Electroencephalography and clinical Neurophysiology , 87 (1993) 421-436 421 © 1993 Elsevier Scientific Publishers Ireland, Ltd. 0013-4694/93/$06.00

EEG RIO 01

Recommended standards for electroretinograms and visual evoked potentials. Report of an IFCN Committee

Gastone G. Celesia (Chicago, IL, USA) (Chairman), Ivan Bodis-Wollner (Omaha, NE, USA), Gian Emilio Chatrian (Seattle, WA, USA),

Graham F.A. Harding (Birmingham, UK), Samuel Sokol (Boston, MA, USA) and Henk Spekreijse (Amsterdam, The Netherlands)

( R e c e i v e d for pub l i c a t i on : 3 J u n e 1993)

Contents

Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 421

(1) Terminology . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 422

(2) Visual evoked potentials (VEPs) . . . . . . . . . . . . . . . . 422 (I) Basic technology . . . . . . . . . . . . . . . . . . . . . . . . . 423

(1) Visual stimulation . . . . . . . . . . . . . . . . . . . . . 423 (2) Calibration of visual stimuli . . . . . . . . . . . . . . 424 (3) Electrodes . . . . . . . . . . . . . . . . . . . . . . . . . . 424 (4) Recording equipment . . . . . . . . . . . . . . . . . . 424

(II) Clinical protocol . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .425 (1) General s ta tements . . . . . . . . . . . . . . . . . . . . 425 (2) Measurements of VEPs . . . . . . . . . . . . . . . . . 425

(III) Report of results . . . . . . . . . . . . . . . . . . . . . . . 426 (IV) Suggested specific protocols . . . . . . . . . . . . . . . 427

(3) Pattern electroretinography (PERG) . . . . . . . . . . . . . 428 (I) Basic technology . . . . . . . . . . . . . . . . . . . . . . . . . 428

(1) Visual stimulation . . . . . . . . . . . . . . . . . . . . . 428 (2) Calibration of visual stimuli . . . . . . . . . . . . . . 428 (3) Electrodes . . . . . . . . . . . . . . . . . . . . . . . . . . 428 (4) Recording equipment . . . . . . . . . . . . . . . . . . 428

(II) Clinical protocol . . . . . . . . . . . . . . . . . . . . . . . . 429 (1) Patient preparation . . . . . . . . . . . . . . . . . . . . 429 (2) Measurements of P E R G . . . . . . . . . . . . . . . . 429

(III) Report of results . . . . . . . . . . . . . . . . . . . . . . . 429 (IV) Suggested specific protocols . . . . . . . . . . . . . . . 429

(4) Full-field flash electroretinography (ERG) . . . . . . . . . 430 (1) Basic technology . . . . . . . . . . . . . . . . . . . . . . . . . 431

(1) Light diffusion . . . . . . . . . . . . . . . . . . . . . . . . 431 (2) Electrodes . . . . . . . . . . . . . . . . . . . . . . . . . . 431 (3) Light sources . . . . . . . . . . . . . . . . . . . . . . . . . 431 (4) Light adjustment and calibration . . . . . . . . . . . 432 (5) Electronic recording equipment . . . . . . . . . . . 432

Correspondence to: Prof. Gastone G. Celesia, Dept. of Neurology, Loyola University School of Medicine, 2160 South First Avenue, Maywood, IL 60153 (USA). Fax: 708-216.56.17.

(ll) Clinical protocol . . . . . . . . . . . . . . . . . . . . . . . . 432 (1) Preparation of the patient . . . . . . . . . . . . . . . 432 (2) ERG measurements and recording . . . . . . . . . 433

(III) Specific responses . . . . . . . . . . . . . . . . . . . . . . 433 (1) ' Rod' response . . . . . . . . . . . . . . . . . . . . . . . 433 (2) Maximal response . . . . . . . . . . . . . . . . . . . . . 433 (3) Oscillatory potentials . . . . . . . . . . . . . . . . . . . 433 (4) Single-flash 'cone ' response . . . . . . . . . . . . . . 433 (5) Flicker response . . . . . . . . . . . . . . . . . . . . . . 433

(5) Visual evoked potentials in pediatrics . . . . . . . . . . . . 434 (I) Pediatric protocol . . . . . . . . . . . . . . . . . . . . . . . . 434

(1) Pattern VEPs . . . . . . . . . . . . . . . . . . . . . . . . 434 (2) Flash VEPs . . . . . . . . . . . . . . . . . . . . . . . . . . 435 (3) ERG . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 435

References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 435

Introduction

V i s u a l f u n c t i o n c a n b e a s s e s s e d e l e c t r o p h y s i o l o g i -

c a l l y w i t h a v a r i e t y o f t e s t s t h a t h a v e b e c o m e r o u t i n e in

n e u r o l o g i c a l a n d o p h t h a l m o l o g i c a l l a b o r a t o r i e s . R e t i -

n a l p h y s i o l o g y a n d p a t h o p h y s i o l o g y c a n b e e v a l u a t e d

b y e l e c t r o r e t i n o g r a p h y . O p t i c n e r v e , v i s u a l p a t h w a y s

a n d v i s u a l c o r t i c e s c a n b e a s s e s s e d b y v i s u a l e v o k e d

p o t e n t i a l s ( V E P s ) . T o i m p r o v e c l i n i c a l c a r e a n d t o

p e r m i t c o m p a r i s o n a m o n g d i f f e r e n t l a b o r a t o r i e s , s t a n -

d a r d i z a t i o n o f p r o t o c o l s is e s s e n t i a l . T h e p r e s e n t r e c -

o m m e n d a t i o n s o u t l i n e b a s i c t e r m i n o l o g y a n d s t a n d a r d

m e t h o d s a n d a d v o c a t e d e s i r a b l e i n s t r u m e n t a t i o n . M o s t

o f t h e g u i d e l i n e s r e p r e s e n t t h e m i n i m a l r equ i r emen t s

f o r r e c o r d i n g b a s i c v i s u a l e v o k e d r e s p o n s e s in a n y

l a b o r a t o r y . H o w e v e r , w e d o n o t i m p l y t h a t m o r e a d -

v a n c e d o r c o m p l e x p r o c e d u r e s s h o u l d n o t b e u s e d .

A l s o , w e h a v e r e f r a i n e d f r o m p r o m u l g a t i n g a " c o o k

422 IFCN COMMITTEE

book" of standards, but rather suggest rigorous meth- ods to describe and measure visual stimuli and re- sponses.

Progress and improvement require individualization of procedures and development of novel methodolo- gies, and these efforts should be encouraged. Yet, minimal standards are needed to assure adequate ap- preciation of visual physiology and delivery of good quality of care.

This set of recommendat ions is divided into 5 parts: (1) terminology; (2) visual evoked potentials (VEPs); (3) pat tern electroretinograms (PERGs); (4) full-field flash electroretinograms (ERGs); and (5) VEPs in pe- diatrics.

These recommended standards are not a mandate for specific procedures for individual patients and do not represent safety standards.

(1) Terminology

"The chief merit of language is clearness" Galen: On the natural faculties I,ii circa 2nd Cent.

"The ill and unfit choice of words wonderfully obstructs the under- standing"

F. Bacon: Novum: Organum, "Aphorisms" XLIII, 1561-1626

To improve communication among scientists and clinicians a standardized nomenclature needs to be adopted. The nomenclature advocated in this report is derived from: (1) established use in the last decade, and (2) introduction of clarifications in areas where conflicting terms have been used.

The electroretinogram is a mass response of the retina to visual stimuli. Presently two types of elec- troretinogram are used clinically: (1) the full-field flash electroretinogram; and (2) the pat tern electroretino- gram. The full-field flash electroretinogram is abbrevi- ated as ERG; and the pattern electroretinogram as PERG. The nomenclature of both electroretinograms should follow the tradition of naming successive wave forms in alphabetical order: a wave, b wave, c wave, etc. The use of terms such as N and P for negative and positive waves causes confusion with the terminology used for visual evoked potential responses and should be avoided. To distinguish between E R G and P E R G the subscript pt (for pattern) should follow the letters a, b, and c for P E R G waves. For example, a wave will indicate the first negative wave of flash E R G whereas apt wave will indicate the first negative wave of the pat tern E R G etc.

