Acta Neurol Scand 2002: 106: 17 Printed in UK.

All rights reserved

Copyright Blackwell Munksgaard 2002

ACTA NEUROLOGICA SCANDINAVICA ISSN 0001-6314

Review article

Guidelines for the use of EEG methodology in the diagnosis of epilepsy

International League Against Epilepsy: Commission Report Commission on European Aairs: Subcommission on European Guidelines
Flink R, Pedersen B, Guekht AB, Malmgren K, Michelucci R, Neville zkara C. Guidelines for the use of EEG B, Pinto F, Stephani U, O methodology in the diagnosis of epilepsy. International League Against Epilepsy: Commission report. Commission on European Aairs: Subcommission on European Guidelines. Acta Neurol Scand 2002: 106: 17. Blackwell Munksgaard 2002. The Commission of European Aairs of the International League Against Epilepsy published Appropriate Standards for Epilepsy Care Across Europe which contained recommendations for the use of electroencephalography (EEG) in the diagnosis of epilepsy (Brodie et al. Epilepsia 1997; 38:1245). The need for a more specic basic document of EEG methodology was recognized and the Subcommission on European Aairs was asked to produce more detailed guidelines to be used across Europe recognizing the range of practices in EEG laboratories. There are many general guidelines published on EEG methodology but this document focuses on the diagnosis of epilepsy. Details from previously published guidelines are included in references and in an appendix. These guidelines are not meant to be used as minimal standards but recommendations that can be applied to all EEG laboratories despite variations in equipment.

R. Flink1, B. Pedersen2, A. B. Guekht3, K. Malmgren4, R. Michelucci5, B. Neville6, F. Pinto7, U. Stephani8, C. zkara9

1 Department of Neuroscience, Clinical Neurophysiology, University Hospital, Uppsala, Sweden; 2Neurologisk Avdelning, Aalborg sygehus Nord, Denmark; 3 Department of Neurology and Neurosurgery, Russian State Medical University, Moscow, Russia; 4Department of Neurology, Institute of Clinical Neuroscience, Sahlgrenska University Hospital, Gteborg, Sweden; 5 Divisione di Neurologia, Ospedale Bellaria, Bologna, Italy; 6Neuroscience Unit, Institute of Child Health, The Wolfson Centre, London, UK; 7Consulta de Epilepsia, Hospital Santa Maria, Lisboa, Portugal; 8Klinik fr Neuropdiatrie der Universitt, Kiel, Germany; 9 Cerrahpasa Tip Fakltesi, Norosirurji Anabilim Dali, Istanbul, Turkey

Key words: Commission Report; European Guidelines; electroencephalography; diagnosis of epilepsy; methodology; EEG reporting Roland Flink, Department of Neuroscience, Clinical Neurophysiology, University Hospital, SE-751 85 Uppsala, Sweden Tel.: +46 18 66 3428 Fax: +46 18 55 6106 e-mail: roland.flink@nc.uas.lul.se Accepted for publication December 6, 2001

Introduction

The primary diagnosis of epilepsy is clinical but electroencephalography (EEG) plays a major role in evaluating epilepsy, recognizing that a normal routine EEG does not exclude the diagnose of epilepsy. The main indication for performing an EEG is clinical suspicion of an epileptic disorder. EEGs can also be useful in the evaluation of encephalopaties (metabolic, infectious,

degenerative) and focal brain lesions (cerebral infarction, haemorrhage, neoplasms). In paediatric practice, the EEG might help to determine the level of maturation of the brain. The EEG is not useful in following the therapeutic eect of antiepileptic drugs (AEDs) as interictal epileptiform activity is aected very little by AEDs. An exception is absence epilepsy where the quantication of spike-wave episodes is helpful in following the eect of treatment. 1

