Abstract
Over 30% epileptic patients are refractory to medication, who are amenable to neurosurgical treatment. Non-invasive brain imaging technologies including video-electroencephalogram (EEG), magnetic resonance imaging (MRI), and magnetoencephalography (MEG) are widely used in presurgical assessment of epileptic patients. This review mainly discussed the current development of clinical MEG imaging as a diagnose approach, and its correlations with the golden standard intracranial electroencephalogram (iEEG). More importantly, this review discussed the possible applications of functional networks in preoperative epileptic foci localization in future studies.
Introduction
Epilepsy is defined as a brain disorder characterized predominantly by recurrent and unpredictable interruptions of normal brain function resulting from abnormal and excessive neuronal discharges (Fisher et al., ). Approximately a third of epileptic patients are refractory to medication, who are amenable to neurosurgical treatment if evidences clarify focal network underlying the epilepsy (Jobst and Cascino, ). The main purpose of presurgery evaluation is to localize the epileptic foci in order to direct a precise surgery plan for an individual patient. Various of non-invasive brain imaging technologies including video-electroencephalogram (EEG), magnetic resonance imaging (MRI), positron emission tomography (PET), single photon emission computerized tomography (SPECT), and magnetoencephalography (MEG) play crucial roles in presurgical assessments of epileptic patients (Duncan et al., ). However, when the non-invasive outcomes are inconsistent, intracranial-electroencephalogram (iEEG), which is the current “golden standard” for localization of seizures, is needed to be conducted to confirm the involvement of suspected brain regions.
Among those non-invasive technologies, MEG has held the promise for localizing epileptic zones for its predictions of favorable post-surgical outcomes (Zhang et al., ; Englot et al., ) and concordance with iEEG results (Almubarak et al., ; Grova et al., ; Murakami et al., ). One of the main advantages of MEG is its high spatial and temporal accuracy. By detecting the changes of magnetic fields produced by the electrical activities of neurons, epileptic spikes could be observed in MEG within only 4 cm2 cortical generators (Mikuni et al., ; Oishi et al., ; Shigeto et al., ). While a minimum area of 10 cm2 was required to detect spikes in scalp EEG (Tao et al., ).
Although MEG has numerous advantages in epileptic foci detection, it still can't replace the iEEG as an independent modality in presurgical assessment (Duncan et al., ). The most common MEG application now is to guide implantation sites for intracranial recordings (Knowlton et al., ; Agirre-Arrizubieta et al., ). One of the most important limitations exist in clinical MEG studies is the immature source reconstructing algorithm which is still developing. In the past decade, both the quantity and quality of MEG studies have remarkable increased, this state of the art technique seems promising to reveal brain mechanism of cognitive processing (Baillet, ). Going deep into understanding of human's brain, a large number of new concepts and perspectives are emerging. How could we take advantage of the exciting progresses made in cognitive MEG studies? This review attempted to discuss the recent development of clinical MEG studies and its application in preoperative epileptic foci localization.
Source reconstruction of meg data
Interictal epileptic discharge (IED), which represents the field summation of potentials from a population of pathologically synchronized bursting neurons, is generally considered as a robust biomarker that identifies the epileptic foci (Oishi et al., , ). However, the raw MEG signals are magnetic fields outside the human scalp (Cohen, ). A combination of MEG signals and structural MRI scans is needed to estimate the location and intensity of IEDs sources inside the brain which is called source localization. After artifact rejection and noise definition, source localization processing could be generally divided into two steps: (1) head modeling which establishes the electromagnetic properties of the sensor array and estimations of the brain sources; (2) source modeling imaging which solves an inverse problem (Tadel et al., ; Baillet, ).
Reconstructing of current dipoles, which are convenient model equivalents to the post-synaptic electrophysiological activity of local neuronal assemblies, is the basic concept in MEG source estimation process (Mosher et al., ). Among the current studies, two main approaches have been well adopted: single equivalent current dipoles (sECD), where the position and amplitude of one sECD is estimated over relatively short time windows, and distributed source modeling (DSM), which attempts to explain a given observed magnetic field by the distribution of source intensities in the brain instead of a single source (Tadel et al., ; Baillet, ).
