www.mjms.usm.my © Penerbit Universiti Sains Malaysia, 2015 For permission, please email:mjms.usm@gmail.com

Abstract Background: Magnetoencephalography(MEG)hasbeenextensivelyusedtomeasuresmall-scaleneuronalbrainactivity.Althoughitiswidelyacknowledgedasasensitivetoolfordecipheringbrainactivityandsourcelocalisation,theaccuracyoftheMEGsystemmustbecriticallyevaluated.Typically,on-sitecalibrationwith theprovidedphantom(Local phantom) isused.However, thismethod is still questionable due to the uncertainty thatmay originate from the phantom itself.Ideally,thevalidationofMEGdatameasurementswouldrequirecross-sitecomparability. Method: Asimplemethodofphantomtestingwasusedtwiceinadditiontoameasurementtaken with a calibrated reference phantom (RefPhantom) obtained from Elekta Oy of Helsinki,Finland.Thecomparisonsoftwomainaspectsweremadeintermsofthedipolemoment(Qpp)andthedifferenceinthedipoledistancefromtheorigin(d)afterthetestsofstatisticallyequalmeansandvariancewereconfirmed. Result: TheresultofQppmeasurementsforthe LocalPhantomandRefPhantom were978(SD24)nAmand988(SD32)nAm,respectively,andwerestilloptimallywithintheacceptedrangeof900to1100nAm.Moreover,theshifteddresultsfortheLocalPhantomandRefPhantomwere1.84mm(SD0.53)and2.14mm(SD0.78),respectively,andthesevalueswerebelowthemaximumacceptancerangeofwithin5.0mmofthenominaldipolelocation. Conclusion: The localphantomseems tooutperform the referencephantomas indicatedby the small standard error of the former (SE 0.094) comparedwith the latter (SE 0.138). TheresultindicatedthatHUSMMEGsystemwasinexcellentworkingconditionintermsofthedipolemagnitudeandlocalisationmeasurementsasthesevaluespassedtheacceptancelimitscriteriaofthephantomtest.

Keywords: magnetoencephalography, validation, phantom

Introduction

Non-invasive techniques have becomeimportant brain mapping techniques in thestudyofbrain function. In the lastfivedecades,the non-invasive functional imaging technologyhas advanced in a wide range of modalities,including magnetoencephalography (MEG) andelectroencephalography (EEG). MEG-recordedmagneticfieldsareproducedbyelectricalactivity

within thebrainand recordedoutside thehead,whereas EEG is based on themeasurements ofpotentialdifferencesonthescalpthatresultfromohmiccurrentsinducedbyelectricalbrainactivity(1).Bothmodalitiesmeasuretheelectromagneticsignalsproducedby the electrical activity in thebrain. However,MEG possessmore advantagesthanEEGbecausetheneuronalsignalsgenerated

Special Issue - Neuroscience

Magnetoencephalography Phantom Comparison and Validation: Hospital Universiti Sains Malaysia (HUSM) RequisiteHazim Omar1, Alwani Liyana ahmad1, Noburo hayashi2, Zamzuri idris3,4, Jafri Malin abdullah3,4

1 Department of Neurosciences, Hospital Universiti Sains Malaysia, Jalan Hospital USM, 16150 Kubang Kerian, Kelantan, Malaysia

2 Elekta K K Shibaura Renasite Tower 3-9-1 Shibaura, Minato-ku, Tokyo 108-0023 Japan

3 Department of Neurosciences, School of Medical Sciences, Universiti Sains Malaysia, 16150 Kubang Kerian, Kelantan, Malaysia