Flink et al. The purpose of the EEG recording is to detect interictal activity and localize the region of interictal activity and/or ictal activity or ictal events. In the presence of epileptiform activity, the EEG recording will also help to determine the type of seizure or epilepsy syndrome. A routine EEG recording in a patient with epilepsy will have no epileptiform activity in about 50% of cases. It is necessary therefore to increase the sensitivity by activation procedures such as hyperventilation, photic stimulation, sleep and sleep deprivation. The EEG recording during sleep is particularly useful when there is a suspicion of epilepsy with partial seizures or the syndrome of benign childhood epilepsy with centrotemporal spikes. Hypsarrhythmia in Wests syndrome is an EEG pattern that changes from wakefulness to sleep. Using activation procedures, abnormalities are found in about 90% of patients with epilepsy (1). If routine EEG and EEG after sleep deprivation still reveals no abnormalities, long-term EEG monitoring may be used, which increases the detection rate of interictal and/or ictal events. Routine EEGs for specic dierential diagnostic purposes should include simultaneous electrocardiography (ECG) monitoring and where appropriate a monitoring of other parameters, i.e. respiration and muscle activity (polygraphic recording).
Equipment and electrodes electrode position

recommendations and can be used with dierent types of equipment from eight-channel EEG machines and those with more channels. As many EEG recordings are still analog, it is useful to have similar montages and examples of the common ones are given in the appendix (see Appendix). Recording parameters such as sensitivity, lter setting (notch lter) and time base should be in accordance with the international guidelines (4). The use of muscle lters and notch lter should be avoided if possible because of the risk of losing low amplitude spike potentials. The normal paper speed is 15 or 30 mm/s. Reduced paper speed allows better detection of slow wave abnormalities. The duration of recording time should be at least 30 min of artefact-free signals. In digital recordings, which allow remontaging, recording time should not be reduced, because the chances of recording interictal epileptiform activity increases with sampling time. The specic international guidelines for recording clinical EEG on digital media should be used (5).
Requirements for paediatric EEG recordings Preterm

The modied combined nomenclature derived from the 1020 system should be used for electrode location and the minimum number of electrodes should be 21 regardless of the number of the channels available on the EEG (2). A 16-channel recording is recommended for focal (partial) epilepsies; however, for diagnosing primary generalized epilepsy the number of channels is not crucial. Even with an eight-channel recording it is possible to obtain and locate the epileptiform discharges with the modied combined system but a longer sampling time using several montages will be required (3). Although there is no general agreement among EEG laboratories about montages, a routine EEG should (at least) include bipolar montages with longitudinal and transverse chains. These chains should be used with equal electrode distances and side-to-side symmetry to avoid the artefact of false amplitude asymmetry. Additional referential montages should be included in a routine recording. The previously mentioned montages are available in dierent settings in the international 2

Although the head may be small, the minimum number of electrodes is nine. The montages performed should be bipolar, longitudinal and transverse including the Cz electrode. These recordings are often performed in surroundings with a lot of electrical interference, giving artefacts. Such artefacts can be reduced by using short electrode cables thus placing the headbox and preamplier close to the head of the child. If possible a polygraphic recording should be obtained including ECG, respiration, eye movements and electromyography (EMG) to record muscle activity (6). The duration of the recording in small children should be at least 1 h to be able to evaluate wakefulness and sleep and reactivity. Photic stimulation in these children may give additional information and should be performed (7). Activation processes in order to evaluate EEG reactivity could include tapping and auditorial stimuli.
Infants and older children

The adult electrode placement should be used as young as possible as more electrodes will enhance sensitivity. A sleep recording will increase the chance of recording epileptiform activity and usually reduce movement and muscle artefacts. In all children despite age, recordings in spontaneous sleep are preferred to induced sleep.