The most common algorithm of the epileptic foci localization from clinical MEG recordings is sECD (Bagic et al., ). A well trained magnetoencephalographer is needed to extracts tightness and orientation of the dipole clusters visually (Salayev et al., ). Ideally, sECD can localize at best the center of the spatially extended generators as a point source (Ebersole, ). A retrospective study used sECD method revealed that patients with tight cluster and stable orientation dipole could significantly predict a better outcomes after surgery (Murakami et al., ).
However, the practicability of sECD method is debated not only for the subjective selection of IED period, but also for the possibility to mislead the presence of large spatially extended generators (Kobayashi et al., ). The latter is sophisticated because the IEDs were sometimes not confined to only one area (Agirre-Arrizubieta et al., ).
In the past decade, DSM has been largely promoted in cognitive MEG studies based on the fact that multiple brain areas would activate simulatonesly when a participant is performing a certain task (Tadel et al., ). DSM has also been applied in presurgical MEG evaluation using the nonlinear Maximum Entropy on the Mean (MEM) algorithm (Chowdhury et al., ; Grova et al., ; Heers et al., ; Zerouali et al., ), which could better predict the cortical regions with IEDs compared with other distributed models (Heers et al., ).
Correlation with intracranial data
To evaluate the effectiveness of epileptic foci localization in MEG, the agreement between the MEG and “gold standard” invasive EEG approaches, which mainly include electrocorticography (ECoG) and stereo-electroencephalogram (sEEG) should be considered. Although most studies focus on the evaluation consistency between MEG and ECoG (Oishi et al., , ; Knowlton et al., ; Almubarak et al., ), more and more studies has paid attention to the correlations between MEG and sEEG (Bouet et al., ; Grova et al., ; Murakami et al., ), which is considered to be minimal invasive and more suitable for preoperative evaluation (Gonzalez-Martinez et al., ; Zhou et al., ).
The effectiveness of MEG is depends on the depths of IED sources. Previous studies showed that MEG could detect 50–90% of spikes detected on iEEG superficial cortex including the lateral temporal cortex (Shigeto et al., ; Oishi et al., ; Agirre-Arrizubieta et al., ), while it could detect about 50–63% iEEG spikes in mesial temporal structures (Santiuste et al., ). These results are in general acceptable for presurgical assessments. Recent studies have revealed that epileptic activity from the insula, which is far away from scalp, can also be detected by MEG (Heers et al., ; Mohamed et al., ).
The limitation of iEEG approaches is it can only monitor a portion of brain areas, while MEG can provide a broad view of whole head activities and therefore may aid the interpretation of iEEG data. Murakami et al. () systematically examined the consistency between the interictal discharges in MEG with interictal and ictal discharges in sEEG in 50 retrospective cases. The results showed that the proportion of seizure-free after epileptic foci resection was significantly higher when sEEG completely sampled the area identified by MEG. This finding suggests that although the MEG is still unable to take the place of iEEG as the standard, more weight should be assigned to MEG outcomes in combination of multiple modalities in preoperative evaluation.
Epilepsy as a network diseases
One of the core questions in cognitive neuroscience is how different brain areas cooperate during perceptual/cognitive processing (Bressler and Menon, ). This hypothesis could also applied when epilepsy is considered as a network diseases (Laufs, ). Recent studies suggested that the brain connectivity differences among individuals could effectively interpret the patients with differing surgical outcomes (Englot et al., ; Munsell et al., ).
Traditional epileptic network is defined as discharge propagation network (Malinowska et al., ), which is correlated with patient surgical outcome (Tanaka et al., ). However, a portion of patients could not be captured epileptic discharges during MEG recording. Functional connectivity studies have been recently applied in epileptic patients based on the assumption that the normal functions of local networks are impaired in the foci areas. Evidence for altered focal functional connectivity in patients with temporal lobe epilepsy has been reported using resting-state functional MRI method (Haneef et al., ; Cataldi et al., ; Maccotta et al., ).