Visual evoked potentials are electrical potential dif- ferences recorded from the scalp in response to visual stimuli; they are abbreviated as VEPs. VEPs to stimuli repeated at low rates are referred to as transient visual

eL, oked potentials. Traditionally, transient z~isual eL~oked potentials are abbreviated as VEPs.

Response components of VEPs should be desig- nated in accordance with their polarity and peak la- tency. Negative waves should be designated N followed by a number indicating the peak latency of the wave in msec, i.e., N70; positive waves should be designated P followed by a number indicating the peak latency in msec, i.e., P100. Although the precise peak latency depends on stimulus characteristics, a standardized labeling scheme is recommended for clarity and sim- plicity.

Steady-state L,isual eL,oked potentials" are responses to visual stimuli at relatively high frequencies (above 3.5/sec). These responses overlap one another and merge into quasi-sinusoidal oscillations that remain constant for the duration of the stimulation. They were defined by Regan 1 as "repeti t ive evoked potentials whose constituent discrete frequency components re- main constant in amplitude and phase over an in- finitely long time period." They are abbreviated as S- VE Ps.

Jean Fourier in 1822 demonstrated that periodic wave forms can be analyzed into the sum of harmoni- cally related sine waves of specific frequencies, phases and amplitudes. Thus, any periodic wave with a rate of F times per second can be described in Fourier analysis by the sum of a series of sine waves whose frequencies are FHz, 2FHz, 3FHz, etc 1"2. The response at the frequency of the stimulus is the fundamental or first harmonic (FHz), the response component at twice the stimulus frequency is called the second harmonic 2FHz and so on. The Fourier analysis is a transformation from the time domain to the frequency domain turning time data into f r equency /phase data. Specifically in the case of S-VEPs, the periodic responses present continously in the VEPs are transformed into discrete harmonic frequency responses in the frequency do- main.

Response components of S-VEPs should be desig- nated in accordance to the Fourier nomenclature F for the fundamental response and 2F for the second har- monic response.

(2) Visual evoked potentials (VEPs)

Visual evoked potentials represent a mass response of cortical and possibly subcortical visual areas. VEPs are employed routinely in the assessment of the func- tional integrity of the visual pathways. The identity of the cortical generators of both pat tern and flash evoked VEPs is neither known nor agreed upon 2 5. The scalp VEPs distribution may reflect the complex interaction of electrical field potentials within multiple cortical

RECOMMENDED STANDARDS FOR ERGs AND VEPs 423

visual areas. Precise knowledge of VEP source location is not essential to the clinical utilization of VEPs 2. The present recommendat ions are organized as follows:

Basic technology. Visual stimulation. Calibration of visual stimuli. Electrodes. Recording equipment.

Clinical protocol. Report ing of results. Suggested specific protocols.

(I) Basic technology

(1) Visual stimulation VEPs can be elicited by either pat terned or unpat-

terned stimuli. The stimulus is processed mostly via multiple parallel pathways. Thus, modification of the visual stimulus will greatly affect VEPs. The committee is cognizant of the occasional need to vary visual stimu- lation as specific problems arise and agrees that this versatility is to be encouraged. The user, however, must be careful not to equate data evoked by one stimulus to data obtained under different stimulation parameters.

Unpat terned stimuli consist most often of strobo- scopic flashes 6. The following measurements must be defined: flash duration, stimulus wave length (color), stimulus intensity (luminance), background intensity (luminance), and stimulus rate. In the case of other stimuli, such as sinusoids or square waves or white noise, the wave shape, frequency (spectrum), modula- tion depth, mean intensity (luminance), color and size of the stimulus field should also be defined.

The recommendat ions of the ISCEV (see pages 425-431 of present document) for the measurements of light sources for flash ERGs should be adhered to.

Patterned stimuli are most frequently used to elicit VEPs in clinical settings.

The following stimulus measurements must be spec- ified:

(a) Type of pattern. The type of pat tern should be specified, i.e., checkerboard, square-wave gratings, sine-wave gratings, etc. The term gratings should al- ways be preceded by a qualifier describing the charac- teristic of the gratings. Sine:wave gratings are charac- terized by a sinusoidal variation in light in one direc- tion. They can be graphically described by a one-di- mensional sine wave and therefore contain power at a single spatial frequency by Fourier analysis 7. Square- wave gratings consist of a series of bars with sharply defined edges and contain power at multiple harmonic spatial frequencies.

We advocate the use of either checkerboard or gratings. When gratings are used, their orientation

(horizontal, vertical, etc.) should be specified s. The pat tern should be achromatic black and white (bright and dark).

(b) Size of the pattern elements. The dimensions of the individual checks or bars should be described by the visual angle that they subtend at the subject's eye. Checks are traditionally expressed in minutes of arc, whereas gratings are customarily reported in cycle per degree. The following formula is used to calculate the visual angle of the pattern elements:

a = t a n - ~ ( W / 2 D ) x 120

where "a" is the visual angle in minutes of arc, " W " is the width of the check in millimeters and " D " is the distance of the pat tern from the corneal surface in millimeters.

Measurements of the visual angle in minutes or degrees can be converted to cycles per degree (cpd) by the formula:

cpd = 6 0 / W

where " W " is the period of the gratings in minutes (one period contains one dark and one light grating). In the cases of checks the formula is:

cpd = 3 0 / W

" W " is the diagonal measure of the check in minutes of arc. The measurement in cpd defines the spatial frequency of the stimulus.

(c) Total field size. The total field size should be expressed in degrees of visual angle. The location of the fixation point in relation to the field should also be noted.

(d) Method of presentation of the pattern. Pattern reversal is advocated to reduce modulated straylight and to maintain a constant mean luminance during stimulation. Pattern onset-offset can have the same advantages. A change in the method of presentation may affect the response, and pat tern reversal VEPs are different than pattern onset-offset VEPs.

(e) Rate of presentation of the pattern. The rate of pattern presentation should be expressed in Hz. With pattern reversal stimuli the reversal rate (in Hz) indi- cates a full cycle reversal, i.e., a temporal frequency of 4 Hz indicates 8 reversals of the pattern equal to a stimulus interval of every 125 msec. In case of pattern onset-offset both onset and offset duration should be given.

(f) Stimulus luminance. The response amplitude and latency will vary with the stimulus luminance. The luminance of a pat tern stimulus is sometimes incor- rectly referred as intensity. The luminance of the field is measured by a photometer and is expressed in can- dela per square meter (cd/m2). Note that 3.43 c d / m 2 is equal to 1 footLambert (fL) photometric unit.

424 IFCN COMMITTEE

The mean luminance should be measured at the center of the field and can be expressed for spatially mirror symmetric stimuli by the formula:

(Lma x -t- L m i n ) / 2

where Lma x and Lmi n represent the maximum and minimum luminance value across the stimulus field.

It is recommended that the mean luminance of the field be at least 100 c d / m 2. The field luminance should be uniform and vary by less than 20% between the center and the periphery of the field. Isoluminance of the stimulus can be controlled easily by putting a diffuser in front of the stimulus.

(g) Background luminance. Background luminance should be kept constant throughout the recording and the same background luminance should be used for each given protocol.

(h) Contrast. Contrast is the difference in lumi- nance between the bright and the dark portion of the pattern and is expressed by the formula:

C = [ (L ...... - Lmin)/(Lmax + Lmin)] X 100%

where C is the contrast in percent and "Lmax" and "Lmin" are the maximum and minimum luminances of the pattern.

(k) Type of stimulator. Pattern stimuli may be dis- played in various ways such as on a TV or video monitor, an oscilloscope, or on a rear projection screen using a projector and a movable mirror. Different TV and video monitors will evoke equivalent responses provided that all of the parameters of the stimuli are equal.

There is no perfect stimulator. Comparable results among laboratories will be possible only if the physical characteristics of the stimuli (including line frame rate of a TV monitor) are matched.

(2) Calibration of v&ual stimuli Visual stimulators should incorporate the following

characteristics: (1) Capability of modifying the type of the pat tern

and be able to generate at least checks, and square- wave gratings. The availability of sine-wave gratings is highly desirable.

(2) Capability of varying both the size of the spatial elements and the rate of presentation.

(3) Ability to change the contrast and the luminance independently of the other variables.

(4) Possibility of locating a fixation marker any- where on the screen.