Use of EEG for the diagnosis of epilepsy

Recording procedure

Standard EEG recording should attempt to include eye-opening, -closure and -blink procedures to exclude artefacts derived from eye movements. Voluntary hand movements can be useful for testing the reactivity of central rhythms. A standard EEG should aim to include the following activation procedures: (i) Hyperventilation depends upon the developmental age, and the level of co-operation of the child and the experience of the sta. (ii) Hyperventilation for 3 min with a continued recording for at least 2 min after cessation of hyperventilation, which may be worth repeating if absence epilepsy is suspected and the rst recording was negative. If there is unexpected abnormal hypersynchronization activity during hyperventilation, the possibility of physiological hypoglycaemia should be excluded. (iii) Photic stimulation should be performed with separate trains of photo ashes of 10 s duration for each frequency and with minimum intervals of 7 s. Each 10 s train consists of 5 s with open eyes followed by closure and 5 s with closed eyes. The distance between the lamp and the nose should be 30 cm. The method described below follows the previous recommendation for screening for and identifying photosensitive subjects (8, 9). Photic stimulation should not be performed during or within 3 min of hyperventilation. Photic stimulation should start at the frequency of 1 Hz and progress to 20 Hz unless generalized epileptiform discharges are evoked. This is immediately followed by a sequence with 60 Hz photic stimulation and then decreasing to 25 Hz. The trains should be performed with the following frequencies: 1, 2, 4, 6, 8, 10, 12, 14, 16, 18, 20, 60, 50, 40, 30 and 25 Hz and in that order. The total duration of photic activation time recommended is at maximum 6 min in patients without any reaction to intermittent photic stimulation. The procedure should be stopped at once if the patient has a warning or if epileptiform activity is evoked in the EEG. Not all EEG machines and photic stimulators can perform ashes up to 60 Hz, but testing using the single train with open eyes, eye closure and closed 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 false negative results. Sleep recordings can increase the accuracy of the diagnoses of epilepsy and should be performed when the standard EEG fails to show any

epileptiform activity and the level of clinical suspicion justies this investigation. Sleep deprivation can be a useful activation procedure. It must be recognized that sleep deprivation may provoke clinical seizures. If natural sleep is not achieved, a number of pharmacological agents can be used to induce sleep. One must be aware of pharmacologically induced EEG changes that occasionally may mimic epileptiform activity (interference of b activity). Adult patients should abstain from sleep the night prior to the recording. In children up to 12 years the sleeping time should be reduced as much as possible before the recording. A procedure used in many laboratories includes partial night sleep deprivation and EEG recording during the postprandial time when a nap is most likely to occur. The sleep recording should include wakefulness, drowsiness and at least 40 min of sleep. The depth of sleep should be at least to stages 2 and 3. The montage must include a bipolar montage with transverse chains which allow the identication of vertex waves additional channels should be used for electrooculography (EOG), surface EMG recorded from submental muscles, ECG and spirogram (polysomnography) in order to identify sleep stages and to distinguish between frontal discharges and eye movements. Polysomnography should always be used for all-night recordings in order to record epileptic nocturnal activity and to dierentiate this activity from non-epileptic events.
Long-term EEG recording

Long-term monitoring refers to EEG recording over an extended period. There are several options in performing long-term monitoring (10). In cases where a standard EEG followed by activation procedures does not show clear evidence of interictal epileptiform activity a prolonged recording time increases the possibility of detection of interictal and ictal events. The duration of such recordings can, depending on the clinical question raised, be from few hours to several days. The standard 1020 system in its modied version is recommended for these recordings. The long-term monitoring can be performed with portable equipment. This type of dynamic EEG recording allows the patient freedom to perform routine activities as an out-patient or inpatient during long-term EEG monitoring. Eightor 16-channel recordings are available. These EEG recordings may be useful to quantify seizure activity in patients with generalized discharges or for supervised monitoring in certain situations, that cannot be reproduced in the EEG laboratory 3

Flink et al. or hospital. The disadvantages include the increased risk of technical problems and lack of accurate simultaneous behavioural correlation. This type of monitoring may preferably be performed as sleep recordings because the artefacts are less during sleep. In order to correlate clinical behaviour and EEG ndings the recording requires video-equipment with synchronized time code, i.e. split screen video EEG. Digital video EEG systems are currently available. Long-term video-EEG monitoring may be the only way to distinguish epileptic from nonepileptic seizures and is mandatory as part of presurgical evaluation. It is of great value if a trained technician or nurse is constantly present during a video-EEG recording to test patient responsiveness, muscle tone, etc., whenever suspect ictal episodes occur. A set of testing procedures including language tests should be available to the examiner, to be administered during suspected ictal episodes.
EEG reporting
EUCARE for providing the subcommission the possibility of a workshop.