4 Center for Neuroscience Services & Research, Universiti Sains Malaysia, 16150 Kubang Kerian, Kelantan, Malaysia

Submitted:20Oct2015Accepted:5Nov2015

20Malays J Med Sci. Dec 2015; Special Issue: 20-28

Special Issue - Neuroscience |HUSMMEGPhantomComparison&Validation

www.mjms.usm.my 21

in the cortex pass several layers of tissue withdifferent electrical properties and complexgeometriesbeforetheyreachthescalp;thus,theelectricalfieldsthatarerecordedatthescalparedistorted. These issues have a weaker influenceon the magnetic fields because the tissuessurroundingthebrainexhibitaconstantmagneticpermeability(2).Nevertheless,eachmodalityhasits own advantages and disadvantages becauseno singlemethod is best suited for all researchorclinicalpurposes.Forexample,bothMEGandEEGhaveabasiclimitationinthattheneuronalsignalsareonlyrecordedfromthescalp,whichislesssensitivethanrecordingfromdeepsources(3).ThelocalisationofEEGandMEGsourceactivitiesinthebrainisdiscombobulatedbecausethereisan indefinite number of source configurationsthat could give rise to the samemeasurements.Thisproblemisknownastheinverseproblem(4). MEGandEEGarecapableofdetectingdirectneuralelectricalcurrentsandmappingfunctionalbrainactivitiesatatemporalresolutionassmallasafractionofamillisecondandaspatialresolutionof severalmillimetres (5). Themagnetic signalsassociated with bioelectrical activity are veryweak; therefore, special techniques are neededto discriminate these signals from extraneousmagneticfields,i.e.,noise.MEGsystemsinvolveanarrayofsensorsthataredividedintotwotypes,i.e.,magnetometers andgradiometers, and eachtype is coupled toa special, low-noiseamplifier.MEG recording must be performed inside amagneticallyshieldedroomtoreduceextraneousmagnetic fields (6). The primary source ofelectromagneticsignalsisthecurrentflowintheapicaldendritesofpyramidalcellsinthecerebralcortex. Because the columnar organisation ofthe cortex is oriented normal to its surface, thecoherent activation of a small area results ina huge number of these pyramidal cells of thecortex being activated in a manner that can bemodelled as an equivalent currentdipole (ECD)(7).Thesecurrentdipolescantypicallybefoundwithin thecorticalgreymatter,andtheyare thebasicelementusedtorepresentneuralactivationinMEG-andEEG-basedinversemethods(1). ECDsand clusters of suchdipoles areusedto represent focal neural activity. ThemodelledECD information is particularly frequently usedto estimate of the locations and amplitudes oftheequivalentdipolesviatheinverseprocedure.These sources can be estimated accurately byperformingMEGrecodingwhenthetruelocationsandtemporalactivitiesofthedipolesareknown(8). To achieve this aim, artificial objects thatmimic human brain activity called “phantoms”

are constructed by the manufacturers of MEGsystems and used to evaluate the accuracies ofMEGmeasurements.Suchphantomsarenormallydelivered in the package with the MEG systemand intended for localanddailymaintenanceofthesystematthesitetoensureverificationoftheMEGmeasurements. The most common use ofsuchprovidedphantomsistopasselectriccurrentthroughelectrodeswithpre-determinedlocationsinside the phantom. The estimated locations ofthe ECDs as measured by the MEG system viainverse method calculations are then comparedtothepre-determined“trueposition”ofthesignaloriginbasedonthestructureofthephantom.Thedivergencebetweentheprojecteddipolepositionandtheoriginoftheelectrodeorthe“trueposition”isconsideredasanevaluationoftheaccuracyofthe MEG system (9). Nonetheless, this methodis questionable because the uncertainty of themeasurementmaynotliewithintheMEGsystembutwithinphantomitself(LocalPhantom). Toconfirmthereliabilityandtraceabilityoflocal phantommeasurements and the efficiencyoftheHUSMMEGsystem,areferencephantomspecificallydesignedbyElektaOywasused.Thisreference phantom possessed shape and dipolestructures that were identical to those of thelocal phantom and was used in multiple MEGsystem sites as a reference. Hence, comparisonof the data uncertainties and dispersions fromthetwophantomsisareasonablemethodfortheevaluationofthedataaccuracywhenthereferencephantom has been conditionally calibrated, andthisapproachcan thusensure confidence in theassessments(10).Therefore, the objectives of this experimentwere to compare the accuracies of the local andreference phantoms and to demonstrate thatHUSM MEG system measurements were validandfellwithintheacceptancelimits.

