel E, Kaufmann J, Guderian S, Szentkuti A, Schott B, Bodammer N, et al: Measures of hippocampal volumes, diffusion and 1H MRS metabolic abnormalities in temporal lobe epilepsy provide partially complementary information. Eur J Neurol 11(3):195-205, 2004

461 36. Goncalves Pereira PM, Oliveira E, Rosado P: Apparent diffusion coefficient mapping of the hippocampus and the amygdala in pharmacoresistant temporal lobe epilepsy. AJNR Am J Neuroradiol 27(3):671683, 2006 37. Londono A, Castillo M, Lee YZ, Smith JK: Apparent diffusion coefficient measurements in the hippocampi in patients with temporal lobe seizures. AJNR Am J Neuroradiol 24(8):1582-1586, 2003 38. Wehner T, Lapresto E, Tkach J, Liu P, Bingaman W, Prayson RA, et al: The value of interictal diffusion-weighted imaging in lateralizing temporal lobe epilepsy. Neurology 68(2):122-127, 2007 39. Rugg-Gunn FJ, Eriksson SH, Symms MR, Barker GJ, Duncan JS: Diffusion tensor imaging of cryptogenic and acquired partial epilepsies. Brain 124(3):627-636, 2001 40. Medina LS, Bernal B, Ruiz J: Role of functional MR in determining language dominance in epilepsy and nonepilepsy populations: a Bayesian analysis. Radiology 242:94-100, 2007 41. Benke T, Koylu B, Visani P, Karner E, Brenneis C, Bartha L, et al: Language lateralization in temporal lobe epilepsy: a comparison between fMRI and the Wada Test. Epilepsia 47(8):1308-1319, 2006 42. Sabbah P, Chassoux F, Leveque C, Landre E, Baudoin-Chial S, Devaux B, et al: Functional MR imaging in assessment of language dominance in epileptic patients. Neuroimage 18(2):460-467, 2003 43. Rabin ML, Narayan VM, Kimberg DY, Casasanto DJ, Glosser G, Tracy JI, et al: Functional MRI predicts post-surgical memory following temporal lobectomy. Brain 127(pt 10):2286-2298, 2004 44. Richardson MP, Strange BA, Thompson PJ, Baxendale SA, Duncan JS, Dolan RJ: Pre-operative verbal memory fMRI predicts post-operative memory decline after left temporal lobe resection. Brain 127(pt 11): 2419-2426, 2004