The stimulus luminance of the field and the Lma x

and Lm~ n of the pat tern must be documented by the manufacturer. The luminance is measured by a spot photometer . The light output may vary with the rate of presentation of the pat tern (repetition rate of the stim- ulus); thus, separate calibration will be needed for the

various rates of presentation. The photometer used must meet the standard for photometric measure- ments. The committee recommends that in the future manufacturers of visual stimulators provide an ade- quate photometer as part of the equipment.

Light output may vary over time as some of the electronics or optics of the system change. The manu- facturer of the stimulators should provide documenta- tion about the stability of the system and warning about sources of instability. Recalibration of the stimu- lators is therefore necessary. The frequency of recali- bration will vary with the system utilized and could be as high as once a week. Self-calibrating units are rec- ommended.

(3) Electrodes Standard disk E E G electrodes are recommended

for recording VEPs. The relative position of skull land- marks and brain in the posterior regions of the head is quite variable. For instance, the distance between the inion (one of the skull landmarks used in the 10-20 international system) and the posterior tip of the cal- carine fissure is variable; the variation zone as mea- sured by MRI was 4 cm in the 16 subjects studied 9. The electrode Oz is therefore located at the middle of the variation zone of the calcarine fissure 9. There is also considerable asymmetry between the right and left occipital poles 5'9'1°.

The recording electrode should be placed at the Oz position of the 10-20 international system. The refer- ence electrode should be located either in the frontal region (electrode Fpz or an electrode 12 cm above the nasion). Alternatively linked ears may be used as refer- ence. The ground electrode can be placed at the vertex (electrode Cz). Electrode impedance should be less than 5000 ,(2.

Utilization of 2 or 4 electrodes placed laterally to the midline occipital electrode is not adequate to study VEP topography and should not be utilized for clinical diagnosis of visual field defects. VEPs in response to hemifield visual stimulation are not yet sufficiently reli- able and sensitive to be utilized in the clinical assessment of retrochiasmatic lesions.

Topographic studies of the amplitude distribution of VEPs require the utilization of a minimum of 16 elec- trodes. Topographic mapping of VEPs is a complex and controversial subject, therefore, the committee be- lieves that utilization of VEP mapping in clinical set- tings is premature.

(4) Recording equipment The bandwidth of the recording system (amplifiers

and preamplifiers) should be at least 0.3-250 Hz ( - 3 dB) with roll-off slopes not exceeding 12 dB/oc tave for the low frequencies and 24 dB/oc tave for the high frequencies. The amplified signal for each channel

RECOMMENDED STANDARDS FOR ERGs AND VEPs 425

should be matched to reduce channel-to-channel vari- ability to 1% or less after computer adjusted gains based on calibration pulses and biocalibration compari- son. The square root of the power of the amplifier "noise" should be preferably less than 1.0 ~V but should not exceed 2.0 ~V and the amplifier should have a DC offset of less than 1 ~V.

The amplifier must be electrically isolated from the patient. The current standards for safety of a biologi- cally recording system used in humans should be ad- hered to. Timing and synchronization of stimulus and data sampling should be matched.

The analog signal should be digitized with a mini- mum sample rate of 200 samples/sec per channel and a 10-bit resolution per sample. The sampling rate fol- lows the Nyquist theorem stating that the sampling rate must be at least twice the highest frequency contained in the analog signal. With sampling rates at 200 or 256 Hz the digital data will have good reproduction up to 100 and 128 Hz respectively. The system should in- clude software for the following procedures: digital filtering, averaging, Fourier analysis. The Fourier anal- ysis should provide the amplitude, phase and power spectra of the biological signals under study, with the phase corrected for the filter setting. For the peak of interest the software should provide information about the amplitude in p~V and the phase in radians or degrees. Automated artefact rejection should be pro- vided.

The display of the system should provide the ability of viewing: (1) the amplified signal of each channel before it is averaged or otherwise manipulated, (2) the averaged signals, and (3) the spectrum of the signal. The system should also provide a noise estimate (pre- ferably with + average). Suitable hardware for printing hard-copies of the results should be provided. Mag- netic or optical devices for storage of the digitized unprocessed data should be part of the recording sys- tem. Manufacturers are encouraged to provide storage of the digitized data. Accurate calibrations must be provided for the whole recording system, from the input to the processed output.

(II) Clinical protocol

(1) General statements The visual system processes information via parallel

pathways that originate in the retina ~1. Thus, the selec- tion of the visual stimuli must relate to the physiologi- cal properties of those specific parallel pathways that one is interested in testing 12-14. It follows that there is no "best method" of stimulation. Rather, stimulation and recording should be tailored to the clinical prob- lem that is being addressed ~2-~5. The committee rec- ommends that each laboratory develops several proto-

cols for various clinical problems. It advocates the use of more than one stimulus size.

Pattern reversal with checks, square-wave gratings a n d / o r sine gratings may be utilized. In the detection of optic neuropathy, for example, stimuli should be chosen that optimally stimulate the central visual field (the fovea). To this end, the pattern size should have a spatial frequency above 2 cpd, i.e., should subtend visual angles of 22 rain or less for checks and 30 min or less for square-wave gratings.

For other clinical problems the parameters of the stimuli may need to be different, or a pattern onset- offset presentation mode may be preferable.

In cases of abnormal VEPs, if the differential diag- nosis includes macular and optic nerve pathology the committee recommends the simultaneous recording of P E R G and VEP to patterned stimuli. A normal PERG in the presence of a clearly delayed or otherwise ab- normal VEP would suggest that the dysfunction is localized beyond the retina.

During pattern stimulation a fixation point should be provided. Before the onset of the recording the pupil diameter and the visual acuity of each eye should be determined. Appropriate corrective lenses must be worn to compensate for refractive errors. Defocusing produces a delay in both pattern reversal and onset- offset VEPs 27. If a deficit in visual acuity is present the technologist should determine whether it can be cor- rected with a "pin hole."

The pupils should not be dilated to prevent interfer- ence with accommodation.

(2) Measurements of VEPs Transient VEPs consist of a series of waves of

alternating polarity. The committee recommends that the waves be named N followed by the latency in msec for negative waves and P followed by the latency in msec for the positive deflections, i.e., N70, P100, N135, etc. Both amplitude and latency should be measured for the most common components of the waves. For VEPs evoked by reversing checks or bars customarily the waves to be measured are N70 and P100.

When simultaneous recording of PERG and VEPs is carried out the inter-peak latency between the bpt wave and the P100 should be measured. The polarity of the VEPs should be specified in the tracings. One replication of the VEP is recommended to demon- strate the reliability of the data.

Steady-state VEPs can be assessed by Fourier analy- sis. For pattern reversal S-VEPs, the 2nd harmonic response (at twice the frequency of the stimulus) should be measured and the amplitude and phase of the response should be determined. Steady-state pattern VEPs to on-off stimulation differ from pattern reversal S-VEPs because they contain large first or fundamen-

426 IFCN COMMITTEE

tal response (at the stimulation rate) and 2nd harmonic response t~. The amplitude and phase of both responses should be measured 2'7'17-21.

Normatice c'alues. We recommend that each labo- ratory establish or confirm its own normal values. Nor- mative values obtained from other institutions may be utilized only if equivalent stimulation and recording methods are employed and only after testing the valid- ity of the adopted normal values on at least 10 locally gathered subjects. The measurements between the two groups should be similar.

Normative values are influenced by age and possibly by gender. Thus a laboratory a n d / o r a manufacturer establishing normal values should gather data from age-matched individuals of the two sexes. A total of 20 subjects should be collected for each decade after about age 6 years. The boundaries of normals should be set at the 95-99% tolerance level limit for 95% of the normal population.

The same methods of statistical analysis of ampli- tude can be used for transient and steady-state VEPs, however, phase data of S-VEPs require special statis- tics that differ from those used in latency analysis due to the circular value of the measure.

Ampli tude of VEPs and S-VEPs and latency of transient VEPs can be analyzed by weighted linear regression analysis or by the non-parametr ic B- spline 22-24. Most VEP data have a non-normal (non- gaussian) distribution with significant skewness and kurtosis. Therefore calculating mean and standard de- viation on the raw data is inaccurate. The data must first be t ransformed to approximate normal distribu- tion. This transformation can be achieved by taking the natural logarithm, the square root, or the reciprocal of values that have abnormal distribution. The mean and standard deviation can then be calculated on the trans- formed data 24. If transformation of the data has been utilized the method of transformation should be re- ported.

