guidelines eeg

Review article

Guidelines for the use of EEG methodologyin the diagnosis of epilepsyInternational League Against Epilepsy: Commission ReportCommission on European Affairs: Subcommission onEuropean Guidelines

Introduction

The primary diagnosis of epilepsy is clinical butelectroencephalography (EEG) plays a major rolein evaluating epilepsy, recognizing that a normalroutine EEG does not exclude the diagnose ofepilepsy. The main indication for performingan EEG is clinical suspicion of an epilepticdisorder. EEGs can also be useful in the evalu-ation of encephalopaties (metabolic, infectious,

degenerative) and focal brain lesions (cerebralinfarction, haemorrhage, neoplasms). In paediat-ric practice, the EEG might help to determine thelevel of maturation of the brain. The EEG is notuseful in following the therapeutic effect ofantiepileptic drugs (AEDs) as interictal epilepti-form activity is affected very little by AEDs. Anexception is absence epilepsy where the quantifi-cation of spike-wave episodes is helpful infollowing the effect of treatment.

Flink R, Pedersen B, Guekht AB, Malmgren K, Michelucci R, NevilleB, Pinto F, Stephani U, Ozkara C. Guidelines for the use of EEGmethodology in the diagnosis of epilepsy. International League AgainstEpilepsy: Commission report. Commission on European Affairs:Subcommission on European Guidelines.Acta Neurol Scand 2002: 106: 1–7. � Blackwell Munksgaard 2002.

The Commission of European Affairs of the International LeagueAgainst Epilepsy published ‘Appropriate Standards for Epilepsy CareAcross Europe’ which contained recommendations for the use ofelectroencephalography (EEG) in the diagnosis of epilepsy (Brodieet al. Epilepsia 1997; 38:1245). The need for a more specific basicdocument of EEG methodology was recognized and theSubcommission on European Affairs was asked to produce moredetailed guidelines to be used across Europe recognizing the range ofpractices in EEG laboratories.There are many general guidelines published on EEG methodology butthis document focuses on the diagnosis of epilepsy. Details frompreviously published guidelines are included in references and in anappendix.These guidelines are not meant to be used as minimal standards butrecommendations that can be applied to all EEG laboratories despitevariations in equipment.

R. Flink1, B. Pedersen2,A. B. Guekht3, K. Malmgren4,R. Michelucci5, B. Neville6,F. Pinto7, U. Stephani8, C. *zkara91Department of Neuroscience, Clinical Neurophysiology,University Hospital, Uppsala, Sweden; 2NeurologiskAvdelning, Aalborg sygehus Nord, Denmark;3Department of Neurology and Neurosurgery, RussianState Medical University, Moscow, Russia; 4Departmentof Neurology, Institute of Clinical Neuroscience,Sahlgrenska University Hospital, G(teborg, Sweden;5Divisione di Neurologia, Ospedale Bellaria, Bologna,Italy; 6Neuroscience Unit, Institute of Child Health, TheWolfson Centre, London, UK; 7Consulta de Epilepsia,Hospital Santa Maria, Lisboa, Portugal; 8Klinik f5rNeurop6diatrie der Universit6t, Kiel, Germany;9Cerrahpasa Tip Fak5ltesi, Norosirurji Anabilim Dali,Istanbul, Turkey

Key words: Commission Report; European Guidelines;electroencephalography; diagnosis of epilepsy;methodology; EEG reporting

Roland Flink, Department of Neuroscience, ClinicalNeurophysiology, University Hospital, SE-751 85Uppsala, SwedenTel.: +46 18 66 3428Fax: +46 18 55 6106e-mail: [email protected]

Accepted for publication December 6, 2001

Acta Neurol Scand 2002: 106: 1–7Printed in UK. All rights reserved

Copyright � Blackwell Munksgaard 2002

ACTA NEUROLOGICASCANDINAVICAISSN 0001-6314

1

The purpose of the EEG recording is to detectinterictal activity and localize the region of inte-rictal activity and/or ictal activity or ictal events. Inthe presence of epileptiform activity, the EEGrecording will also help to determine the type ofseizure or epilepsy syndrome.

