Acta Neurol Scand 2002: 106: 1–7 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 Affairs: 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 Affairs: Subcommission on European Guidelines. Acta Neurol Scand 2002: 106: 1–7. Blackwell Munksgaard 2002. The Commission of European Affairs 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 specific basic document of EEG methodology was recognized and the Subcommission on European Affairs 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; 4 Department 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; Report; European Guidelines; electroencephalo electroencephalography; graphy; diagnosis diagnosis of epilepsy; epilepsy; methodology; methodology; EEG reporting 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:
[email protected] Accepted for publication December 6, 2001
Introduction The primar primary y diagno diagnosis sis of epilep epilepsy sy is clinic clinical al but electroencephalography (EEG) plays a major role in evaluati evaluating ng epilepsy, epilepsy, recognizing recognizing that a normal normal rout routin inee EEG EEG does does not not excl exclud udee the the diag diagno nose se of epil epilep epsy sy.. The The main main indi indica cati tion on for for perf perfor ormi ming ng an EEG EEG is clin clinic ical al susp suspic icio ion n of an epil epilep epti ticc disord disorder. er. EEGs can also be useful useful in the evaluevaluation ation of encephalo encephalopati paties es (metabol (metabolic, ic, infectio infectious, us,
degene degenerat rative ive)) and focal focal brain brain lesion lesionss (cereb (cerebral ral infarcti infarction, on, haemorrh haemorrhage, age, neoplasm neoplasms). s). In paediatpaediatric practice, the EEG might help to determine the level of maturation of the brain. The EEG is not usef useful ul in foll follow owin ing g the the ther therap apeu euti ticc effec effectt of antiepil antiepileptic eptic drugs (AEDs) (AEDs) as interict interictal al epilepti epilepti-form activity is affected very little by AEDs. An exception exception is absence absence epilepsy epilepsy where the quantifiquantificati cation on of spik spikee-wa wave ve epis episod odes es is help helpfu full in following following the effect effect of treatment 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 West’s 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 specific differential 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 The ‘modified combined nomenclature’ derived from the 10–20 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 modified 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 different settings in the international 2
recommendations and can be used with different 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, filter setting (notch filter) and time base should be in accordance with the international guidelines (4). The use of muscle filters and notch filter 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 specific international guidelines for recording clinical EEG on digital media should be used (5).
Requirements for paediatric EEG recordings Preterm
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 preamplifier 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 staff. (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 first 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 flashes 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 flashes 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 justifies 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 identification 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 differentiate 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 10–20 system in its modified 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 findings 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 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, specific EEG patterns, effect of activation procedures. (iv) Clinical interpretation; this puts the EEG findings in a clinical context (the clinical significance of the EEG findings, 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).
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;110:1671–97. 2. KLEM GH, LU ¨DERS HO, JASPER HH, ELGAR C. The tentwenty 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.):3–6. 3. AMERICAN ELECTROENCEPHALOGRAPHIC SOCIETY. Guideline Seven: a proposal for standard montages to be used in clinical EEG. J Clin Neurophysiol 1994;11:30–6. 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.):7–10. 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.):11–4. 6. AMERICAN ELECTROENCEPHALOGRAPHIC SOCIETY. Guideline Two: minimum technical standards for pediatric electroencephalography. J Clin Neurophysiol 1994;11: 6–9. 7. DE W EERD 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.):149–57. 8. KASTELEIJN-NOLST TRENITE ´ DGA, BINNIE CD, HARDING GFA, WILKINS A. Photic stimulation: standardization of screening methods. Epilepsia 1999;40(Suppl. 4):75–9. 9. KASTELEIJN-NOLST TRENITE ´ DGA, BINNIE CD, HARDING GFA et al. Medical technology assessment photic stimulation. Standardization of screening methods. Neurophysiol Clin 1999;29:318–24. 10. AMERICAN ELECTROENCEPHALOGRAPHIC SOCIETY. Guideline Twelve: guidelines for long-term monitoring for epilepsy. J Clin Neurophysiol 1994;11:88–110. 11. NOACHTAR S, BINNIE C, EBERSOLE J, MAUGERIE `RE F, SAKAMOTO A, W ESTMORELAND B. A glossary of terms most commonly used by clinical electroencephalographers and proposal for the report form for the EEG findings. Recommendations for the Practice of Clinical Neurophysiology. Guidelines of the International Federation of Clinical Physiology. Electroenceph Clin Neurophysiol 1999; 52(Suppl.):21–41.