Measurements obtained from the right and the left eye cannot be treated as independent random vari- ables 23. The values from the two eyes are highly corre- lated. Therefore, they must either be tabulated sepa- rately or the mean value of the two eyes used as a single number 23. As the values of N70 and P100 are influenced by age and sex, a more accurate method to establish normal boundaries is the use of a "projected value" derived from weighted regression anlysis 22-2a.

S-VEP phase is measured by the angle value in a circular distribution, which is divided into 360 equal intervals or degrees. However, the position of the zero point in the circle is arbitrary. Zero can be placed at the top of the circle as for example in the display of time of the day circularly distributed in equal intervals of 24 h. Due to this circular distribution the data cannot be evaluated by a linear scale and calculating

routine mean and standard deviation is inappropri- ate 19"21'25'26. The mean angle, the measure of concen- tration and the angular dispersion should be calcu- lated 26. The boundaries of normality can then be based on the mean angle and angular dispersion (also re- ferred as circular standard deviation) 21. A measure of concentration close to 1 indicates that all the data are in the same direction, whereas a measure of concentra- tion close to zero indicates that there is a great deal of dispersion and that a mean angle cannot be defined.

The statistical method used to obtain normal bound- aries should be stated.

(III) Report of results

All reports should contain basic information about: (1) patient; (2) clinical status; (3) technical data; (4) normative values; (5) results; (6) interpretation.

(1) Patient information should include: name, age, gender and patient identification number.

(2) The clinical question being addressed should be provided in the report. The following information about the patient vision should be included in the report:

(a) Visual acuity with the corrective lenses used during the recording. If visual acuity is less than 20/20 (6/6) , pin-hole acuity is measured.

(b) Pupil size in mm at time of recording. (c) Visual field defects. (d) Ability to fixate. (e) Status of the eye optics (i.e., presence or ab-

sence of cataracts, corneal opacities or other problems that may interfere with visualization of the pattern).

(3) Technical data. The following stimulus parame- ters should be reported: type and size of the pattern, rate of presentation, mean luminance of the field, size of the field, and contrast. For flash VEPs the flash intensity, the rate of stimulation and background lumi- nance should be specified. The report should state which equipment was used, and when it was calibrated, and whether monocular or binocular stimulation was used.

(4) Normative values. (a) Transient VEPs. The labo- ratory normal values for amplitude and latency of the N70 and P100 should be reported. Normal values should include the boundaries of normality. (b) Steady-state VEPs. The laboratory normal values for amplitude and phase of the 2nd harmonic response should be reported. Normal values should include the mean amplitude, the phase circular mean and the boundaries of normality.

(5) Results. The report should include the ampli- tude, latency or phase values of the VEPs a n d / o r S-VEPs for each eye. Representative wave forms of the response should be provided with calibration signals for time and amplitude.

R E C O M M E N D E D S T A N D A R D S FOR ERGs A N D VEPs 427

(6) Interpretation. Interpretat ion should include a statement about the normality or abnormality of the test. The type of abnormality should be described. The VEP data should not be interpreted in isolation of the clinical picture. VEPs data are ancillary to the neuro- ophthalmological examination. For instance, an abnor- mally prolonged P100 latency limited to one eye may indicate dysfunction of the optic pathways only when ocular and retinal pathology have been excluded by appropriate ophthalmological examination. VEP ab- normalities are not diagnostic of specific diseases. Dif- ferent pathological processes affecting the same anatomi- cal locus can produce similar physiological disturbances.

Flash VEPs have considerable variation in morphol- ogy and amplitude among different individuals and they should, therefore, be interpreted with extreme caution. Demonstrat ion of flash VEPs in patients sus- pected of blindness provides evidence that some visual input reaches the cortex. It does not indicate that the visual system is intact nor that visual perception is preserved.

These recommendat ions may require revisions every 5 years to keep abreast of the rapid changes in technol- ogy.

(IV) Suggested specific protocols

The following standardized protocols are suggested as the minimum requirement to obtain reliable and reproducible VEPs.

The committee recommends the use of pattern-re- versal checkerboard or vertical gratings stimulation. The following criteria should be adhered to:

(1) Recording The recommended system bandpass is 1.0 to 250-300

Hz with filter roll-off slopes not exceeding 12 d B / o c - tave for the low frequency setting and 24 dB/oc tave for the high frequency setting.

A montage consisting of 2 derivations is sufficient for recording VEPs.

VEPs are widely distributed over the scalp from the vertex to the inion. The following montage is sug- gested: O z - F p z and Oz -A1-A2 (linked ears).

The ground electrode should be placed at Cz. The suggested montage will assure the recording of

a reproducible VEP, even in the occasional cases when the potential gradient of the visual response is very prominent at the vertex.

For transient VEPs an analysis time of 250 msec is recommended with averaging of 100 individual trials. At least two averages should be obtained and superim- posed to verify reproducibility of the results.

For S-VEP it is recommended that the analysis time be 2 sec. Direct Fourier transform of the signal is

suggested with analysis of the second harmonic compo- nent (2FHz).

(2) Stimulation The committee recommends that a minimum of 3

stimuli be routinely tested. The following parameters are suggested for either

alternating checks or alternating vertical gratings. Grat- ings can be either sine or square-wave. Checks, sine gratings and square-wave gratings are not equal, thus a given laboratory should select one of these alternatives. The recommended values of the stimulus are:

- Contrast between 50 and 80%. - Full field size: greater than 8 °. - Size of the pat tern elements: 14-16 min, 28-32

min, 56-64 min. The smaller size of 14-16' is optimal to stimulate the fovea, but is very easily affected by visual acuity changes. The wider size patterns may also stimulate the parafoveal region thus may elicit normal responses in cases of foveal dysfunction.

- Rate of presentation of the pattern: for transient VEPs the preferred rate is 1 Hz (producing a reversal every 500 msec); for S-VEPs the sug- gested rate is 4 or 8 Hz (producing a reversal rate of 8 or 16 Hz respectively).

- Mean luminance of the center field should be at least 50 c d / m 2.

- B a c k g r o u n d luminance: under photopic condi- tions should be 20-40 c d / m 2.

(3) Results The amplitude and latency of the N70 and P100 of

VEPs for each eye and for each stimulus size should be reported. The interpeak latency for N70 and P100 should also be reported. Transient VEP morphology is variable and the interpreter should be familiar with the different shapes of normal VEPs. In 0.5% of normals VEPs obtained to checks have a P100 with a W-shaped configuration with P100 subdivided into 2 peaks 27. However, both peaks have latencies within the normal boundaries. The best way to determine which peak may be measured as P100 is to obtain VEPs to at least 3 different size patterns. Usually the larger cheks will yield only one P100 peak.

If the laboratory utilizes S-VEPs, the amplitude and phase of 2FHz for each eye and each stimulus size should be reported.

When an abnormality is reported it should be speci- fied whether it was present for each pat tern size or if it was limited to one or two pattern sizes.

The committee suggests that, when an abnormality of VEPs is detected, simultaneous recording of P E R G and VEP be carried out to bet ter delineate the loca- tion of the abnormality. In case of demyelination of the

428

optic nerve, for instance, typically P E R G is normal whereas VEP is delayed.

(3) Pattern electroretinography (PERG)

Pattern electroretinography (PERG) is used to mon- itor the functional integrity of the proximal retina. Although there is no complete agreement on the gen- erators of P E R G it is agreed that it involves the proximal 30% of the retina corresponding to the gan- glion cell layer, the inner proximal layer and the inner nuclear layer 28. Most of the data in the literature suggest that the major contribution to P E R G is from ganglion cells and possibly from amacrine cells 2~-31.

These recommendations may require revisions every 5 years to keep abreast of the rapid changes in technol- ogy.

The recommendations are organized as follows: Basic technology:

Visual stimulation Calibration of visual stimuli Electrodes Recording equipment

Clinical protocol: Patient preparation Suggested specific protocols

(D Basic technology

Precise description of pattern stimulation is neces- sary in view of the great variety of possible patterns that may be used. Changes in the stimulus parameters may drastically affect the nature of PERG. It is incum- bent on users to demonstrate that changes in the stimuli do not alter or modify PERG. Comparison between results from different laboratories can only be done when utilizing identical or equivalent stimuli and recording procedures.

(1) Visual stimulation The committee recommends the use of checks, ver-

tical sine-wave or square-wave gratings. The recom- mendations described previously (pages 422-423) for the measurements of pattern stimuli in evoking VEPs are also valid for PERGs.