Unlike the extensive studies using function MRI modality, the resting-state MEG studies are rare. Krishnan et al. () conducted a preliminary MEG study and suggested that accurate localization of the epileptogenic foci may be accomplished using non-invasive spontaneous resting-state signals without the need to capture IEDs. Zerouali et al. () used resting-state MEG analysis to recognize insular epileptic foci and revealed that anterior and posterior sub-regions of the insula have distinct patterns of functional connectivity, which could further guide different surgery plans.
A combination of resting-state fMRI and MEG studies of functional connectivity could be adopted to investigate the pathophysiology of epileptic networks. However, whether there is potential benefit in prediction of epilepsy surgery outcome in individual level is not established yet. To achieve this goal, a large database of normal resting-state mapping should be established. With a temporally rich source of information on brain network dynamics, MEG was called upon to join in the Human Connectome Project (Larson-Prior et al., ). Three years later, Niso et al. () published an open archive resting-state MEG data of 97 health participants, which mainly focus on the distributions of typical oscillation frequency bands (delta, theta, alpha, beta, and gamma). Interestingly, a previous study shown that coherent neural activity change in the beta band might be causally involved in epilepsy (Heers et al., ), suggesting an exciting practicability of this MEG normative repository in preoperative evaluation for individual epilepsy patient.
To sum up, with the development of cognitive neuroscience, our understanding of normal and abnormal functional mechanism of the brain will get better and better. Using advanced MEG analysis technology, objective biomarks to locate epileptic foci is likely to replace the traditional subjective judgments in the upcoming future.
Statements
Author contributions
QW, PT, and GL all contributed to the writing and revision of this manuscript.
Funding
This work was supported by the National Natural Science Foundation of China (81671285), the China Postdoctoral Science Foundation (2016M601066), the Capital Health Research and Development of Special (2016-1-8012), Beijing Municipal Science and Technology Commission (Z161100000516230, Z161100002616016).
Conflict of interest
The authors declare that the research was conducted in the absence of any commercial or financial relationships that could be construed as a potential conflict of interest.
References
1
Agirre-ArrizubietaZ.HuiskampG. J. M.FerrierC. H.vanHuffelenA. C.LeijtenF. S. (2009). Interictal magnetoencephalography and the irritative zone in the electrocorticogram. Brain132, 3060–3071. 10.1093/brain/awp137
2
Agirre-ArrizubietaZ.ThaiN. J.ValentínA.FurlongP. L.SeriS.SelwayR. P.et al. (2014). The value of magnetoencephalography to guide electrode implantation in epilepsy. Brain Topogr.27, 197–207. 10.1007/s10548-013-0330-x
3
AlmubarakS.AlexopoulosA.Von-PodewilsF.WangZ. I.KakisakaY.MosherJ. C.et al. (2014). The correlation of magnetoencephalography to intracranial EEG in localizing the epileptogenic zone: a study of the surgical resection outcome. Epilepsy Res.108, 1581–1590. 10.1016/j.eplepsyres.2014.08.016
4
BagicA. I.KnowltonR. C.RoseD. F.EbersoleJ. S.ACMEGS Clinical Practice Guideline (CPG) Committee (2011). American clinical magnetoencephalography society clinical practice guideline 1: recording and analysis of spontaneous cerebral activity. J. Clin. Neurophysiol. 28, 348–354. 10.1097/WNP.0b013e3182272fed
