This Article


Creative Commons License
Except where otherwise noted, this work is licensed under Creative Commons Attribution-NonCommercial 4.0 International License.

Value of Proton-MR-Spectroscopy in the Diagnosis of Temporal Lobe Epilepsy; Correlation of Metabolite Alterations With Electroencephalography

1 MRI Department, S.B. Ankara Diskapi Yildirim Beyazit Education and Research Hospital, Ankara, Turkey
2 Vascular Interventional Department, S.B. Ankara Diskapi Yildirim Beyazit Education and Research Hospital, Ankara, Turkey
*Corresponding author: Volkan Kizilgoz, MRI Department, S.B. Ankara Diskapi Yildirim Beyazit Education and Research Hospital, Irfan Bastug St.,, Ankara, Turkey. Tel.: +90-5057994013, Fax: +90-3123220006, E-mail:
Iranian Journal of Radiology. 2012 March; 9(1): 1-11. , DOI: 10.5812/iranjradiol.6686
Article Type: Research Article; Received: Apr 3, 2010; Revised: Dec 10, 2011; Accepted: Dec 19, 2011; epub: Mar 25, 2012; ppub: Mar 2012
Running Title: Proton-MR-Spectroscopy in the Diagnosis of TLE


Background: Epilepsy, a well-known mostly idiopathic neurologic disorder, has to be
correctly diagnosed and properly treated. Up to now, several diagnostic approaches have
been processed to determine the epileptic focus.

Objectives: The aim of this study was to discover whether proton-MR-spectroscopic imaging
(MRSI) aids in the diagnosis of temporal lobe epilepsy in conjunction with classical
electroencephalography (EEG) findings.

Patients and Methods: Totally, 70 mesial temporal zones consisting of 39 right hippocampi
and 31 left hippocampi of 46 patients (25 male, 21 female) were analyzed by proton
MRSI. All patients underwent a clinical neurologic examination, scalp EEG recording and
prolonged video EEG monitoring. Partial seizures on the right, left or both sides were
recorded in all patients. All patients were under medical treatment and none of the patients
underwent amygdalohippocampectomy and similar surgical procedures.

Results: The normal average lactate (Lac), phosphocreatine, N-acetyl aspartate (NAA),
creatine (Cr), choline (Cho), myo-inositol, glutamate and glutamine (Glx) peaks and Nacetyl
aspartate/Cr, NAA/ Cho + Cr, Cho/Cr ratios were measured from the healthy opposite
hippocampi or from the control subjects. The Lac, glutamate and glutamine (Glx),
myo-inositol, phosphocreatine and NAA metabolites plus Cho/Cr ratio showed statistical
difference between the normal and the epileptic hippocampi. Cho, Cr metabolites plus
NAA/Cr, NAA/ Cho + Cr ratios were almost the same between the groups. The sensitivity
of Proton-MR-Spectroscopy for lateralization of the epileptic foci in all patients was 96%
and the specificity was 50%.

Conclusions: Proton-MRSI can easily be considered as an alternative modality of choice
in the diagnosis of temporal lobe epilepsy and in the future; Proton-MR-Spectroscopy
may become the most important technique used in epilepsy centers.

Keywords: Magnetic Resonance Spectroscopy Epilepsy Temporal Lobe Electroencephalography

1. Background

Epilepsy is a group of neurological abnormalities in which the most common and fundamental characteristic is recurrent and mostly unprovoked seizures (1). The subgroups are idiopathic epilepsies which occur in a structurally normal brain and have a presumably genetic cause and symptomatic epilepsies which are caused by a known structural, focal or diffuse abnormality (1, 2). The most common class of symptomatic epilepsies is the mesial temporal lobe epilepsy (MTLE) frequently associated with hippocampal sclerosis (1-4). The most important tools in the diagnosis of MTLE are the clinical neurological features and the intensive video-EEG monitoring (1, 3-6). Neuropsychological evaluation, brain magnetic resonance imaging (MRI), positron emission tomography (PET) and single photon emission computed tomography (SPECT) imaging are the other diagnostic approaches in the assessment of epileptic foci (3-7). Cerebral MRI usually shows atrophy or high signal intensity in the affected hippocampus or amygdala; however, in some patients with MTLE, it may reveal no abnormalities (1, 3, 7-9). Proton-MR-Spectroscopy (H-MRS), a non-invasive technique, has been suggested for its usefulness in the evaluation of MTLE and epileptogenic areas. By detection of metabolites in the mesial temporal structures (hippocampus and the amygdala), H-MRS can give sensitive and reliable results, may aid in the lateralization of epileptic foci and the diagnosis of MTLE (1, 3, 4, 6-11). Lateralization is a procedure that is used to determine the hemisphere which is responsible for the genesis of seizures (1, 4, 6, 7, 11).