The pattern stimulus must have the following mea- surements defined: type of pattern, size of the elements of the pattern, total field size, method of presentation of the pattern, rate of presentation of the pattern, stimulus luminance, background luminance, and type of stimula- tor.

(2) Calibration of visual stimuli The recommendation for calibration of visual stim-

uli advocated for VEPs also applies for PERGs (see page 423).

IFCN COMMITTEE

(3) Electrodes The ideal electrode should be non-invasive, electri-

cally stable and should not modify the optic character- istic of the eye. None of the presently available elec- trodes meet all these criteria.

Two types of electrodes are acceptable for recording PERGs: corneal contact lens electrodes and conjuncti- val contact electrodes. Skin electrodes severely reduce the amplitude of the PERGs and should not be used.

Corneal contact lens electrode are identical to the one recommended for flash ERG by ISCEV (pages 425-431) and should therefore meet the same require- ments. The contact lens often modifies the eye refrac- tion, therefore the eye must be refracted after the corneal lens is in place and corrective lenses employed to restore normal visual acuity or restore vision to the baseline status during the recording session.

Conjunctival contact electrode consists of electrodes placed usually within the lower lid in direct contact with the conjunctiva. The prototype of these electrodes is the gold foil electrode 32. These electrodes do not interfere with the eye optics, thus, do not modify the patient refraction, however, their signal is less stable and noisier than that obtained with contact corneal lens electrodes 33.

The reference electrode is either incorporated in the contact lens-speculum of the corneal lens electrode (such as in the bipolar Burian-Allen electrode) thus making contact with the conjunctiva or is a skin elec- trode placed near the outer canthus of the correspond- ing eye. Reference placed at the ear lobe may record signals from the visual pathways and should not be used.

A separate skin ground electrode should be posi- tioned on the scalp. Suggested locations are: the fore- head, the vertex or the ear.

The skin electrode should be applied with collodion and have an impedance below 5000 ~(2.

Electrode cleaning. Contact electrodes will be ex- posed to tears. They must be properly cleaned to prevent transmission of infectious agents. The cleaning protocol should follow current practice standards for devices that are contaminated by body fluids.

Corneal contact electrodes may produce corneal abrasions if kept in position for a too long period of time. The committee recommends that the corneal elec- trode be kept in place no longer than 30 min.

(4) Recording equipment The recording system bandpass should be at least

0.3 Hz and 250-300 Hz ( - 3 dB) with roll-off slopes not exceeding 12 dB/oc tave for the low frequencies and 24 dB/oc tave for the high frequencies.

Other features of the recording system should fol- low the recommendations for VEPs described on page 424.

RECOMMENDED STANDARDS FOR ERGs AND VEI~s 429

(II) Clinical protocol

(1) Patient preparation Good cooperation from the patient is essential for

recording PERGs. Detailed explanation of the proce- dure to the patient will reduce anxiety and foster cooperation. To obtain the PERG, subjects must be able to resolve the pattern and thus corrective lenses must be worn to compensate for refractive errors. The pupils should not be dilated to avoid interference with accommodation. When corneal contact lenses are used, the eye must be refracted after the lens is in place, to assure the best corrected visual acuity.

The eye should be first anesthetized with an accept- able topical ophthalmic anesthetic before a contact electrode (whether corneal or conjunctival) is placed over the eye. A fixation point at the center of the stimulating field should be provided.

At the onset of the recording the pupil diameter and the visual acuity of each eye should be determined.

(2) Measurements of PERG Traditionally ERGs were recorded with the positive

response upward and the nomenclature used consisted of the letters of the alphabet from a to d. The commit- tee endorses the continuation of this tradition. The first small negative wave should therefore be named apt wave, followed by the bpt war, e and Cpt waL'e. The subscript "pt" indicates that the stimulus evoking ERG was a pattern. The subscript will prevent confusion with the flash ERG.

Both amplitude and latency (or "implicit time") should be measured for the Opt wave and the Cpt wave. Because the apt wave is often absent or small, it cannot reliably be measured.

NormatiL,e values. We recommend that each labo- ratory establish or confirm its own normal values. Nor- mative values obtained from other institutions may be utilized only if equivalent stimulation and recording methods are employed and only after testing the valid- ity of the adopted normal values on at least 10 locally gathered subjects. The measurements between the two groups should be similar.

The boundaries of normalities should be set at the 95-99% tolerance level for 95% of the normal popula- tion. A more detailed description of the statistics is reported under the section of VEPs on page 428.

(2) The clinical question being addressed should be provided in the report. The following information about the patient visual examination should be included in the report: (a) visual acuity with the corrective lenses used during the recording; if visual acuity is less than 20 /20 (6 /6) pin hole acuity is measured; (b) pupil size in millimeters at time of recording; (c) visual field defects; (d) ability to fixate; (e) status of the eye optics (i.e., presence or absence of cataracts, corneal opacities or other problems that may interfere with viewing the pattern).

(3) Technical data. The following stimulus parame- ters should be reported: type, size and contrast of the pattern, rate of presentation, mean luminance of the field. The report should state whether monocular or binocular stimulation was used. The type of ERG electrode and the montage used should be reported.

(4) Normative values. The laboratory normal values for amplitude and latency of the bpt and cpt waves should be reported. Normal values should include the limits of normal for the laboratory.

(5) Results. The report should include the ampli- tude and latency of the P ERG from each eye. A representative wave form of the responses should be provided containing calibration signals for time and amplitude.

(6) Interpretation. Interpretation should include a statement about the normality or abnormality of the test. The type of abnormalities should be described. Clinical correlations should be made with caution.

P E R G abnormalities are not diagnostic of a specific disease or lesion, thus clinical correlation statements should not imply that the results indicate a specific disorder. The correlation statement may contain word- ings indicating that the results provide support for a clinically suspected disease or process, but should con- tain the caveat that "do not exclude other possibilities."

(IV) Suggested specific protocols

The committee recommends that PERGs be recorded whenever pattern VEPs have been abnormal. Simultaneous recording of PERGs and VEPs is sug- gested with a 2-channel recording system. The commit- tee recommends the use of pattern-reversal checker- board or vertical gratings stimulation. The following criteria should be adhered to.

(IIl) Report ofresults

All reports should contain as a basic minimum infor- mation about: (1) patient; (2) clinical status; (3) techni- cal data; (4) normative values; (5) results; (6) interpre- tation.

(1) Patient information should include: name, age and gender and patient identification number.

(1) Recording The recommended system bandpass is 1.0 to 250-300

Hz ( - 3 dB) with filter roll-off slopes not exceeding 12 dB/oc tave for the low frequencies and 24 dB/oc tave for the high frequencies.

The following montage is suggested: Oz-Fpz; ERG electrode-reference. The reference should be located at 1-2 cm laterally from the outer canthus of the same

430 IFCN COMMITTEE

eye as the E R G electrode. If bipolar Burian-Allen electrodes are used, bipolar recording can be obtained between the contact corneal electrode and that part of the speculum that makes contact with the inner surface of the eyelid. The ground electrode should be placed at Cz.

Transient P E R G and VEP are recommended. The analysis time of 250 msec is recommended with averag- ing of 200 individual trials. A repeated trial to verify reproducibility of the results should always be ob- tained. The averaged responses of the two trials should be superimposed.

(2) Stimulation The committee recommends that a minimum of 3

stimuli be routinely presented. These should consist of alternating checks, alternat-

ing vertical sine-wave gratings, or alternating square- wave gratings. Because these stimuli produce different results, a given laboratory should select one of these alternatives. Irrespective of which of these 3 stimuli are used, the following parameters are suggested:

- Contrast: between 50 and 80%. - Full-field size: greater than 8 °. - Size of the individual pat tern elements: 14-16

min, 28-32 rain, 56-64 min. The smaller size of 14-16 ' is optimal to stimulate the fovea, but is very easily affected by visual acuity changes. The wider size patterns may also stimulate the parafoveal region thus may be normal in cases of foveal dysfunction.

- Rate of pat tern presentation: 1 Hz (producing a reversal every 500 msec).

- Mean luminance of the center field: at least 50 c d / m 2.

- B a c k g r o u n d luminance: 20-40 c d / m 2 under photopic conditions.

(3) Results The amplitude and latency of the bpt wave, the cpt

wave of PERGs and N70 and P100 of VEPs for each eye and for each stimulus size should be reported. The inter-peak interval between the bpt wave and the P100 should be calculated.