Materials and Methods

Phantom The system performance was examinedusing the providedphantom,which contains 32simulateddipolesandfourpre-setheadpositionindicator (HPI) coils. The hemispheric shape ofthe phantom was designed to create equivalentmagneticfielddistributionsbasedonassembledequilateral triangular linecurrents thatgeneratetangential current dipoles inside the sphere.This arrangement provisionally created suchthat the vertex of the triangle and the origin oftheconductingspherecoincide.Thecurrentsarethen fed through an inbuilt generator to excite

22 www.mjms.usm.my

Malays J Med Sci. Dec 2015; Special Issue: 20-28

theartificialdipolesviaadipolemultiplexer.Thedipolesareactivatedsequentiallyupto32timestogeneratetypicalevokedresponsesthatarethenanalysedforaccuracy. Two designated phantoms were used inthe validation procedure, i.e., a local phantom(LocalPhantom)thatwassuppliedwiththeMEGsystem and a reference NM24058N phantom(serial number: 101861; RefPhantom) providedbyElektaOy ofHelsinki Finland that had beencalibrated by the Department of MechanicalEngineering and Industrial Systems of theTampereUniversityofTechnology(Figure1).

Methods

The phantom tests were performedas instructed in the Elekta Neuromag DataAcquisitionUser’sManualNM23065A-A(11). Thephantomwasconnectedtoanexcitationmultiplexer attached to 25-pin socket connectorlocated under the side cover of the gantry.After loading the built-in measurement settingparameters located in the file /neuro/dacq/setup/phantom.fif, the acquisition programmeproceeded with the digitisation of the HPI coil(Figure2).Thesecardinalpointswereverifiedto

coincidewiththeHPIcoilbeforethephantomwasplacedintheprobeunitofthegantryhelmet.Thesetupwascarefullyarrangedsuchthatthefrontcoil pointed upward, and the 32-pair cable andHPIcoilwerefitintotherespectiveoutletsundergantry side cover (Figure 3). Later, the dipoleutilityprogrammewasexecuted,which involvedtheactivationofthefirstdipole,andthedisplayoftherawsinesignalandthetrigger.Subsequently,the average option buttonwas clicked to recordthe average of 100 epochs for a single dipole.ThecurrentversionoftheHUSMMEGsoftwareallows for the continuous recordingof a totalof32dipoleswithouttheneedtomanuallyresetthechannelforeachdipole. The measurement file was createdautomaticallywhentherecordingwascompleted.Sourcemodellingsoftwarewasusedtoanalysethedipolefitting inwhich thesetting for thesphereof origin was (0,0,0) in the head coordinatesystem, and the baseline was set from -50 to 0ms.ThelocalisationresultwascomparedwiththenominaldataprovidedbytheMEGsystem.Twomainaspectsof thecomparisonwereexamined;i.e.,theamplitudeofthedipole(Qpp)shouldbewithin the rangeof900nAmto1100nAm,andthe displacement (mm) of the dipole from thenominal location (d) should not have exceeded5mm.Themeasurementof thedipolewas thenrepeatedusingtheotherphantom(RefPhantom).Thesetwomeasurementswerethenexaminedtogaininsightintothecorrelationandvalidityofthecomparison using a simple unpaired t test afterthe normality of the data distribution has beenconfirmedvisually(Figure4)andstatistically.

Results

Thefullresultsofdipolemomentmagnitude(Qpp)anddifferenceindistancefromtheoriginofthephantomarereportedinTable1awhilethenormality test results of skewness and kurtosisaredisplayed inTable 1b.Later theZ-score canbe calculated in determining the type of datadistribution. The Z-scores of the skewness andkurtosis tests that fell within the range of -1.19to +1.96 indicated that the data were normallydistributed(12,13). Additionally, as demonstrated in Figure5, the dots that appeared along the line in theq-q plot indicated that the observed data wereapproximatelynormallydistributed. A normality test conducted with SPSSwas applied to acquire information aboutthe distribution of the data. The distributioncharacteristics were determined based on

Figure 1: Reference phantom (RefPhantom)NM24058N(Serialnumber:101861)provided by Elekta Oy, HelsinkiFinland. 32 built in simulateddipoles and four presetting headpositionindicatorcoils(HPI).

Special Issue - Neuroscience |HUSMMEGPhantomComparison&Validation

www.mjms.usm.my 23

the skewness and kurtosis test scores, whichindicated approximately normal distributions ofallofthedatasets.Basedonthephantomdipolemoment (Qpp) skewness scores of -0.122 (SE=0.414)and0.41(SE=0.414)andkurtosisscoresof-0.201(SE=0.809)and0.161(SE=0.809)forthe local phantom and the reference phantom,respectively, the values were considered smalland close to zero. To statistically quantify howfar the skewness and kurtosis departed fromtheir standard errors (SEs), the Z-values werecalculatedbydividingtheskewnessandkurtosisscoresbytherespectivestandarderrors(SE). The same approach was applied to thedifference in the dipole distance (d) from thenominallocationofthephantom.Thedifferencesinthedipoledistanceforthelocalphantom(dloc)andthereferencephantom(dref)were-0.022(SE=0.414)and-0.723(SE=0.414),respectively,inthe skewness tests, and scores of -0.220 (SE =0.809)and0.122(SE=0.809),respectively,wereobservedinthekurtosistest(Table1b).