5
BailletS. (2017). Magnetoencephalography for brain electrophysiology and imaging. Nat. Neurosci.20, 327–339. 10.1038/nn.4504
6
BouetR.JungJ.DelpuechC.RyvlinP.IsnardJ.GuenotM.et al. (2012). Towards source volume estimation of interictal spikes in focal epilepsy using magnetoencephalography. Neuroimage59, 3955–3966. 10.1016/j.neuroimage.2011.10.052
7
BresslerS. L.MenonV. (2010). Arge-scale brain networks in cognition: emerging methods and principles. Trends Cog. Sci.14, 277–290. 10.1016/j.tics.2010.04.004
8
CataldiM.AvoliM.Villers-SidaniE. (2013). Resting state networks in temporal lobe epilepsy. Epilepsia54, 2048–2059. 10.1111/epi.12400
9
ChowdhuryR. A.LinaJ. M.KobayashiE.GrovaC. (2013). MEG source localization of spatially extended generators of epileptic activity: comparing entropic and hierarchical bayesian approaches. PLoS ONE8:e55969. 10.1371/journal.pone.0055969
10
CohenD. (1968). Magnetoencephalography: evidence of magnetic fields produced by alpha-rhythm currents. Science161, 784–786. 10.1126/science.161.3843.784
11
DuncanJ. S.WinstonG. P.KoeppM. J.OurselinS. (2016). Brain imaging in the assessment for epilepsy surgery. Lancet Neurol.15, 420–433. 10.1016/S1474-4422(15)00383-X
12
EbersoleJ. S. (1997). Defining epileptogenic foci: past, present, future. J. Clin. Neurophysiol.14, 470–483. 10.1097/00004691-199711000-00003
13
EnglotD. J.HinkleyL. B.KortN. S.ImberB. S.MizuiriD.HonmaS. M.et al. (2015a). Global and regional functional connectivity maps of neural oscillations in focal epilepsy. Brain138, 2249–2262. 10.1093/brain/awv130
14
EnglotD. J.NagarajanS. S.ImberB. S.RaygorK. P.HonmaS. M.MizuiriD.et al. (2015b). Epileptogenic zone localization using magnetoencephalography predicts seizure freedom in epilepsy surgery. Epilepsia56, 949–958. 10.1111/epi.13002
15
FisherR. S.Van Emde BoasW.BlumeW.ElgerC.GentonP.LeeP.et al. (2005). Response: definitions proposed by the international league against epilepsy (ILAE) and the international bureau for epilepsy (IBE). Epilepsia46, 1701–1702. 10.1111/j.1528-1167.2005.00273_4.x
16
Gonzalez-MartinezJ.BulacioJ.AlexopoulosA.JehiJ.BingamanW.NajmI.et al. (2013). Stereoelectroencephalography in the “difficult to localize” refractory focal epilepsy: early experience from a north american epilepsy center. Epilepsia54, 323–330. 10.1111/j.1528-1167.2012.03672.x
17
GrovaC.AiguabellaM.ZelmannR.LinaJ. M.HallJ. A.KobayashiE.et al. (2016). Intracranial EEG potentials estimated from MEG sources: a new approach to correlate MEG and iEEG data in epilepsy. Hum. Brain Mapp.37, 1661–1683. 10.1002/hbm.23127
18
HaneefZ.LenartowiczA.YehH. J.EngelJ.Jr.SternJ. M. (2012). Effect of lateralized temporal lobe epilepsy on the default mode network. Epilepsy Behav.25, 350–357. 10.1016/j.yebeh.2012.07.019
19
HeersM.ChowdhuryR. A.HedrichT.DubeauF.HallJ. A.LinaJ. M.et al. (2016). Localization accuracy of distributed inverse solutions for electric and magnetic source imaging of interictal epileptic discharges in patients with focal epilepsy. Brain Topogr.29, 162–181. 10.1007/s10548-014-0423-1
20
HeersM.HedrichT.AnD.DubeauF.GotmanJ.GrovaC.et al. (2014). Spatial correlation of hemodynamic changes related to interictal epileptic discharges with electric and magnetic source imaging. Hum. Brain Mapp.35, 4396–4414. 10.1002/hbm.22482
21
HeersM.RamppS.StefanH.UrbachH.ElgerC. E.von LeheM.et al. (2012). MEG-based identification of the epileptogenic zone in occult peri-insular epilepsy. Seizure21, 128–133. 10.1016/j.seizure.2011.10.005