When an abnormality is described, the following questions should be answered: (1) is the abnormality localized to VEPs or does it include both P E R G and VEPs? (2) is the abnormality present for each pat tern size or is it limited to one or two pat tern sizes? (3) is the inter-peak interval normal or delayed?

The interpretation of the results may indicate a dysfunction of the proximal retina if the PERGs are abnormal, whereas a normal P E R G usually indicates that the abnormality involves the optic nerve or optic pathways. Correlation with the clinical data is recom- mended.

( 4 ) F u l l - f i e l d f l a s h e l e c t r o r e t i n o g r a p h y ( E R G )

Full-field flash electroretinography (occasionally re- ferred as f lash-ERG) is a widely used electrophysiolog- ical test to assess the functional integrity of the retina and specifically the status of the rods, cones, or both photoreceptor systems.

The committee endorses the standard of E R G es- tablished by the International Society for the Clinical Electrophysiology of Vision (ISCEV) in 198933 and reproduced here with permission.

"Full-field electroretinography (ERG) is a widely used ocular electrophysiologic test, and a basic proto- col should be standardized so that certain responses will be recorded comparably throughout the world. We propose standards for 5 commonly obtained responses:

(1) A maximal response developed in the dark- adapted eye. (2) A response developed by the rods (in the dark- adapted eye). (3) Oscillatory potentials. (4) A response developed by the cones. (5) Responses obtained to a rapidly repeated stimu- lus (flicker). While this document is intended as a guide to prac-

tice and will assist in interpretation of ERGs, we recognize that there are many additional techniques and protocols that certain laboratories may choose to use. The standard describes simple technical proce- dures that allow reproducible ERGs to be recorded under a few defined conditions. Different procedures can provide equivalent E R G responses. It is incumbent on users of alternative techniques to demonstrate that their procedures do in fact produce signals that are equivalent in basic wave form, amplitude, and physio- logic significance to the standard.

Our intention is that the standard method and re- sponses be used widely, but not to the exclusion of other responses or additional tests that individual labo- ratories may choose or continue to use. We also recog- nize that the investigation of certain eye conditions may not require all five of the standard responses. In addition, specialized types of E R G (e.g., focal ERG, early receptor potential, pattern ERG, bright-flash ERG, prolonged-flash ERG, and DC recording) are not covered by the standard.

Because of the rapid rate of change of E R G tech- niques, these standards will be reviewed every 4 years. We have made recommendations that commercial recording equipment have the capability to record ERGs under conditions that are outside the present standard but that are nevertheless either widely used or likely to be needed in the future. Note that this document is not a safety standard and does not man- date particular procedures for individual patients.

RECOMMENDED STANDARDS FOR ERGs AND VEPs 431

The organization of this report is as follows:

Basic technology Light diffusion Electrodes Light sources Light adjustment and calibration Electronic recording equipment

Clinical protocol Preparat ion of the patient E R G measurement and reporting

Description of the five standard responses Rod response Maximal response Oscillatory potential Single-flash cone response Flicker response

(I) Basic technology

(1) Light diffusion We believe strongly that full-field (Ganzfeld) dome

stimulation should be used. With focal flashes, the area of retinal illumination is not uniform, and its extent is unknown (although focal flashes may be used for cer- tain specialized E R G tests). Ocular diffusers (e.g., 100-diopter or opalescent contact lenses) are also less desirable, since there is no way to accurately measure the illumination produced by such devices.

(2) Electrodes Recording electrodes. Corneal contact lens elec-

trodes are recommended for basic full-field recording; they should have an optical opening as large as possi- ble and incorporate a device to hold the lids apart. The corneal surface should be protected during electrode use with a non-irritating and non-allergenic ionic con- ductive solution that is relatively non-viscous (e.g., ~< 0.5% methyl cellulose). Skin electrodes and other types of corneal electrodes (e.g., wicks, fibers, and foils) are less stable than corneal electrodes and may not be comparable for E R G amplitude and wave form mea- surements.

Reference electrodes. Reference electrodes may be incorporated into the contact lens-speculum assembly to make contact with the conjunctiva ("bipolar elec- trodes"). Otherwise, skin electrodes can be placed cen- trally on the forehead or near each orbital rim as a reference for the corresponding eye. Other positions may not be equivalent.

Ground electrodes. A separate skin electrode should be attached to an indifferent point and con- nected to ground. Typical locations are on the fore- head or ear.

Skin electrode characteristics. Skin electrodes used for reference or ground should have 10 kY2 or less

resistance measured at 30-200 Hz when applied. The skin should be prepared by cleaning and use of suitable conductive paste to ensure good electrical connection.

Electrode stability. Whatever corneal and reference electrode system is used, the baseline voltage in the absence of light stimulation should be stable. Some electrode systems may need to be made of non-polariz- able material to achieve this stability.

Electrode cleaning. Recording of E R G involves the exposure of corneal electrodes to tears and exposure of the skin electrodes to blood if there has been any abrasion of the skin surface. We advise that electrodes be suitably cleaned after each use to prevent transmis- sion of infectious agents. The cleaning protocol should follow current standards for devices that contact skin and tears.

(3) Light sources Stimulus duration. The standard is based on stim-

uli of duration considerably shorter than the integra- tion time of any photoreceptor. Thus, the light stimulus should consist of flashes having a maximum duration of 5 msec ~. Short-flash durations may be obtained from gas discharge tubes, from stroboscopes, and potentially from other devices.

Stimulus wat,elength. The stroboscopic flash tubes in use have a color tempera ture near 7000°K and should be used with domes or diffusers that are visibly white. Colored filters are used by some laboratories to enhance the separation of rod and cone responses, but this is not part of the standard 2

Stimulus strength. A standard system is defined as one that produces a stimulus strength (in luminance time) at the surface of the Ganzfeld bowl of at least 1.5-3.0 c d / m 2 / s e c (note that these are photometric units and that 3.43 c d / m 2= 1 fL) 3. A flash of this

1 Prolonged flash ERGs are currently used for studying slow poten- tials and off-responses that are outside the scope of the standard. We recognize that one can adjust the intensities of long flashes to produce response amplitudes equivalent to those produced by brief flashes; thus, standard ERG responses can be obtained from such longer stimuli. However, this procedure requires careful comparison of " V / l o g I" curves and particular care to avoid interference from off-responses and signal attenuation by light adaptation (i.e., the interstimulus interval must be appropriately lengthened). The verifi- cation of equivalence to the standard ERG is recommended only for laboratories with special needs. z Chromatic stimuli offer certain advantages in the separation of cone and rod responses, but the calibration of colored stimuli and the relation of the responses produced to the standard ERG require special procedures. We recommend that who flashes be used for the standard responses in addition to other stimuli that may be used. 3 This measurement can in practice be made with inexpensive light- meters that integrate the flash output overtime (see 'Light Adjust- ment and Calibration' section). Technically, the standard describes a source that delivers at the cornea the same number of quanta during the period of its flash as would be produced in 1 sec by the Ganzfeld bowl when continuously illuminated by a source that produces a luminance of 1.5-3.0 c d / m 2.

432 IFCN C O M M I T T E E

strength will be called the standard flash (SF). Background illumination. In addition to producing

flashes, the stimulator must be capable of producing a steady and even background luminance of at least 17-34 c d / m 2 (5-10 fL) across the full field. For this standard, a white background is used, but we recognize that colored backgrounds may also be used.

(4) Light adjustment and calibration Adjustment of stimulus and background intensity.

Methods of modifying both the stimulus and back- ground intensity must be provided. We recommend that a standard system be capable of attenuating flash strength from the SF over a range of at least 3 log units, either continuously or in steps of no more than 0.25 log unit. The method of attenuation should not change the wave length composition of the light. The background luminance needs sufficient adjustability to calibrate and recalibrate the intensity to the levels specified below under "Single-Flash Cone Response." It is preferable that the color temperature of the background should not alter with intensity. We recog- nize that the stimulus and background requirements for a full range of E R G testing are both more extensive and more stringent, and we recommend that equip- ment manufacturers exceed the minimum standard 4.