Discussion

Allofthedataforboththelocalandreferencephantomswereslightlyskewedandkurtotic,buttheydidnotsignificantlydifferfromnormal.Wethus assumed that our datawere approximatelynormally distributed in terms of skewness andkurtosis. The Z-scores for the skewness andKurtosis tests fell in the rangeof -1.19 to+1.96,which indicated that the data were normallydistributed(12,13). Visual inspection of the histograms furtherindicatedthatallthedatasetsapproximatedtheshapeofanormalcurve.TheQ-Qplotsofthedataalso provided evidence of normal distributionsbecausetheseriesofdotindicatorsfellalongtheline. It is important to demonstrate that thedependent variable data are approximatelynormallydistributedforeachcategorysothattheappropriatetestcanbeusedforthecomparison.Unpairedorindependentparametricttestswereusedtoexploreandinvestigatetherelationshipsofthedatasets. A t test between the local phantom andreferencephantomwasanappropriateapproachtoaffirmingthatthemeasurementsofthedipolemoment(Qpp)anddipolelocalizationwerevalidbecause both phantoms had the same design,whichcontained32artificialdipoles.ALevene’stestyieldedaF(62)=1.784andP=0.187,whichindicated that the variances of both phantomswereapproximatelyequal(Table2a,b).Thus,the

Figure 2: The digitisation of the phantomincluding the head positioningindicator(HPI)coil(4point).

Figure 3: Phantom is carefully set into sensorhelmetoftheprobeunitandpushedagain helmet. HPI coil is fitted intooutletunderrightgantrysidecover.

24 www.mjms.usm.my

Malays J Med Sci. Dec 2015; Special Issue: 20-28

Figure5:(a)Thequantile-quantile(Q-Q)plotof32dipolemoment(Qpploc)forlocalphantomoverlaidwithnormaldistributionline.(b)Thequantile-quantile(Q-Q)plotof32dipoledifferenceindistance fromnominal location (dloc) for localphantomoverlaidwithnormaldistributionline. (c) The quantile-quantile (Q-Q) plot of 32 dipole moment (Qppref) for referencephantomoverlaidwithnormaldistribution line.(d)Thequantile-quantile(Q-Q)plotof32dipoledifference indistance fromnominal location (dref) for referencephantomoverlaidwithnormaldistributionline.

a b c d

Figure4:(a)Thehistogramof32dipolemoment(Qpploc)forlocalphantomwithmean978.72nAm,stddev.24.32nAmoverlaidwithnormaldistributioncurve.(b)Thehistogramof32dipoledifferenceindistancefromnominallocation(dloc)forlocalphantomwithmean1.84m,stddev.0.53moverlaidwithnormaldistributioncurve.(c)Thehistogramof32dipolemoment(Qppref)forreferencephantomwithmean988.38nAm,stddev.32.72nAmoverlaidwithnormaldistributioncurve.(d)Thehistogramof32dipoledifferenceindistancefromnominallocation(dref)forreferencephantomwithmean2.14m,stddev.0.78moverlaidwithnormaldistributioncurve.

a b c d

Figure6: (a)Theboxplot of dipolemoment (Qpp) for both local phantomand referencephantomillustratedallthedatasincludingtheoutlierarestillwithinpermittedrangeof900nAmto1100nAm.(b)Theerrorbarofdipolemoment(Qpp)forbothlocalphantomandreferencephantom illustrated the extensively overlapping error bar which means no significantdifferenceindatavalues.Allthedatavaluesarestillwithinpermitablerange.(c)Theboxplotofdipoledifferenceindistancefromnominallocation(d)forlocalandreferencephantomillustratedallthedatasincludingtheoutlierarestillunderpermittedvalue(5mm).(d)Theerrorbarcomparisonofdipoledifferenceindistancefromnominallocation(d)forlocalandreferencephantomshoedtheextensivelyoverlappingerrorbarwhichmeansnosignificantdifferenceindatavalues.Allthedatavaluesarestillbelowmaximumacceptedrange.