A routine EEG recording in a patient withepilepsy will have no epileptiform activity inabout 50% of cases. It is necessary therefore toincrease the sensitivity by activation proceduressuch as hyperventilation, photic stimulation, sleepand sleep deprivation. The EEG recording duringsleep is particularly useful when there is asuspicion of epilepsy with partial seizures or thesyndrome of benign childhood epilepsy withcentrotemporal spikes. Hypsarrhythmia in West’ssyndrome is an EEG pattern that changes fromwakefulness to sleep. Using activation procedures,abnormalities are found in about 90% of patientswith epilepsy (1). If routine EEG and EEG aftersleep deprivation still reveals no abnormalities,long-term EEG monitoring may be used, whichincreases the detection rate of interictal and/orictal events.

Routine EEGs for specific differential diagnosticpurposes should include simultaneous electrocar-diography (ECG) monitoring and where appropri-ate a monitoring of other parameters, i.e.respiration and muscle activity (polygraphicrecording).

Equipment and electrodes – electrode position

The ‘modified combined nomenclature’ derivedfrom the 10–20 system should be used for electrodelocation and the minimum number of electrodesshould be 21 regardless of the number of thechannels available on the EEG (2).

A 16-channel recording is recommended forfocal (partial) epilepsies; however, for diagnosingprimary generalized epilepsy the number of chan-nels is not crucial. Even with an eight-channelrecording it is possible to obtain and locate theepileptiform discharges with the modified com-bined system but a longer sampling time usingseveral montages will be required (3).

Although there is no general agreement amongEEG laboratories about montages, a routine EEGshould (at least) include bipolar montages withlongitudinal and transverse chains. These chainsshould be used with equal electrode distances andside-to-side symmetry to avoid the artefact offalse amplitude asymmetry. Additional referentialmontages should be included in a routine record-ing. The previously mentioned montages areavailable in different settings in the international

recommendations and can be used with differenttypes of equipment from eight-channel EEGmachines and those with more channels. Asmany EEG recordings are still analog, it isuseful to have similar montages and examples ofthe common ones are given in the appendix (seeAppendix).

Recording parameters such as sensitivity, filtersetting (notch filter) and time base should be inaccordance with the international guidelines (4).The use of muscle filters and notch filter should beavoided if possible because of the risk of losing lowamplitude spike potentials. The normal paperspeed is 15 or 30 mm/s. Reduced paper speedallows better detection of slow wave abnormalities.The duration of recording time should be at least30 min of artefact-free signals.

In digital recordings, which allow remontaging,recording time should not be reduced, because thechances of recording interictal epileptiform activityincreases with sampling time. The specific inter-national guidelines for recording clinical EEG ondigital media should be used (5).

Requirements for paediatric EEG recordings

Preterm

Although the head may be small, the minimumnumber of electrodes is nine. The montages per-formed should be bipolar, longitudinal and trans-verse including the Cz electrode. These recordingsare often performed in surroundings with a lot ofelectrical interference, giving artefacts. Such arte-facts can be reduced by using short electrode cablesthus placing the headbox and preamplifier close tothe head of the child. If possible a polygraphicrecording should be obtained including ECG,respiration, eye movements and electromyography(EMG) to record muscle activity (6).

The duration of the recording in small childrenshould be at least 1 h to be able to evaluatewakefulness and sleep and reactivity. Photic sti-mulation in these children may give additionalinformation and should be performed (7). Activa-tion processes in order to evaluate EEG reactivitycould include tapping and auditorial stimuli.

Infants and older children

The adult electrode placement should be used asyoung as possible as more electrodes will enhancesensitivity. A sleep recording will increase thechance of recording epileptiform activity andusually reduce movement and muscle artefacts. Inall children despite age, recordings in spontaneoussleep are preferred to induced sleep.

Flink et al.

2

Recording procedure

Standard EEG recording should attempt to includeeye-opening, -closure and -blink procedures toexclude artefacts derived from eye movements.Voluntary hand movements can be useful fortesting the reactivity of central rhythms.