Stimulus and background calibration. The stimulus strength (in luminance time) produced by each flash on the surface of the Ganzfeld bowl must be documented by the user or manufacturer, ideally with an integrating photometer (luminance meter) placed at the location of the eye. The light output per flash of most strobo- scopes varies with the flash repetition rate; therefore, separate calibrations will need to be made for single and repetitive stimuli. The photometer must record the luminance of the Ganzfeld bowl's surface, must meet international standards for photometric measurements based on the photopic luminous efficiency function (photopic luminosity curve), and must be capable of recording the total output of very brief flashes. The committee recommends that in the future, manufactur- ers of stimulators provide a suitable photometer as part of the equipment. Background luminance may be measured with the same instrument, in the non-in- tegrating mode.

4 We recommend that the flash source of commercial ins t ruments be capable of generating strengths 1 log unil~ above the SF and be at tenuable through 6 log units below the SF. Regardless of whether at tenuation is achieved by filters or electronic means, we strongly recommend that commercial units incorporate a means of inserting additional colored and neutral density filters to meet a variety of individual (and unforeseen) needs. We also suggest that background luminance be adjustable to perform electro-oculography with the same equipment. Commercial units should also allow the insertion of colored and neutral filters into the background illumination system to meet a variety of needs.

Recalibration. Light output from the dome may vary with time from changes in the flash tube, the tube power source, the background light bulbs, the attenua- tion systems, or the paint in the dome. This may be especially critical for background illumination provided by incandescent sources. Responsibility for electronic stability and warnings about sources of instability should rest with the manufacturers of the equipment; however, at present this cannot be presumed. The frequency with which recalibration of flashes and back- grounds is required will vary from system to system and could be as high as weekly for some units. Self-calibrat- ing units are to be encouraged.

(5) Electronic recording equipment Amplification and display systems. We recommend

that the bandpasses of the amplifier and preamplifiers include the range of 0.3-300 Hz and be adjustable for oscillatory potential recordings and special require- ments. We advise that the input impedance of the preamplifiers be at least 1 MS2. Amplifiers should generally be AC coupled and capable of handling off- set potentials that may be produced by the electrodes 5

Display system. We strongly recommend that the equipment that provides the final record be able to represent, without attenuation, the full amplifier band- pass. Good resolution can be achieved with oscillo- scopes or computer-aided systems but not with direct pen recorders. With the computer-aided systems it is important that responses be displayed promptly so that the operator can continuously monitor stability and make adjustments during the test procedure.

Patient isolation. We believe that the amplifiers must be electrically isolated from the patient, accord- ing to current standards for safety of biologic recording systems used clinically.

(H) Clinical protocol

(1) Preparation of the patient Pupillary dilatation. We recommend that pupils be

maximally dilated for all E R G recordings in this stan- dard and that pupil size be noted when dilatation is, for any reason, less than maximal.

Initial dark adaptation. Dark adaptation for at least 20 min is required to achieve a relatively stable physio- logic condition and relatively maximal scotopic re- sponses. The ERG recording electrodes can be in- serted under dim red light at the end of this period to minimize corneal irritation from the electrodes.

Pre-exposure to light. We advise that fluorescein angiography or fundus photography be avoided before

5 Direct current amplification can produce identical responses but is extremely difficult to use because of drift in baseline and in offset potentials; we strongly advise AC recording except for laboratories with special requirements and expertise.

RECOMMENDED STANDARDS FOR ERGs AND VEPs 433

E R G testing, but if these examinations have been performed, a period of dark adaptation of 1 h is needed. It is usually preferable to record scotopic responses to weak flashes before the mixed and cone responses to more intense flashes, to minimize light adaptation, and to reduce the time that the patient wears the contact lens electrode.

Fixation. A fixation point is useful but not essen- tial. Some patients will not be able to see it, and the Ganzfeld dome minimizes the need for accurate fixa- tion. In the absence of fixation, patients can be in- structed to look straight ahead and keep their eyes steady.

(2) ERG measurements and recording Measurement of the ERG. Both amplitude and im-

plicit time should be measured for selected E R G sig- nals. For practical purposes the parameters most often measured are the cone, the rod, and the maximal b wave amplitudes and the cone or flicker b wave time- to-peak. According to current convention, the a wave amplitude is measured from baseline to a wave trough, the b wave amplitude from a wave trough to b wave peak, and the b wave t ime-to-peak from flash onset to the peak of the wave.

Normal ualues. We recommend that each labora- tory establish or confirm normal values for its own equipment and patient population and that all E R G reporting (whether for local records, publication, or even for non-standard responses) include normal val- ues and the limits of normal. Some manufacturers may choose to distribute norms for their standard protocols. An effort is under way to establish worldwide norms.

Reporting the ERG. Standardization of E R G re- porting is critical to the goal of having comparable data worldwide. We recommend that reports or communica- tions of E R G data include a representative wave form of each of the standard responses (if performed) dis- played with amplitude and time calibrations and la- beled with respect to stimulus variables and the state of light or dark adaptation. The strength of stimulation ( c d / m 2 / s e c ) and light adaptat ion ( c d / m 2) should be given in absolute values. The reporting forms should indicate whether the recordings meet the international standard. We recommend that the basic numerical measurements listed above be extracted from the data and listed along with the normal values and their variances (that which must be provided on all reports).

(III) Specific responses

(1) 'Rod' response We recommend that the rod response be the first

signal measured after dark adaptation, since it is the most sensitive to light adaptation. The standard stimu- lus is a dim white flash of strength 2.5 log units below the white SF (see above); we advise a minimum inter-

val of 2 sec between flashes. Blue is equally appropri- ate if equated to the white standard z

(2) Maximal response The maximal response is to be produced by the

white SF, in the dark-adapted eye. We recommend an interval of at least 5 sec between stimuli. This response is normally produced by a combination of cone and rod systems.

(3) Oscillatory potentials Oscillatory potentials are obtained from the dark-

adapted eye, using the same white SF. However, the high-pass filter must be reset to 75-100 Hz, so that an overall bandpass of 75-100 Hz on the low end and 300 Hz or above at the high end is achieved. Filters should at tenuate sufficiently to achieve this result. Users should be aware of and test for artifacts (e.g., phase shifts or ringing) that may be produced by present-day filters. The oscillatory potential response varies with stimulus repetition rate and changes after the first stimulus. To standardize the response, we recommend that flashes be given 15 sec apart and that only the second or subsequent responses be retained or aver- aged.

(4) Single-flash 'cone' response We propose the white SF as the stimulus and advise

that to achieve stable and reproducible cone responses, the rods be suppressed by a background with a lumi- nance of 17-34 c d / m 2 (5-10 fL) measured at the surface of the Ganzfeld bowl. We recommend that the higher value of the background be chosen if the stimu- lus flash is at the upper end of the allowable SF range and the lower background value chosen if the flash stimulus is at the lower end of the range. The intention is that all standard E R G recording systems have an identical intensity ratio between the SF and the rod- suppressing background, equivalent numerically to 3.0 ( c d / m 2 / s e c ) divided by 34 (cd/m2). We recommend that patients light adapt to the background luminance for 10 min before recording the cone ERG, since the cone responses may increase during this period.

(5) Flicker response Flicker responses are to be obtained with SF stimuli,

under the same rod-suppressing background illumina- tion, after recording the single-flash cone response. Recording the flicker response in the light-adapted state reduces discomfort and allows the photopic adap- tation to be standardized. We advise strongly that flashes be presented at a rate of 30 st imuli /sec, and that the first few responses should be discarded so that stable conditions are reached. Some flash tubes do not produce full output while flickering, and separate cali- bration or a change in neutral density, filtering may be needed to keep as closely as possible to the standard."

434 1FCN C O M M I T T E E

Quantitative standardized E R G is an important tool in the study of genetic retinopathy and may be used to identify the carrier of the gene responsible for the retinal dysfunction 35-3v.

Other specialized types of E R G such as: (a) focal ERG; (b) early receptor potentials, bright flash ERGs and DC E R G recordings are sometimes used in study- ing retinal physiology. These procedures are still inves- tigational and are not covered by the proposed stan- dards.

(5) Visual evoked potentials in pediatrics

Visual evoked potentials in children vary in relation to age 3s-4°. P100 to pattern VEP is consistently pres- ent at all ages, with longer latency at birth and gradu- ally reaching adult values. The latency of P100 to small checks reaches adult values by age 539 . It is recom- mended that the same protocol utilized for adults be used for children 5 years or older.

For children under age 5 and for infants a specific protocol is suggested. The requirements for measure- ments of visual stimulation and for recordings of VEPs advocated for the adults apply to children. However, it is imperative that the infant and child fixate on the stimulus thus great care should be involved in the testing.