a b c d

Special Issue - Neuroscience |HUSMMEGPhantomComparison&Validation

www.mjms.usm.my 25

standardttestresultswereused.Theresultoftheindependent t test was not significant (t (62) =-1.34,P=0.185),whichindicatedthattherewasno significantdifferencebetween thephantoms.The 95% confidence interval of the differencebetweenthemeanswas-24.06to4.75. Similarly, the test was applied to thedifferences in the dipole distances from thephantom origins (d) for both phantoms. ALevene’s test revealed that the difference indistance(d)hadanequalvarianceasindicatedbyanF(62)=3.141andP=0.081,whichindicatedanon-significantresult(Table2c,d).Hence,thestandardttestwasusedandyieldedthescoreoft(62)=-1.818,p=0.074,whichindicatedthattherewas no significant difference in the mean (d)betweenthebothphantoms.The95%confidenceintervalofthedifferencewas-0.63to0.03. Basedonthesetestresults,itcanbeassumedthat both phantoms were similar in dipolemagnitude and dipole location characteristics.Consequently,thecomparisonsoftheaccuraciesofthedipolemagnitudeanddipolelocationwereverified. Accordingly, the test results were thencomparedtotheacceptedlimitsforthephantomtest. If the test result falls within the acceptedvalue range, the local phantom (LocalPhantom)andtheMEGsystemmeetthenationalstandardin Europe. The dipole moments (Qpps) for theLocalPhantom and RefPhantom were 978 (SD24) nAm and 988 (SD 32) nAm, respectively.AlthoughtherewasanoutlierintheRefPhantomas indicated in the dipole moment comparisonboxplot, all of the dipolesmagnitudeswere stillwithin the accepted range of 900 to 1100 nAm.The samepattern canbeobserved in thedipolemoment comparison error bar. Impressively,the LocalPhantom exhibited an accuracy andprecision that were superior to those of theRefPhantom in terms of the dipole magnitude.Thisfindingisillustratedbytheobservationthatthe dipole moment comparison standard errorof the LocalPhantomwas distinctly smaller (SE4.299)thanthatoftheRefPhantom(SE5.783). Correspondingly, the differences in thedipole distances from the origin (d) for theLocalPhantomandtheRefPhantomwere1.84(SD0.53)mmand2.14 (SD0.78)mm, respectively.Again, there was an outlier in the RefPhantomdata regarding the difference in the distance ofthe comparison boxplot. However, the outlierand the rest of thedata (d)were still below themaximumacceptablerange,whichwaswithin5.0mmofthenominaldipolelocation.Theerrorbarforthedistancefromthenominaldipole(d)was

Table 1a: The test result of 32 dipolemeasurementforbothphantoms

Dipole Qpploc(nAm)

dloc(m)

Qppref(nAm)

dref(m)

1 980 1.6 1038 2.42 983 1.3 1020 2.13 965 1.3 998 1.84 974 1.2 1069 3.15 997 2.4 1002 2.66 956 1.9 993 2.57 1001 1.9 1011 2.18 1018 2.1 1035 3.19 946 2.2 966 2.410 966 1.7 950 2.811 962 1.3 1027 3.512 978 0.7 941 1.913 989 2.1 995 2.714 972 1.9 981 1.815 969 2.1 988 1.616 963 1.8 992 1.317 999 1.9 985 218 957 1.6 960 2.219 951 1.4 931 2.320 928 2.2 922 2.921 989 2.5 987 2.522 982 2.6 986 2.423 981 1.7 972 2.924 980 2.1 957 2.725 1016 2.6 1024 2.226 1007 1.9 982 0.927 963 1.1 968 0.528 991 1.6 971 0.629 1017 3 1032 330 1024 2.6 1001 1.931 988 1.7 970 1.532 927 0.9 974 0.4Average 978.72 1.84 988.38 2.14Thetestresultof32dipolemomentmagnitude(nAm)forlocalphantom(Qpploc)andreferencephantom(Qppref)aswellastheresultfordifferenceindistancefromorigin(m)oflocalphantom(dloc)andreferencephantom(dref).