References
1. BINNIE CD, STEFAN H. Modern electroencephalography: its role in epilepsy management. Clin Neurophysiol 1999;110:167197. DERS HO, JASPER HH, ELGAR C. The ten2. KLEM GH, LU twenty electrode system of the International Federation. Recommendations for the Practice of Clinical Neurophysiology: Guidelines of the International Federation of Clinical Physiology. Electroenceph Clin Neurophysiol 1999;52(Suppl.):36. 3. AMERICAN ELECTROENCEPHALOGRAPHIC SOCIETY. Guideline Seven: a proposal for standard montages to be used in clinical EEG. J Clin Neurophysiol 1994;11:306. 4. EBNER A, SCIARNETTA G, EPSTEIN C, NUWER M. EEG instrumentation. Recommendations for the Practice of Clinical Neurophysiology: Guidelines of the International Federation of Clinical Physiology. Electroenceph Clin Neurophysiol 1999;52(Suppl.):710. 5. NUWER MR, CORNI G, EMERSON R et al. IFCN standards for digital recording of clinical EEG. Recommendations for the Practice of Clinical Neurophysiology: Guidelines of the International Federation of Clinical Physiology. Electroenceph Clin Neurophysiol 1999; 52(Suppl.):114. 6. AMERICAN ELECTROENCEPHALOGRAPHIC SOCIETY. Guideline Two: minimum technical standards for pediatric electroencephalography. J Clin Neurophysiol 1994;11: 69. 7. DE WEERD AW, DESPLAND PA, PLOUIN P. Neonatal EEG. Recommendations for the Practice of Clinical Neurophysiology. Guidelines of the International Federation of Clinical Physiology. Electroenceph Clin Neurophysiol 1999;52(Suppl.):14957. DGA, BINNIE CD, HARDING 8. KASTELEIJN-NOLST TRENITE GFA, WILKINS A. Photic stimulation: standardization of screening methods. Epilepsia 1999;40(Suppl. 4):759. DGA, BINNIE CD, HARDING 9. KASTELEIJN-NOLST TRENITE GFA et al. Medical technology assessment photic stimulation. Standardization of screening methods. Neurophysiol Clin 1999;29:31824. 10. AMERICAN ELECTROENCEPHALOGRAPHIC SOCIETY. Guideline Twelve: guidelines for long-term monitoring for epilepsy. J Clin Neurophysiol 1994;11:88110. ` RE F, 11. NOACHTAR S, BINNIE C, EBERSOLE J, MAUGERIE SAKAMOTO A, WESTMORELAND B. A glossary of terms most commonly used by clinical electroencephalographers and proposal for the report form for the EEG ndings. Recommendations for the Practice of Clinical Neurophysiology. Guidelines of the International Federation of Clinical Physiology. Electroenceph Clin Neurophysiol 1999; 52(Suppl.):2141.

The EEG report should be a response to the clinical question asked by the referring doctor. Apart from describing the waveforms and frequencies of the EEG signal there should be a clinical interpretation. A standardized report form is advocated and should include the following headings: (i) Information concerning the status of the patient, neurological condition, medication, last seizure, and clinical question. (ii) Information concerning the EEG recording; number of electrodes, use of special electrodes, recording conditions, level of consciousness of the patient during the recording, activation procedures, artefacts noted. (iii) Description of EEG; postcentral rhythm, background activity, asymmetries, epileptiform activity, specic EEG patterns, eect of activation procedures. (iv) Clinical interpretation; this puts the EEG ndings in a clinical context (the clinical signicance of the EEG ndings, prognosis, etc.), and responds to the clinical question. It is strongly advised to use the glossary proposed by the International Federation of Clinical Neurophysiology (11).
Acknowledgements
The authors wish to thank Guiliano Avanzini and Martin Brodie for valuable comments on the manuscript, and

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 recording machine is available.