2. Objectives

In this study, the aim was to discover the potential benefits of H-MRS in the lateralization of epileptic foci in the mesial temporal parts, mainly the hippocampi, correlated with the EEG findings and the clinical features.

3. Patients and Methods

Forty-six consecutive patients (25 male-21 female) in the age range of 11-62 years (mean, 35 ± 4 years) were included in this study. Thirty-nine right hippocampi and 31 left hippocampi, totally 70 mesial temporal zones were analyzed in these patients. Twenty-two of the patients had bilateral and 24 had unilateral hippocampal involvement. All patients were evaluated by the neurology department of our hospital. In the MRI, they had no neoplastic or traumatic lesions. Only one patient had a subacute hematoma in the left temporal lobe. All patients underwent a clinical neurologic examination, scalp EEG recording and prolonged video EEG monitoring. Complex partial seizures on the right, left or both sides were recorded in all patients and none of the patients had a second seizure focus. All patients were under medical treatment and none of the patients underwent amygdalohippocampectomy and similar surgical procedures. Follow up of the patients and duration of epilepsy were about 2 months to 26 years. In this group; there were chronic epileptics-late epilepsy patients and also patients with acute onset seizures. All the patients and the healthy control group participated in this study with their own consent.

All the MRIs and spectroscopic analysis were carried out via 1.5 T (Philips Achieva, 8 Channel, Philips Medical System, Netherlands) scanner using a standard head coil. Routine brain MRI was performed using 3D-Flair, T2 weighted fast field echo (FFE) and T2 weighted turbo spin echo (TSE) coronal sequences. Axial T1 and T2 weighted TSE sequences were also applied. Multi-voxel spectroscopic imaging (MV-MRS) was performed using point-resolved spectroscopy (PRESS) with a standard volume of interest (VOI) size, 0.5 × 0.5 × 1 cm for all hippocampi. We positioned the voxel over the mesial temporal lobe to cover most of the hippocampus and tried to minimize partial-volume effects resulting from other neighbouring tissues including amygdala, cerebral spinal fluid (CSF) of the ventricles and parahippocampal gyrus. The voxel mostly comprised the head and anterior part of the hippocampi. Time domain data were multiplied by a Gaussian function of 90 (Center 0, halfwidth 256 ms), 2D Fourier transformed phase and base-line corrected, quantified by means of frequency domain curve fitting with the assumption of a Gaussian line shape using Phil-ips software. A 0-4.35 ppm was analyzed and metabolite signal peaks were centered as follows; N-acetyl aspartate (NAA) at 2 ppm, creatine (Cr) at 3-3.1 ppm, phosphocreatine (Cr2) at 3.8-3.9 ppm, choline (Cho) at 3.2 ppm, lactate (Lac) at 1.3-1.4 ppm, glutamate and glutamine (Glx) at 2.45 ppm, glycine and myo-inositol (Gly-MI) at 3.6-3.75 ppm (1, 7, 10-12). Detailed Glx-MI-Cr2 metabolite concentrations were briefly presented in TE: 26 ms and NAA-Cho-Cr-Lac metabolites were mostly analyzed in TE: 144 ms. Automatic shimming of the linear x, y, z channels was used to optimize field homogeneity, water resonance and water suppression pulses were optimized. Proton spectra were recorded in the coronal plane with T2 weighted images via TR; 1500 ms, TE; 26 and 144 ms, FOV; 24 × 24 cm, 1-1.5 cm section thickness, 256 × 256 matrix and 24 × 24 phase encoding. Spectral analysis and last processing were carried out using Philips software (Philips Achieva, 8 Channel, Netherlands) work-shop. All the MRSI data and metabolite concentrations were evaluated by two radiologists, 3 and 7 years experienced, together with consensus so there was no intra- and interobserver variability.

Lateralization of epileptic foci using MRSI was performed in two ways; the first lateralization involved comparison of metabolite concentrations and metabolite peak ratios in the ipsilateral hippocampal region with those on the opposite side, this was applied when there was unilateral involvement in EEG. When there was bilateral involvement in EEG, we compared the hippocampal data of patients upon the mesial temporal zone data of five healthy control subjects. The control group was selected from non-epileptic patients with their own willingness to participate in this research who were referred to our unit for cranial MRI due to headache, dizziness, vertigo and similar symptoms. These five healthy controls had normal cerebral and cerebellar MRI findings. Values more than 2 standard deviations below the mean data of control subjects were considered abnormal. The control group was composed of healthy subjects with an unknown seizure history and no EEG abnormality, aged between 20 and 45 years (mean, 32 ± 3 years).