26 www.mjms.usm.my

Malays J Med Sci. Dec 2015; Special Issue: 20-28

Table1b:Thenormalitytestresultofthephantoms N Minimum Maximum Mean SD Skewness Kurtosis

Statistic Statistic Statistic Statistic Statistic Statistic Std.

Error

Z-

score

Statistic Std.

Error

Z-score

Qpploc 32 927 1024 978.72 24.321 -0.122 0.414 -0.295 -0.163 0.809 -.201

dloc 32 0.7 3.0 1.841 0.5333 -0.022 0.414 -0.053 -0.220 0.809 -.271

Qppref 32 922 1069 988.38 32.715 0.241 0.414 0.582 0.161 0.809 .199

dref 32 0.4 3.5 2.144 0.7779 -0.723 0.414 -1.743 0.122 0.809 .151

Thenormalitytestresultofrespectivedipolemomentmagnitude(Qpp)(nAm)andthedifferenceindistancefromorigin(d)(m)forlocalphantom(loc)andreferencephantom(ref).

Table2a:Thegeneralstatistictestresultofdipolemomentmagnitude(Qpp)Phantom N Mean SD SEM

Qpp LocalPhantom 32 978.7188 24.32075 4.29934RefPhantom 32 988.3750 32.71455 5.78317

Thestatistictestresultofdipolemomentmagnitude(Qpp)(nAm)forlocalphantom(loc)andreferencephantom(ref).

Table2b:Thettestresultofdipolemomentmagnitude(Qpp)(nAm)ofthephantomsIndependentSamplesTest

Levene’sTestforEqualityofVariances

t-testforEqualityofMeans 95%ConfidenceIntervaloftheDifference

F Sig. t df Sig.(2-tailed)

MeanDifference

Std.ErrorDifference

Lower Upper

Qpp Equalvariancesassumed

1.784 0.187 -1.340 62 0.185 -9.65625 7.20621 -24.06125 4.74875

Equalvariancesnotassumed

-1.340 57.248 0.186 -9.65625 7.20621 -24.08507 4.77257

Thettestresultofdipolemomentmagnitude(Qpp)(nAm)forlocalphantom(loc)andreferencephantom(ref)demonstratedthatLevene’ssignificantscore0.187whichishigherthanP=0.05.Thismeansthatbothdipolephantomdataishavingtheequalvariance.

Table2c:Thegeneralstatistictestresultofdipoledifferenceindistancefromorigin(d)(m)ofthephantoms

GroupStatistics

Phantom N Mean Std.Deviation Std.ErrorMean

d LocalPhantom 32 1.8406 0.53332 0.09428RefPhantom 32 2.1438 0.77790 0.13751

Generalstatistictestresultofdipoledifferenceindistancefromorigin(d)(m)forlocalphantom(loc)andreferencephantom(ref).

Special Issue - Neuroscience |HUSMMEGPhantomComparison&Validation

www.mjms.usm.my 27

Table2d:ttestresultofdipoledifferenceindistancefromorigin(d)(m)ofthephantoms.IndependentSamplesTest

Levene’sTestforEqualityofVariances

t-testforEqualityofMeans 95%ConfidenceIntervaloftheDifference

F Sig. t df Sig.(2-tailed)

MeanDifference

Std.ErrorDifference

Lower Upper

d Equalvariancesassumed

3.141 0.081 -1.818 62 0.074 -0.30312 0.16673 -0.63641 0.03016

Equalvariancesnotassumed

-1.818 54.869 0.075 -0.30312 0.16673 -0.63727 0.03102

The t test result of dipole difference in distance from origin (d) (m)for local phantom (loc) and reference phantom (ref)demonstratedthatLevene’ssignificantscore0.081whichishigherthanP=0.05.Thismeansthatbothdipolephantomdataishavingtheequalvariance.

narrowerfortheLocalPhantom(SE0.094)thantheRefPhantom(SE0.138),whichagainindicatedthattheLocalPhantomperformedbetterthantheRefPhantom.

Conclusion

The local phantom measurements werevalidated by demonstrating that the collecteddata were comparable to those of the referencephantom in terms of the statistically equivalentmeans and variances; thus, the HUSM MEGsystem measurements were validated as beingwithintheacceptablelimitsofthephantomtest.Therefore, this system was assumed to be inexcellent working condition in terms of dipolemagnitudeanddipolelocalisationmeasurements.The measurements from this HUSM MEGsystemwereacceptableaccordingtothenationalstandardinEurope.