A standard EEG should aim to include thefollowing activation procedures:

(i) Hyperventilation depends upon the develop-mental age, and the level of co-operation of thechild and the experience of the staff.(ii) Hyperventilation for 3 min with a continuedrecording for at least 2 min after cessation of hy-perventilation, which may be worth repeating ifabsence epilepsy is suspected and the first recordingwas negative. If there is unexpected abnormal hy-persynchronization activity during hyperventila-tion, the possibility of physiological hypoglycaemiashould be excluded.(iii) Photic stimulation should be performed withseparate trains of photo flashes of 10 s duration foreach frequency and with minimum intervals of 7 s.Each 10 s train consists of 5 s with open eyes fol-lowed by closure and 5 s with closed eyes. Thedistance between the lamp and the nose should be30 cm. The method described below follows theprevious recommendation for screening for andidentifying photosensitive subjects (8, 9).

Photic stimulation should not be performed duringor within 3 min of hyperventilation. Photic stimu-lation should start at the frequency of 1 Hz andprogress to 20 Hz unless generalized epileptiformdischarges are evoked. This is immediately fol-lowed by a sequence with 60 Hz photic stimulationand then decreasing to 25 Hz. The trains should beperformed with the following frequencies: 1, 2, 4, 6,8, 10, 12, 14, 16, 18, 20, 60, 50, 40, 30 and 25 Hzand in that order. The total duration of photicactivation time recommended is at maximum6 min in patients without any reaction to intermit-tent photic stimulation. The procedure should bestopped at once if the patient has a warning or ifepileptiform activity is evoked in the EEG.

Not all EEG machines and photic stimulatorscan perform flashes up to 60 Hz, but testing usingthe single train with open eyes, eye closure andclosed eyes should use the available frequencies,especially in children. A reduced program of 1, 4, 6,10, 14, 16 and 20 Hz is used in several laboratories;however, one must be aware of the risk of falsenegative results.

Sleep recordings can increase the accuracy ofthe diagnoses of epilepsy and should be per-formed when the standard EEG fails to show any

epileptiform activity and the level of clinicalsuspicion justifies this investigation. Sleep depriva-tion can be a useful activation procedure. It mustbe recognized that sleep deprivation may provokeclinical seizures. If natural sleep is not achieved, anumber of pharmacological agents can be used toinduce sleep. One must be aware of pharmacolog-ically induced EEG changes that occasionallymay mimic epileptiform activity (interference ofb activity).

Adult patients should abstain from sleep thenight prior to the recording. In children up to12 years the sleeping time should be reduced asmuch as possible before the recording. A procedureused in many laboratories includes partial nightsleep deprivation and EEG recording during thepostprandial time when a nap is most likely tooccur. The sleep recording should include wakeful-ness, drowsiness and at least 40 min of sleep. Thedepth of sleep should be at least to stages 2 and 3.

The montage must include a bipolar montagewith transverse chains which allow the identifica-tion of vertex waves – additional channels shouldbe used for electrooculography (EOG), surfaceEMG recorded from submental muscles, ECG andspirogram (polysomnography) in order to identifysleep stages and to distinguish between frontaldischarges and eye movements. Polysomnographyshould always be used for all-night recordings inorder to record epileptic nocturnal activity and todifferentiate this activity from non-epileptic events.

Long-term EEG recording

Long-term monitoring refers to EEG recordingover an extended period. There are several optionsin performing long-term monitoring (10).

In cases where a standard EEG followed byactivation procedures does not show clear evidenceof interictal epileptiform activity a prolongedrecording time increases the possibility of detectionof interictal and ictal events. The duration of suchrecordings can, depending on the clinical questionraised, be from few hours to several days. Thestandard 10–20 system in its modified version isrecommended for these recordings.

The long-term monitoring can be performedwith portable equipment. This type of dynamicEEG recording allows the patient freedom toperform routine activities as an out-patient or in-patient during long-term EEG monitoring. Eight-or 16-channel recordings are available. These EEGrecordings may be useful to quantify seizureactivity in patients with generalized discharges orfor supervised monitoring in certain situations,that cannot be reproduced in the EEG laboratory

Use of EEG for the diagnosis of epilepsy

3

or hospital. The disadvantages include the in-creased risk of technical problems and lack ofaccurate simultaneous behavioural correlation.

This type of monitoring may preferably beperformed as sleep recordings because the artefactsare less during sleep.