All statistical analyses were performed using a software program, Statistical Package for the Social Sciences (SPSS for Windows ver. 15.0, Chicago-Illinois). Measurements of metabolite peaks, their alterations and metabolite ratios were analyzed by paired t-test for both hippocampal involvement and single-sampling t-test for all patients including bilateral and unilateral involvement. P < 0.05 was considered to be statistically significant. Receiver operating characteristic (ROC) curve analysis for metabolite peaks, ratios and cut-off points with regard to area under curve (AUC) were also determined.

5. Discussion

The new imaging techniques over the past 20 years have enabled the non-invasive analysis of abnormalities, preceding epileptic seizures and also allowed the use of these approaches in order to reveal the etiologic mechanisms of epilepsy (1, 3). Many diagnostic imaging tools such as brain MRI, PET, SPECT and H-MRS may be administered and correlated with the results of classical EEG and used in the analysis of epilepsy (1, 5, 7). MRI and H-MRS offer a lot of qualitative and quantitative data which can help to localize the epileptic lesions and provide an insight into the biophysical and biochemical processes related to epileptic seizures (1, 4-6). The early reports for H-MRS in patients with MTLE were published in the 1990s which tried to lateralize the focal epileptic foci (1, 7, 9, 12-14). Examination of patients with epilepsy Proton-MR-Spectroscopy in the Diagnosis of TLEusing MRSI is now focused on the observation of changes in NAA, Cr, Cr2, Cho, Glx and GABA (Gama-aminobutyric acid) signals, their correlations with the results of MRI, EEG findings and the clinical features (1, 3, 5-7, 10, 11). In this paper, H-MRS was discussed as a method for lateralization or localization of the epileptogenic zone in MTLE. It was the best-recognized surgically proven epilepsy syndrome; the epileptogenic zone consisted of the amygdalo-hippocampal complex (1, 3-13). Approximately, 60% of MTLE patients, were predicted with severe unilateral hippocampal atrophy and a typical macroscopic feature of hippocampal sclerosis, also called Ammon’s horn sclerosis or mesial temporal sclerosis (MTS); ‘’Hippocampi of these patients revealed astrogliosis with prominent neuronal cell losses histopathologically’’ (1, 5, 10, 14).

MRI and at the same time H-MRS, which are non-invasive techniques, play an extremely important role in the diagnosis of hippocampal sclerosis and MTLE; MRI shows an atrophic-small or firm hippocampus and also hyper- intense, diffuse edematous areas at the hippocampi on T2W images (1, 4, 6, 8, 13). H-MRS predicted the metabolite peaks, measured in different parts of both hippocampi and NAA decrease in the spectrum usually provides the lateralization and localization of the epileptic focus (1, 5-8, 10, 12-14). As mentioned above, the most important signal in the MRSI spectra is NAA which describes the neuronal loss and dysfunction. NAA is synthesized in the mitochondria of neurons from aspartate and acetyl-CoA and its synthesis is controlled by aspartate-N-acetyltransferase (1, 5, 14, 15). The other important metabolites, especially for the treatment of epilepsy are Glx and GABA, of which Glx is the main excitatory neurotransmitter and GABA is the main inhibitory neurotransmitter (1, 3). With the impaired function of GABA, Glx becomes higher lead- ing to hyperexcitability and spontaneous epileptic activity; therefore, administration of GABAergic anti-epileptic drugs can easily improve and control the seizures (1, 3, 15). Previous H-MRS studies have shown elevation of Glx in patients with MTLE without the evidence of hippocampal sclerosis on MRI (3, 16).


We are grateful to Mr. Ahmet Yilmaz, Erdeniz Yurdakul, Şukru Yilmaz and Egemen Alper for their assistance to manage the MRSI procedures of all patients, processing the figures and statistical analysis of the output.


Implication for health policy/practice/research/medical education: Proton-MR-Spectroscopy can be considered as an alternative modality of choice in the diagnosis of temporal lobe epilepsy. Seizures and convulsions have to be treated urgently as fast as possible and by the rapid development of technology plus medical sciences, Proton-MR-Spectroscopy may become the most important technique used in epilepsy centers in order to analyse the hippocampusamygdaloid complex and to search for the effectiveness of medical anti-epileptic therapy.
Please cite this paper as: Aydin H, Oktay NA, Kizilgoz V, Altin E, Tatar IG, Hekimoglu B. Value of Proton-MR-Spectroscopy in the Diagnosis of Temporal Lobe Epilepsy; Correlation of Metabolite Alterations With Electroencephalography. Iran J Radiol. 2012;9(1):1-11.
Financial Disclosure : There was no financial disclosure for the research.
Funding/Support: There was no funding support for the research.