Use of EEG for the diagnosis of epilepsy (ii) Use longitudinal bipolar, transversal bipolar and referential montages during each EEG recording. (iii) Duration of recording with one montage should be at least 2 min. (iv) The electrode connections (bipolar) should run in straight (unbroken) lines and the interelectrode distance should be kept equal. (v) Use a left above right order of derivations, i.e. on the recording page left-sided leads should be placed above right-sided leads for either alternating pairs of derivations or blocks of derivations. (vi) Tracings from the more anterior electrodes should be placed above those from the more posterior 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 be explored.
Pediatric montages

Because of the size of head, a reduced number of electrodes is usually used in neonates and young infants. A minimum of nine electrodes should be used (including Fp1, Fp2, C3, Cz, C4, T3, T4, O1 and O2). An eight-channel EEG machine is usually not enough as most of these recordings will require two or even more channels devoted to record polygraphic variables such as ECG and respiration. Montage using nine electrodes, combining longitudinal and transverse derivations. The longitudinal derivation block can be used with an

Proposed montages for eight- and 16-channel EEG machines Eight-channel recordings Bipolar montages Channel 1 2 3 4 5 6 7 8 Fp1-F3 F3-C3 C3-P3 P3-O1 Fp2-F4 F4-C4 C4-P4 P4-O2 Longitudinal Fp1-F7 F7-T3 T3-T5 T5-O1 Fp2-F8 F8-T4 T4-T6 T6-O2 F7-Fp1 Fp1-Fp2 Fp2-F8 C3-Cz Cz-C4 T5-O1 O1-O2 O2-T6 Transverse F7-F3 F3-Fz Fz-F4 F4-F8 T3-C3 C3-Cz Cz-C4 C4-T4 T3-C3 C3-Cz Cz-C4 C4-T4 T5-P3 P3-Pz Pz-P4 P4-T6 Referential montages F3-A1 C3-A1 P3-A1 O1-A1 F4-A2 C4-A2 P4-A2 O2-A2 Fp1-A1 F7-A1 T3-A1 T5-A1 Fp2-A2 F8-A2 T4-A2 T6-A2

Sixteen-channel recordings Bipolar montages Channel 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 Longitudinal Fp1-F3 F3-C3 C3-P3 P3-O1 Fp2-F4 F4-C4 C4-P4 P4-O2 Fp1-F7 F7-T3 T3-T5 T5-O1 Fp2-F8 F8-T4 T4-T6 T6-O2 F7-Fp1 Fp1-Fp2 Fp2-F8 F7-F3 F3-Fz Fz-F4 F4-F8 T3-C3 C3-Cz Cz-C4 C4-T4 T5-P3 P3-Pz Pz-P4 P4-T6 O1-O2 Transverse Fp1-Fp2 F7-F3 F3-Fz Fz-f4 F4-F8 A1-T3 T3-C3 C3-Cz Cz-C4 C4-T4 T4-A2 T5-P3 P3-Pz Pz-P4 P4-T6 O1-O2 F7-Fp1 Fp2-F8 F7-F3 F3-Fz Fz-F4 F4-F8 T3-C3 C3-Cz Cz-C4 C4-T4 T5-P3 P3-Pz PZ-P4 P4-T6 T5-O1 O2-T6 Referential montages Fp1-A1 F3-A1 C3-A1 P3-A1 Fp2-A2 F4-A2 C4-A2 P4-A2 F7-A1 T3-A1 T5-A1 O1-A1 F8-A2 T4-A2 T6-A2 O2-A2 Fp1-A1 Fp2-A2 F3-A1 F4-A2 C3-A1 C4-A2 P3-A1 P4-A2 F7-A1 F8-A2 T3-A1 T4-A2 T5-A1 T6-A2 O1-A1 O2-A2

Flink et al. eight-channel EEG machine, and when switching to the transverse block of derivations (channels 912) there will be possibilities for polygraphic channels even with the eight-channel machine.
Recording parameters

Electrodes The impedance (contact resistance) in the electrodes should be less than 5 kW, to reduce the noise artefacts and other interference.