In order to correlate clinical behaviour and EEGfindings the recording requires video-equipmentwith synchronized time code, i.e. split screen videoEEG. Digital video EEG systems are currentlyavailable. Long-term video-EEG monitoring maybe the only way to distinguish epileptic from non-epileptic seizures and is mandatory as part ofpresurgical evaluation. It is of great value if atrained technician or nurse is constantly presentduring a video-EEG recording to test patientresponsiveness, muscle tone, etc., whenever suspectictal episodes occur. A set of testing proceduresincluding language tests should be available to theexaminer, to be administered during suspected ictalepisodes.

EEG reporting

The EEG report should be a response to theclinical question asked by the referring doctor.Apart from describing the waveforms and frequen-cies of the EEG signal there should be a clinicalinterpretation.

A standardized report form is advocated andshould include the following headings:

(i) Information concerning the status of thepatient, neurological condition, medication, lastseizure, and clinical question.(ii) Information concerning the EEG recording;number of electrodes, use of special electrodes,recording conditions, level of consciousness of thepatient during the recording, activation proce-dures, artefacts noted.(iii) Description of EEG; postcentral rhythm,background activity, asymmetries, epileptiformactivity, specific EEG patterns, effect of activationprocedures.(iv) Clinical interpretation; this puts the EEGfindings in a clinical context (the clinical signifi-cance of the EEG findings, prognosis, etc.), andresponds to the clinical question.

It is strongly advised to use the glossary pro-posed by the International Federation of ClinicalNeurophysiology (11).

AcknowledgementsThe authors wish to thank Guiliano Avanzini and MartinBrodie for valuable comments on the manuscript, and

EUCARE for providing the subcommission the possibility ofa workshop.

References1. BINNIE CD, STEFAN H. Modern electroencephalography:

its role in epilepsy management. Clin Neurophysiol1999;110:1671–97.

2. KLEM GH, LUDERS HO, JASPER HH, ELGAR C. The ten-twenty electrode system of the International Federation.Recommendations for the Practice of Clinical Neurophy-siology: Guidelines of the International Federation ofClinical Physiology. Electroenceph Clin Neurophysiol1999;52(Suppl.):3–6.

3. AMERICAN ELECTROENCEPHALOGRAPHIC SOCIETY. Guide-line Seven: a proposal for standard montages to be used inclinical EEG. J Clin Neurophysiol 1994;11:30–6.

4. EBNER A, SCIARNETTA G, EPSTEIN C, NUWER M. EEGinstrumentation. Recommendations for the Practice ofClinical Neurophysiology: Guidelines of the InternationalFederation of Clinical Physiology. Electroenceph ClinNeurophysiol 1999;52(Suppl.):7–10.

5. NUWER MR, CORNI G, EMERSON R et al. IFCN stand-ards for digital recording of clinical EEG. Recommen-dations for the Practice of Clinical Neurophysiology:Guidelines of the International Federation of ClinicalPhysiology. Electroenceph Clin Neurophysiol 1999;52(Suppl.):11–4.

6. AMERICAN ELECTROENCEPHALOGRAPHIC SOCIETY. Guide-line Two: minimum technical standards for pediatricelectroencephalography. J Clin Neurophysiol 1994;11:6–9.

7. DE WEERD AW, DESPLAND PA, PLOUIN P. Neonatal EEG.Recommendations for the Practice of Clinical Neurophy-siology. Guidelines of the International Federation ofClinical Physiology. Electroenceph Clin Neurophysiol1999;52(Suppl.):149–57.

8. KASTELEIJN-NOLST TRENITE DGA, BINNIE CD, HARDING

GFA, WILKINS A. Photic stimulation: standardization ofscreening methods. Epilepsia 1999;40(Suppl. 4):75–9.

9. KASTELEIJN-NOLST TRENITE DGA, BINNIE CD, HARDING

GFA et al. Medical technology assessment photic stimu-lation. Standardization of screening methods. Neuro-physiol Clin 1999;29:318–24.

10. AMERICAN ELECTROENCEPHALOGRAPHIC SOCIETY. Guide-line Twelve: guidelines for long-term monitoring for epi-lepsy. J Clin Neurophysiol 1994;11:88–110.