  • 1. Hajek M, Dezortova M, Krsek P. (1)H MR spectroscopy in epilepsy. Eur J Radiol. 2008;67(2):258-267. [DOI] [PubMed]
  • 2. Hajek M, Dezortova M. Introduction to clinical in vivo MR spectroscopy. Eur J Radiol. 2008;67(2):185-193. [DOI] [PubMed]
  • 3. Simister RJ, McLean MA, Barker GJ, Duncan JS. Proton MR spectroscopy of metabolite concentrations in temporal lobe epilepsy and effect of temporal lobe resection. Epilepsy Res. 2009;83(2-3):168-176. [DOI] [PubMed]
  • 4. Cohen-Gadol AA, Pan JW, Kim JH, Spencer DD, Hetherington HH. Mesial temporal lobe epilepsy: a proton magnetic resonance spectroscopy study and a histopathological analysis. J Neurosurg. 2004;101(4):613-20. [DOI] [PubMed]
  • 5. Vielhaber S, Niessen HG, Debska-Vielhaber G, Kudin AP, Wellmer J, Kaufmann J, et al. Subfield-specific loss of hippocampal N-acetyl aspartate in temporal lobe epilepsy. Epilepsia. 2008;49(1):40-50. [DOI] [PubMed]
  • 6. Hammen T, Schwarz M, Doelken M, Kerling F, Engelhorn T, Stadlbauer A , et al. 1H-MR spectroscopy indicates severity markers in temporal lobe epilepsy: correlations between metabolic alterations, seizures, and epileptic discharges in EEG. Epilepsia. 2007;48(2):263-269. [DOI] [PubMed]
  • 7. Thompson JE, Castillo M, Kwock L, Walters B, Beach R. Usefulness of proton MR spectroscopy in the evaluation of temporal lobe epilepsy. AJR Am J Roentgenol. 1998;170(3):771-776. [PubMed]
  • 8. Li LM, Cendes F, Antel SB, Andermann F, Serles W, Dubeau F, et al. Prognostic value of proton magnetic resonance spectroscopic imaging for surgical outcome in patients with intractable temporal lobe epilepsy and bilateral hippocampal atrophy. Ann Neurol. 2000;47(2):195-2000. [PubMed]
  • 9. Cendes F, Caramanos Z, Andermann F, Dubeau F, Arnold DL . Proton magnetic resonance spectroscopic imaging and magnetic resonance imaging volumetry in the lateralization of temporal lobe epilepsy: a series of 100 patients. Ann Neurol. 1997;42(5):737-746. [DOI] [PubMed]
  • 10. Li LM, Cendes F, Andermann F, Dubeau F, Arnold DL. Spatial extent of neuronal metabolic dysfunction measured by proton MR spectroscopic imaging in patients with localization-related epilepsy. Epilepsia. 2000;41(6):666-674. [DOI] [PubMed]
  • 11. Hammen T, Dolken M, Schwarz M, Kerling F, Engelhorn T, Stadlbauer A, et al. Correlation between metabolic alterations in 1H-MR spectroscopy and epileptic activity in patients with temporal lobe epilepsy. Clin Neurophysiol. 2007;118(4) [DOI]
  • 12. Achten E, Boon P, Van De Kerckhove T, Caemaert J, De Reuck J, Kunnen M. Value of single-voxel proton MR spectroscopy in temporal lobe epilepsy. AJNR Am J Neuroradiol. 1997;18(6):1131-1139. [PubMed]
  • 13. Ende GR, Laxer KD, Knowlton RC, Matson GB, Schuff N, Fein G, et al. Temporal lobe epilepsy: bilateral hippocampal metabolite changes revealed at proton MR spectroscopic imaging. Radiology. 1997;202(3):809-817. [PubMed]
  • 14. Vermathen P, Ende G, Laxer KD, Knowlton RC, Matson GB, Weiner MW. Hippocampal N-acetylaspartate in neocortical epilepsy and mesial temporal lobe epilepsy. Ann Neurol. 1997;42(2):194-199. [DOI] [PubMed]
  • 15. Petroff OA, Mattson RH, Rothman DL. Proton MRS: GABA and glutamate. Adv Neurol. 2000;83:261-271. [PubMed]
  • 16. Simister RJ, Woermann FG, McLean MA, Bartlett PA, Barker GJ, Duncan JS. A short-echo-time proton magnetic resonance spectroscopic imaging study of temporal lobe epilepsy. Epilepsia. 2002;43(9):1021-1031.
  • 17. Kuzniecky R, Palmer C, Hugg J, Martin R, Sawrie S, Morawetz R, et al. Magnetic resonance spectroscopic imaging in temporal lobe epilepsy: neuronal dysfunction or cell loss? Arch Neurol. 2001;58(12):2048-2053. [DOI] [PubMed]
  • 18. McLean MA, Woermann FG, Simister RJ, Barker GJ, Duncan JS. In vivo short echo time 1H-magnetic resonance spectroscopic imaging (MRSI) of the temporal lobes. Neuroimage. 2001;14(2):501-509. [DOI] [PubMed]