Acknowledgement

The authors would like to thank JukkaKnuutilaoftheElektaOy,Helsinki,forassistancewith the designing and calibration of thereferencephantom.ThisresearchwassupportedbyHospitalUniversitiSainsMalaysiainensuringtheoptimaldiagnosticservicestothepatientsandresearch.

Funds

None.

Conflicts of Interests

None.

Authors’ Contributions

Conceptionanddesign:HO,ALA,ZI,JMAAnalysisandinterpretationofthedata:HO,ALA,NHDraftingofthearticle:HO,ALACritical revision of the article for importantintellectualcontent:ALA,ZI,JMAFinalapprovalofthearticleZI,JMAProvisionofstudymaterialsorpatients:NHStatisticalexpertise,administrative,technical,orlogisticsupport,collectionandassemblyofdata:HO,NH

Correspondence

MrHazimOmarBASc Medical Physics (USM), MSc TranslationalNeuroimaging(UniversityofNottingham)DepartmentofNeurosciencesHospitalUniversitiSainsMalaysiaUniversitiSainsMalaysia16150KubangKerianKelantan,MalaysiaTel:+609-7674046Fax:+609-7673833Email:hazim@usm.my

28 www.mjms.usm.my

Malays J Med Sci. Dec 2015; Special Issue: 20-28

References

1. DarvasF,PantazisD,YildirimE,LeahyRM.MappinghumanbrainfunctionwithMEGandEEG:methodsand validation. Neuroimage. 2004;23 (Suppl 1):S289–299.doi:10.1016/j.neuroimage.2004.07.014.

2. FernandoLopesdasilva.EEGandMEG:RelevancetoNeuroscience. Neuron. 2013;80(5):1112–1128. doi:0.1016/j.neuron.2013.10.017.

3. HansenPC,KringelbachML,SalmelinR. MEG : an introduction to methods. New York (NY): OxfordUniversityPress.xii.2010.p.436.

4. Hämäläinen M, Hari R, Ilmoniemi RJ, KnuutilaJ, Lounasmaa OV. Magnetoencephalography-theory, instrumentation, and applications tononinvasive studies of the working human brain.Rev Mod Phys. 1993;65(2):413–497. doi: 10.1103/RevModPhys.65.413.

5. Sung CJ.MEG and EEG fusion in Bayesian frame.Electronics and Information Engineering (ICEIE).2010;2:295–299.doi:10.1109/ICEIE.2010.5559785.

6. Papanicolaou A, Eduardo MC, Billingsley-MashallR, Pataraia E. Simos PG. A review of clinicalapplications of magnetoencephalography. Int Rev Neurobiol. 2005;68:223–247. doi: 10.1016/S0074-7742(05)68009-9.

7. OkadaYC,WuJ,KyuhouS.GenesisofMEGsignalsinamammalianCNSstructure.Electroencephalogr.Clin Neurophysiol. 1997;103(4):474–485. doi: 10.1016/S0013-4694(97)00043-6.

8. Leahy RM, Mosher JC, Spencer M, HuangMX, Lewine JD. A study of dipole localizationaccuracy for MEG and EEG using a human skullphantom. Electroencephalogr. Clin Neurophysiol. 1998;107(2):159–173.doi: 10.1016/S0013-4694(98)00057-1.

9. OyamaD.Dryphantomformagnetoencephalography-Configuration, calibration, and contribution. J Neurosci Methods. 2015;251:24–36. doi: 0.1016/j.jneumeth.2015.05.004.

10. Metrology BJCfGi. Evaluation of measurement data – Guide to the expression of uncertainty in measurement JCGM 100:2008, in (GUM 1995 with minor corrections).Paris;2008:JCGM100:2008:

11. OyEN.Magnetoencephalography Data Acquisition User's Manual.Helsinki,Finland;Elekta:2007.

12. Kim HY. Statistical notes for clinical researchers:assessing normal distribution (2) using skewnessandkurtosis. Restor Dent Endod.2013;38(1):52–54.doi:10.5395/rde.2013.38.1.52.

13. CramerD.Fundamental statistics for social research: step-by-step calculations and computer techniques using SPSS for Windows. London [England], NewYork;Routledge.1998.p.456.