Channel 1 2 3 4 5 6 7 8 9 10 11 12 ECG Respiration EOG Optional

Montage Fp1-T3 T3-O1 Fp2-T4 T4-O2 Fp1-C3 C3-O1 Fp2-C4 C4-O2 T3-C3 C3-Cz Cz-C4 C4-T4

Amplication The EEG machine is equipped with dierential ampliers with common mode rejection. A typical setting for the gain for the EEG is 7 lV/mm leading to deections of 320 mm for input voltages of 20140 lV. The ability of the ampliers to suppress voltages common to both electrodes is called the common mode rejection, it will for example reduce the noise from 50 or 60 Hz line current. The ability of an amplier to reject in phase and amplify out of phase potentials denes the common mode rejection ration of the amplier, which is always >80 dB.

Recording procedure Awake, test reactivity, intermittent eye-opening 10 min Drowsy/sleep 10 min Hyper-ventilation 3 + 3 min Photic stimulation 6 min (maximum)

Photic stimulation Time Flash Frequency Eye open Eye closed Continued Time Flash Frequency Eye open Eye closed Continued Time Flash Frequency Eye open Eye closed Continued Time Flash Frequency Eye open Eye closed 5s 50 Hz open closed 5s 7s Rest 5s 40 Hz open closed 5s 7s Rest 5s 30 Hz open closed 5s 7s Rest 5s 25 Hz open closed 5s 7s Rest 5s 16 Hz open closed 5s 7s Rest 5s 18 Hz open closed 5s 7s Rest 5s 20 Hz open closed 5s 7s Rest 5s 60 Hz open closed 5s 7s Rest 5s 8 Hz open closed 5s 7s Rest 5s 10 Hz open closed 5s 7s Rest 5s 12 Hz open closed 5s 7s Rest 5s 14 Hz open closed 5s 7s Rest 5s 5s 7s Rest 1 Hz open closed 2 Hz open closed 5s 5s 7s Rest 4 Hz open closed 5s 5s 7s Rest 6 Hz open closed 5s 5s 7s Rest

Use of EEG for the diagnosis of epilepsy

Polarity The polarity convention in recording with dierential ampliers is the following:

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

paper speed is preferably used in order to detect slow-wave abnormalities. The use of the higher paper speed might help in detecting asynchronicity in bilateral discharges.
Report form

Information by referring physician Patient ID Clinical history Neurological condition Medication Clinical question

Filtersetting For the majority of EEG recordings the signal frequency lies between 1 and 70 Hz and the bandwidth of the recording channels should therefore correspond to this frequency range. If a narrower bandwidth is used information will be lost and with a wider bandwidth, noise in the recorded data will contain irrelevant information. The low frequency lter should not be higher than 1 Hz (corresponding time constant of 0.16 s) and the high frequency lter should be 70 Hz. The use of notch lters (50 or 60 Hz) can distort sharp signals (i.e. spikes and sharp waves) and should only be used if other measures to reduce 50 or 60 Hz noise fail. The use of additional channels recording EMG, ECG, EOG, respiration, movements, etc., requires individual setting of gain and lters for each channel. Paper speed The print out of the recording (whether analog or digital) should use a time base (paper speed) of 15 or 30 mm/s with the option of 60 mm/s. The lower

Information by EEG technician Level of consciousness, vigilance, co-operation of the patient Clinical symptoms (jerking, moving, etc.) Activation procedures Technical artefacts

EEG description Background activity Postcentral rhythm Asymmetries Focal ndings Epileptiform activity, type, appearance and location Special EEG patterns Eect of activation procedures

EEG interpretation (keep it short) Interpretation of the EEG results in light of the clinical diagnosis and questions of the referring physician (clinical signicance, prognosis, etc.).