11. NOACHTAR S, BINNIE C, EBERSOLE J, MAUGERIERE F,SAKAMOTO A, WESTMORELAND B. A glossary of terms mostcommonly used by clinical electroencephalographers andproposal for the report form for the EEG findings. Rec-ommendations for the Practice of Clinical Neurophysiol-ogy. Guidelines of the International Federation of ClinicalPhysiology. Electroenceph Clin Neurophysiol 1999;52(Suppl.):21–41.

Appendix

Recording montages

Basic rules for choosing montages:

(i) Use the full 21 electrode placements of the 10–20 System, even if only an eight-channel recordingmachine is available.

Flink et al.

4

(ii) Use longitudinal bipolar, transversal bipolarand referential montages during each EEGrecording.(iii) Duration of recording with one montageshould be at least 2 min.(iv) The electrode connections (bipolar) shouldrun in straight (unbroken) lines and the interelec-trode distance should be kept equal.(v) Use a ‘left above right’ order of deriva-tions, i.e. on the recording page left-sided leadsshould be placed above right-sided leads for eitheralternating pairs of derivations or blocks of deri-vations.(vi) Tracings from the more anterior electrodesshould be placed above those from the more pos-terior electrodes on the recording page.(vii) During photic stimulation, frontopolar,frontal and occipital regions should be explored.

(viii) During hyperventilation, frontal, central,occipital and middle temporal regions should beexplored.

Pediatric montages

Because of the size of head, a reduced number ofelectrodes is usually used in neonates and younginfants. A minimum of nine electrodes should beused (including Fp1, Fp2, C3, Cz, C4, T3, T4, O1and O2). An eight-channel EEG machine is usuallynot enough as most of these recordings will requiretwo or even more channels devoted to recordpolygraphic variables such as ECG and respir-ation.

Montage using nine electrodes, combininglongitudinal and transverse derivations. The lon-gitudinal derivation block can be used with an

Proposed montages for eight- and 16-channel EEG machinesEight-channel recordings

Bipolar montages

Channel Longitudinal Transverse Referential montages

1 Fp1-F3 Fp1-F7 F7-Fp1 F7-F3 T3-C3 F3-A1 Fp1-A12 F3-C3 F7-T3 Fp1-Fp2 F3-Fz C3-Cz C3-A1 F7-A13 C3-P3 T3-T5 Fp2-F8 Fz-F4 Cz-C4 P3-A1 T3-A14 P3-O1 T5-O1 C3-Cz F4-F8 C4-T4 O1-A1 T5-A15 Fp2-F4 Fp2-F8 Cz-C4 T3-C3 T5-P3 F4-A2 Fp2-A26 F4-C4 F8-T4 T5-O1 C3-Cz P3-Pz C4-A2 F8-A27 C4-P4 T4-T6 O1-O2 Cz-C4 Pz-P4 P4-A2 T4-A28 P4-O2 T6-O2 O2-T6 C4-T4 P4-T6 O2-A2 T6-A2

Sixteen-channel recordings

Bipolar montages

Channel Longitudinal Transverse Referential montages

1 Fp1-F3 F7-Fp1 Fp1-Fp2 F7-Fp1 Fp1-A1 Fp1-A12 F3-C3 Fp1-Fp2 F7-F3 Fp2-F8 F3-A1 Fp2-A23 C3-P3 Fp2-F8 F3-Fz F7-F3 C3-A1 F3-A14 P3-O1 F7-F3 Fz-f4 F3-Fz P3-A1 F4-A25 Fp2-F4 F3-Fz F4-F8 Fz-F4 Fp2-A2 C3-A16 F4-C4 Fz-F4 A1-T3 F4-F8 F4-A2 C4-A27 C4-P4 F4-F8 T3-C3 T3-C3 C4-A2 P3-A18 P4-O2 T3-C3 C3-Cz C3-Cz P4-A2 P4-A29 Fp1-F7 C3-Cz Cz-C4 Cz-C4 F7-A1 F7-A110 F7-T3 Cz-C4 C4-T4 C4-T4 T3-A1 F8-A211 T3-T5 C4-T4 T4-A2 T5-P3 T5-A1 T3-A112 T5-O1 T5-P3 T5-P3 P3-Pz O1-A1 T4-A213 Fp2-F8 P3-Pz P3-Pz PZ-P4 F8-A2 T5-A114 F8-T4 Pz-P4 Pz-P4 P4-T6 T4-A2 T6-A215 T4-T6 P4-T6 P4-T6 T5-O1 T6-A2 O1-A116 T6-O2 O1-O2 O1-O2 O2-T6 O2-A2 O2-A2


5

eight-channel EEG machine, and when switchingto the transverse block of derivations (channels9–12) there will be possibilities for polygraphicchannels even with the eight-channel machine.