22
JobstB. C.CascinoG. D. (2015). Resective epilepsy surgery for drug-resistant focal epilepsy: a review. JAMA313, 285–293. 10.1001/jama.2014.17426
23
KnowltonR. C.RazdanS. N.LimdiN.ElgavishR. A.KillenJ.BlountJ.et al. (2009). Effect of epilepsy magnetic source imaging on intracranial electrode placement. Ann. Neurol.65, 716–723. 10.1002/ana.21660
24
KobayashiK.YoshinagaH.OhtsukaY.GotmanJ. (2005). Dipole modeling of epileptic spikes can be accurate or misleading. Epilepsia46, 397–408. 10.1111/j.0013-9580.2005.31404.x
25
KrishnanB.VlachosI.WangZ. I.MosherJ.NajmI.BurgessR.et al. (2015). Epileptic focus localization based on resting state interictal MEG recordings is feasible irrespective of the presence or absence of spikes. Clin. Neurophysiol.126, 667–674. 10.1016/j.clinph.2014.07.014
26
Larson-PriorL. J.OostenveldR.Della PennaS.MichalareasG.PriorF.Babajani-FeremiA.et al. (2013). Adding dynamics to the human connectome project with MEG. Neuroimage80, 190–201. 10.1016/j.neuroimage.2013.05.056
27
LaufsH. (2012). Functional imaging of seizures and epilepsy: evolution from zones to networks. Curr. Opin. Neurol.25, 194–200. 10.1097/WCO.0b013e3283515db9
28
MaccottaL.HeB. J.SnyderA. Z.EisenmanL. N.BenzingerT. L.AncesB. M.et al. (2013). Impaired and facilitated functional networks in temporal lobe epilepsy. Neuroimage Clin.2, 862–872. 10.1016/j.nicl.2013.06.011
29
MalinowskaU.BadierJ. M.GavaretM.BartolomeiF.ChauvelP.BénarC. G. (2014). Interictal networks in magnetoencephalography. Hum. Brain Mapp.35, 2789–2805. 10.1002/hbm.22367
30
MikuniN.NagamineT.IkedaA.TeradaaK.TakiW.KimuraJ.et al. (1997). Simultaneous recording of epileptiform discharges by MEG and subdural electrodes in temporal lobe epilepsy. Neuroimage5, 298–306. 10.1006/nimg.1997.0272
31
MohamedI. S.GibbsS. A.RobertM.BouthillierA.LerouxJ.-M.NguyenD. K. (2013). The utility of magnetoencephalography in the presurgical evaluation of refractory insular epilepsy. Epilepsia54, 1950–1959. 10.1111/epi.12376
32
MosherJ. C.LewisP. S.LeahyR. M. (1992). Multiple dipole modeling and localization from spatio-temporal MEG data. IEEE Trans. Biomed Eng.39, 541–557. 10.1109/10.141192
33
MunsellB. C.WeeC. Y.KellerS. S.WeberdB.ElgerdC.da SilvaaA. T.et al. (2015). Evaluation of machine learning algorithms for treatment outcome prediction in patients with epilepsy based on structural connectome data. Neuroimage118, 219–230. 10.1016/j.neuroimage.2015.06.008
34
MurakamiH.WangZ. I.MarashlyA.KrishnanB.PraysonR. A.KakisakaY.et al. (2016). Correlating magnetoencephalography to stereo-electroencephalography in patients undergoing epilepsy surgery. Brain139:aww215. 10.1093/brain/aww215
35
NisoG.RogersC.MoreauJ. T.MoreauaJ. T.ChenaL.-Y.MadjarbC.et al. (2015). Omega: the open MEG archive. Neuroimage124(Pt B), 1182–1187. 10.1016/j.neuroimage.2015.04.028
36
OishiM.KameyamaS.MasudaH.MasudaH.TohyamaJ.KanazawaO.et al. (2006). Single and multiple clusters of magnetoencephalographic dipoles in neocortical epilepsy: significance in characterizing the epileptogenic zone. Epilepsia47, 355–364. 10.1111/j.1528-1167.2006.00428.x
37
OishiM.OtsuboH.KameyamaS.MorotaN.MasudaH.KitayamaM.et al. (2002). Epileptic spikes: magnetoencephalography versus simultaneous electrocorticography. Epilepsia43, 1390–1395. 10.1046/j.1528-1157.2002.10702.x
38
SalayevK. A.NakasatoN.IshitobiM.ShamotoH.KannoA.IinumaK. (2006). Spike orientation may predict epileptogenic side across cerebral sulci containing the estimated equivalent dipole. Clin. Neurophysiol.117, 1836–1843. 10.1016/j.clinph.2006.05.003