Appendix Recording montages
Basic rules for choosing montages:
Acknowledgements The authors wish to thank Guiliano Avanzini and Martin Brodie for valuable comments on the manuscript, and 4
(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
Longitudinal
1 2 3 4 5 6 7 8
Fp1-F3 F3-C3 C3-P3 P3-O1 Fp2-F4 F4-C4 C4-P4 P4-O2
Transverse 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
Referential montages
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
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
Transverse 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
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
Referential montages 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
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
5
Flink et al.
eight-channel EEG machine, and when switching to the transverse block of derivations (channels 9–12) there will be possibilities for polygraphic channels even with the eight-channel machine.
Channel
Recording parameters
Electrodes
The impedance (contact resistance) in the electrodes should be less than 5 k W, to reduce the noise artefacts and other interference.
Montage
1 2 3 4 5 6 7 8 9 10 11 12 ECG Respiration EOG Optional
Fp1-T3 T3-O1 Fp2-T4 T4-O2 Fp1-C3 C3-O1 Fp2-C4 C4-O2 T3-C3 C3-Cz Cz-C4 C4-T4
Amplification
The EEG machine is equipped with differential amplifiers with common mode rejection. A typical setting for the gain for the EEG is 7 lV/mm leading to deflections of 3–20 mm for input voltages of 20–140 lV. The ability of the amplifiers 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 amplifier to reject in phase and amplify out of phase potentials defines the common mode rejection ration of the amplifier, which is always >80 dB.
Recording procedure Awake, test reactivity, intermittent eye-opening 10 min
Drowsy/sleep
Hyper-ventilation
Photic stimulation
10 min
3 + 3 min
6 min (maximum)
Photic stimulation Time Flash Frequency Eye open Eye closed
5s
5s
7s
5s
5s
Rest 1 Hz open
7s
5s
5s
Rest 2 Hz open
closed
7s
5s
5s
Rest 4 Hz open
closed
7s Rest
6 Hz open closed
closed
Continued Time Flash Frequency Eye open Eye closed
5s
5s
7s Rest
8 Hz open
5s
5s
7s Rest
10 Hz open closed
5s
5s
7s Rest
12 Hz open closed
5s
5s
7s Rest
14 Hz open closed
closed
Continued Time Flash Frequency Eye open Eye closed
5s
5s
7s Rest
16 Hz open
5s
5s
7s Rest
18 Hz open closed
5s
5s
7s Rest
20 Hz open closed
5s
5s
7s Rest
60 Hz open closed
closed
Continued Time Flash Frequency Eye open Eye closed
6
5s
5s
50 Hz open
7s Rest
5s
5s
40 Hz open closed
7s Rest
5s
5s
30 Hz open closed
7s Rest
5s
5s
25 Hz open closed
closed
7s Rest
Use of EEG for the diagnosis of epilepsy Polarity
The polarity convention in recording with differential 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 downward deflection. (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 upward deflection.
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 filter should not be higher than 1 Hz (corresponding time constant of 0.16 s) and the high frequency filter should be 70 Hz. The use of notch filters (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 filters for each channel.
Information by EEG technician
Paper speed
EEG interpretation (keep it short)
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
Interpretation of the EEG results in light of the clinical diagnosis and questions of the referring physician (clinical significance, prognosis, etc.).
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 findings Epileptiform activity, type, appearance and location Special EEG patterns Effect of activation procedures
7