Table 1

Metabolite (Lac, NAA, Cho, Cre, Glx, MI, Cr2, Cho/Cr, NAA/Cho + Cr, NAA/Cr) Peak Concentrations of the Right and Left Hippocampus, Analysed by t test (P < 0.05)

Measurement Location No. Mean ± SD Test value = 0
t SD P value
Lac a
Right 39 3.86 ± 2.05 11.608 38 < 0.001
Left 31 4.41 ± 2.68 9.163 30 < 0.001
Right 39 0.67 ± 0.58 -0.137 38 < 0.001
Left 31 0.73 ± 0.67 -0.322 30 < 0.001
Cho a
Right 39 0.75 ± 0.72 1.895 38 0.066
Left 31 0.65 ± 0.55 1.266 30 0.215
Cre a
Right 39 0.72 ± 0.76 1.799 38 0.080
Left 31 0.56 ± 0.43 0.724 30 0.474
Glx a
Right 39 1.15 ± 1.19 4.646 38 < 0.001
Left 31 1.16 ± 1.07 4.773 30 < 0.001
MI a
Right 39 0.76 ± 0.61 6.137 38 < 0.001
Left 31 0.60 ± 0.42 5.951 30 < 0.001
Cr2 a
Right 39 0.78 ± 0.95 4.219 38 < 0.001
Left 31 0.54 ± 0.54 4.323 30 < 0.001
Cho/Cr a
Right 39 1.44 ± 0.82 4.076 38 < 0.001
Left 31 1.51 ± 0.77 3.956 30 < 0.001
NAA/Cho + Cr a
Right 39 0.71 ± 0.90 -2.011 38 0.052
Left 31 1.00 ± 1.31 -0.020 30 0.984
NAA/ Cr a
Right 38 1.44 ± 1.97 -0.024 38 0.981
Left 31 2.20 ± 2.59 1.624 30 0.115
a Abbreviations: Cho, Choline; Cho/Cr, Choline/Creatine; Cre, Creatine; Cr2, Phosphocreatine; Glx, Glutamin-Glutamate; Lac, Lactate; MI, Myo-inositol; NAA, N-Asetyl Aspartate; NAA/Cho + Cr, N-Asetyl Aspartate/Choline + Creatine; NAA/Cr, N Asetyl Aspartate/Creatine

Table 2

Sensitivitiy and Specificity of MRS

ocation MRS a Negative + EEG a Negative MRS Negative + EEG Positive MRS Positive + EEG Negative MRS Positive + EEG Positive Total
Both 1 (25) 3 (75) 1 (2) 65 (98) 70 (100)
Right 1 (25) 3 (75) 0 (0) 35 (100) 39 (100)
Left 0 (0) 0 (0) 1 (0) 29 (100) 31 (100)
a Abbreviation: EEG, electro-encephalography; MRS, proton MR spectorscopy

Figure 1

A 17-year-old girl presenting with seizure
H-MRS regarding major and dominant lactate peak at the left hippocampus

Figure 2

ROC curve analysis for lactate peak

Figure 3

A 19-year-old man with a complaint of headache and a history of seizure
MRSI revealed a markedly decreased N-acetyl aspartate peak in the right hippocampus.

Figure 4

Roc curve analysis for Cho peak

Figure 5

Roc curve analysis for Cre peak

Figure 6

A 22-year-old girl with positive EEG and clinical suspicion for epilepsy
H-MRS presented a dominancy of huge glutamate and glutamine peak in the right hipocampus.

Figure 7

ROC curve analysis for Glx peak

Figure 8

ROC curve analysis for MI peak

Figure 9

A 40-year-old man with late epilepsy
MRSI predicted a dominant phosphocreatine peak in the left hippocampus.

Figure 10

A 30-year-old man with dizziness, mild confusion and a history of epilepsy
H-MRS revealed increased Cho/Cre ratio in the left hippocampus.

Figure 11

ROC curve analysis for Cho/Cr ratio

Figure 12

ROC curve analysis for NAA/Cr ratio