Recording parameters

ElectrodesThe impedance (contact resistance) in the elec-trodes should be less than 5 kW, to reduce the noiseartefacts and other interference.

AmplificationThe EEG machine is equipped with differentialamplifiers with common mode rejection. Atypical setting for the gain for the EEG is7 lV/mm leading to deflections of 3–20 mm forinput voltages of 20–140 lV. The ability of theamplifiers to suppress voltages common to bothelectrodes is called the common mode rejection,it will for example reduce the noise from 50 or60 Hz line current. The ability of an amplifierto reject in phase and amplify out of phasepotentials defines the common mode rejectionration of the amplifier, which is always>80 dB.

Channel Montage

1 Fp1-T32 T3-O13 Fp2-T44 T4-O25 Fp1-C36 C3-O17 Fp2-C48 C4-O29 T3-C310 C3-Cz11 Cz-C412 C4-T4ECGRespirationEOGOptional

Recording procedure

Awake, test reactivity, intermittent eye-opening Drowsy/sleep Hyper-ventilation Photic stimulation

10 min 10 min 3 + 3 min 6 min (maximum)

Photic stimulation

Time 5 s 5 s 7 s 5 s 5 s 7 s 5 s 5 s 7 s 5 s 5 s 7 s

Flash Rest Rest Rest RestFrequency 1 Hz 2 Hz 4 Hz 6 HzEye open open open open openEye closed closed closed closed closed

Continued

Time 5 s 5 s 7 s 5 s 5 s 7 s 5 s 5 s 7 s 5 s 5 s 7 sFlash Rest Rest Rest RestFrequency 8 Hz 10 Hz 12 Hz 14 HzEye open open open open openEye closed closed closed closed closed

Continued


Continued


Flink et al.

6

PolarityThe polarity convention in recording with differ-ential amplifiers is the following:

(i) If input 1 is negative with respect to input 2,there is an upward deflection.(ii) If input 1 is positive, there is a downwarddeflection.(iii) If input 2 is negative with respect to input 1,there is a downward deflection.(iv) If input 2 is positive, there is an upwarddeflection.

FiltersettingFor the majority of EEG recordings the signalfrequency lies between 1 and 70 Hz and thebandwidth of the recording channels should there-fore correspond to this frequency range. If anarrower bandwidth is used information will belost and with a wider bandwidth, noise in therecorded data will contain irrelevant information.The low frequency filter should not be higher than1 Hz (corresponding time constant of 0.16 s) andthe high frequency filter should be 70 Hz. The useof notch filters (50 or 60 Hz) can distort sharpsignals (i.e. spikes and sharp waves) and shouldonly be used if other measures to reduce 50 or60 Hz noise fail.

The use of additional channels recording EMG,ECG, EOG, respiration, movements, etc., requiresindividual setting of gain and filters for eachchannel.

Paper speedThe print out of the recording (whether analog ordigital) should use a time base (paper speed) of 15or 30 mm/s with the option of 60 mm/s. The lower

paper speed is preferably used in order to detectslow-wave abnormalities. The use of the higherpaper speed might help in detecting asynchronicityin bilateral discharges.

Report form

Information by referring physicianPatient IDClinical historyNeurological conditionMedicationClinical question

Information by EEG technicianLevel of consciousness, vigilance, co-operation ofthe patientClinical symptoms (jerking, moving, etc.)Activation proceduresTechnical artefacts

EEG descriptionBackground activityPostcentral rhythmAsymmetriesFocal findingsEpileptiform activity, type, appearance and loca-tionSpecial EEG patternsEffect of activation procedures

EEG interpretation (keep it short)Interpretation of the EEG results in light of theclinical diagnosis and questions of the referringphysician (clinical significance, prognosis, etc.).


7

guidelines eeg

Documents