39
SantiusteM.NowakR.RussiA.AranconT.OliverB.AyatsE.SchelerG.GraetzG.et al. (2008). Simultaneous magnetoencephalography and intracranial EEG registration: technical and clinical aspects. J. Clin. Neurophysiol.25, 331–339. 10.1097/WNP.0b013e31818e7913
40
ShigetoH.MoriokaT.HisadaK.NishioS.IshibashiH.KiraD.-I.et al. (2002). Feasibility and limitations of magnetoencephalographic detection of epileptic discharges: simultaneous recording of magnetic fields and electrocorticography. Neurol. Res.24, 531–536. 10.1179/016164102101200492
41
TadelF.BailletS.MosherJ. C.MosherJ. C.PantazisD.LeahyR. M. (2011). Brainstorm: a user-friendly application for MEG/EEG analysis. Comput. Intell. Neurosci.2011, 879716. 10.1155/2011/879716
42
TanakaN.PetersJ. M.ProhlA. K.TakayaS.MadsenJ. R.BourgeoisB. F.et al. (2014). Clinical value of magnetoencephalographic spike propagation represented by spatiotemporal source analysis: correlation with surgical outcome. Epilepsy Res.108, 280–288. 10.1016/j.eplepsyres.2013.11.006
43
TaoJ. X.BaldwinM.Hawes-EbersoleS.EbersoleJ. S. (2007). Cortical substrates of scalp EEG epileptiform discharges. J. Clin. Neurophysiol.24, 96–100. 10.1097/WNP.0b013e31803ecdaf
44
ZeroualiY.PouliotP.RobertM.MohamedI.BouthillierA.LesageF.et al. (2016). Magnetoencephalographic signatures of insular epileptic spikes based on functional connectivity. Hum. Brain Mapp.37, 3250–3261. 10.1002/hbm.23238
45
ZhangR.WuT.WangY.LiuH.ZouY.LiuW.et al. (2011). Interictal magnetoencephalographic findings related with surgical outcomes in lesional and nonlesional neocortical epilepsy. Seizure20, 692–700. 10.1016/j.seizure.2011.06.021
46
ZhouJ.HuangL. T.ZhaiF.GuanY, -G.BaoM.ZhaoM.et al. (2017). The role of ROSA navigated intracranial electrode implantation technique on precise epileptogenic zone localization. Neuropsychiatry7, 258–264. 10.4172/Neuropsychiatry.1000208
Summary
Keywords
magnetoencephalographic (MEG), intracranial electroencephalogram (iEEG), distributed source modeling (DSM), resting-state, functional connectivity
Citation
Wang Q, Teng P and Luan G (2017) Magnetoencephalography in Preoperative Epileptic Foci Localization: Enlightenment from Cognitive Studies. Front. Comput. Neurosci. 11:58. doi: 10.3389/fncom.2017.00058
Received
15 May 2017
Accepted
12 June 2017
Published
28 June 2017
Volume
11 - 2017
Edited by
Yang Xia, University of Electronic Science and Technology of China, China
Reviewed by
Honghui Zhang, Northwestern Polytechnical University, China; Qishao Lu, Beihang University, China
Updates
Copyright
© 2017 Wang, Teng and Luan.
This is an open-access article distributed under the terms of the Creative Commons Attribution License (CC BY). The use, distribution or reproduction in other forums is permitted, provided the original author(s) or licensor are credited and that the original publication in this journal is cited, in accordance with accepted academic practice. No use, distribution or reproduction is permitted which does not comply with these terms.
*Correspondence: Guoming Luan luangm3@163.com
Disclaimer
All claims expressed in this article are solely those of the authors and do not necessarily represent those of their affiliated organizations, or those of the publisher, the editors and the reviewers. Any product that may be evaluated in this article or claim that may be made by its manufacturer is not guaranteed or endorsed by the publisher.