(1) Pediatric protocol

It is suggested that infant and children below age 5 be first tested with pattern reversal or pat tern onset- offset and only if no responses are obtained with this type of stimulation then the protocol changed to the use of flashes. Whereas the use of pattern stimuli may permit the estimation of the infant visual acuity, the use of flashes will only determine the presence or absence of light perception.

(1) Pattern VEPs Two testers should be available when recording pat-

tern VEPs; one operates the equipment, the other monitors the patient 's fixation. The observer monitor- ing the patient 's fixation can stand behind the pat tern generator and watch for a centered pupillary reflex. The observer should have a remote-control switch to stop the averager when the patient is not fixating. The test room should be separate from the room with the recording equipment and should be dark; the infant or child should only be able to see the screen.

(a) Recording. The recommended system bandpass is 1.0 to 250-300 Hz ( - 3 dB) with filter roll-off slopes not exceeding 12 dB/oc t ave for the low frequencies and 24 dB/oc tave for the high frequencies. A montage

consisting of 2 derivations is sufficient for recording VEPs.

VEPs are widely distributed over the scalp from the vertex to the inion. The following montage is sug- gested: Oz-Fpz ; Oz-A1-A2 (linked ears). The ground electrode should be placed at Cz.

An analysis time of 250 msec is recommended with averaging of 100 individual trials.

(b) Stimulation. The committee recommends that a minimum of 5 stimuli be routinely tested.

To speed the procedure it is recommended that initially a check size appropriate for the infant's age be selected first, and larger and smaller checks be used depending on the outcome. It is often possible to obtain a check size series and estimate VEP acuity. At least 5 check sizes should be used to fit a reliable curve, for example, 7.5, 30, 60, 120, 240 rain. For infants, an intermediate (60 rain) check size is used first. If a response is obtained, a smaller (30 rain) check is used; if no signal is obtained, a larger (120 rain) check size is used.

The following parameters are suggested for either alternating or onset-offset checks. Check pattern rever- sal or onset-offset gives responses with different char- acteristics, thus a given laboratory should select one of the two alternatives.

Regardless of this choice, the following stimulus parameters should be used: contrast between 50 and 80%; full-field size: greater than 8°; size of the pattern elements: as outlined above; rate of presentation of the pattern: random when infant or child is fixating but not faster than every 300 msec: mean luminance of the center field should be at least 100 cd /m2; background luminance. The room should be dark to prevent visual distraction and difficulty with fixation, the infant should only be able to see the screen.

Patients are first tested binocularly, starting with 15 min checks. If necessary, check size is increased until a reliable binocular signal is obtained. After a reliable signal has been obtained, each eye is tested, preferably the eye thought to have better acuity first.

(c) Results. The amplitude and latency of the N70 and P100 of VEPs for each eye and for each stimulus size should be reported.

Since the amplitude and peak latency of specific components of the pat tern VEP change dramatically during the first 6 months of life, normative data should be obtained for each month of life up to 6 months. At 1 month, checks of 120-140 rain give optimal re- sponses; at 2 months, 60-120 rain; at 3-5 months, 30-60 rain; and at 6 months and older, 15-20 rain 3s 4o. When an abnormality is reported, it should be speci- fied whether it was present for each pattern size or if it was limited to one or two pattern sizes.

It may be possible if a curve of the amplitude responses to 5 stimuli is available to estimate the

R E C O M M E N D E D S T A N D A R D S FOR ERGs AND VEPs 435

infant's visual acuity. However, this technique is often too time-consuming for clinical use. An alternative is to measure the difference in amplitude and latency be- tween eyes for one check size. Although a visual acuity value cannot be estimated, useful information about the function of each eye may be obtained.

(2) Flash VEPs When VEPs are not detected with pattern stimula-

tion, flash stimuli should be used. A full-field (Ganzfeld) dome stimulation or light

diffusers that are visibly white should be used. Adher- ence to meticulous standardization of the stimulus is recommended. The basic technology for flash stimula- tion described for ERGs similarly applies here. White flashes should be utilized, with the frequency of flashes at 1, 3, 6, 10 and 20 f lashes/sec. There is wide variabil- ity in the amplitude and latency of transient VEPs to flashes in infants. Therefore, rather than using ampli- tude and latency data to transient VEPs, it is prefer- able to determine if the response follows the increasing stimulus rate. If it does, it is assumed that the patient has vision of at least light perception.

(3) ERG ERG in children aged 6 and older can be obtained

with the same methodology used for adults. Below age 6 obtaining ERGs is very difficult; it may not be prudent to try to use contact corneal lens electrodes because of the danger of corneal abrasions and eye trauma. Skin electrodes may be placed in the periorbic- ular area. However, periorbicular electrodes signifi- cantly reduce the amplitude of the ERG and give variable responses. Interpretation of ERGs from peri- orbicular electrodes is therefore fraught with potential interpretation errors and need to be used with consid- erable caution.

References

1 Regan, D. Some characteristics of average steady-state and tran- sient responses evoked by modulated light. Electroenceph. clin. Neurophysiol., 1966, 20: 238-248.

2 Regan, D. Human Brain Electrophysiology. Evoked Potentials and Evoked Magnetic Fields in Science and Medicine. Elsevier, Amsterdam, 1989:672 pp.

3 Celesia, G.G., Polcyn, R.E., Holden, J.E., Nickles, R.J. and Gatley, J.S. Visual evoked potentials and PET mapping: can the neuronal potential generators be visualized? Electroenceph. clin. Neuophysiol., 1982, 54: 243-256.

4 Darcey, T.M., Ary, J.P. and Fender, D.H. Spatio-temporal visu- ally evoked scalp potentials in response to partial-field pat terned stimuli. Electroenceph. clin. Neurophysiol., 1980, 50: 348-355.

5 Van Dijk, B.W. and Spekreijse H. Localization of electric and magnetic sources of brain activity. In: J.E. Desmedt (Ed.), Visual Evoked Potentials. Elsevier, Amsterdam, 1990: 57-74.

6 Harding, G.F.A. The visual evoked potential in neuro-ophthalmic disorders. In: J.E. Desmedt (Ed.), Visual Evoked Potentials. Elsevier, Amsterdam, 1990: 147-167.

7 De Valois, R.L. and De Valois, K.K. Spatial Vision. Oxfl)rd University Press, Oxford, 1988:381 pp.

8 Camisa, J., Mylin, L.H. and Bodis-Wollner, 1. The effect of st imulus orientation on the visual evoked potential in multiple sclerosis. Ann. Neurol., 1981, 10: 532-539.

9 Steinmetz, H., Furst, G. and Meyer. B. Craniocerebral topogra- phy within the international 10-20 system. Electroenceph. clin. Neurophysiol., 1989, 72: 499-506.

10 Stensaas, S.S., Edomgton, D.K. and Dobelle. W.tt . The topogra- phy and variability of the primary visual cortex in man. J. Neuro- surg., 1974, 40: 747-755.

11 Kaplan, E., Lee, B.R. and Shapley, R.M. New views of primate retinal function. In: N. Osborne and G. Chader (Eds.), Progress in Retinal Research, Vol. 9. Pergamon Press. Oxford, 19~)0: 273-336.

12 Bodis-Wollner, I., Brannan, JR . , Ghilardi, M.F. and Mylin. L.H. The importance of physiology to visual evoked potentials. In: J.E. Desmedt (Ed.), Visual Evoked Potentials. Elsevier, Amsterdam, 1990: 1-24.

13 Celesia, G.G. and Brigell, M.G. Pattern visual stimulation in prechiasmatic lesions, in: J.E. Desmedt (Ed.), Visual Evoked Potentials. Elsevier. Amsterdam, 1990:75 85.

14 Spekreijse, B. and Apkarian. P. The use of a system analysis approach to ERG and VEP assessment. Vision Res.. 1986. 26: 195 219.

15 Bodis-Woltner, I., Hendley, C.D., Mylin, LH. and Thornton, J. Visual evoked potentials and the visuogram in multiple sclerosis. Ann. Neurol., 1979, 5: 40-47.

16 Bodis-Wollner, I., Bobak, P., Harnois. C. amt Mylin, L. Method- ological aspects of signal-to-noise evaluation of s imultaneous electroretinogram and visual evoked potential recordings. Doc. Ophthalmol. Proc. Ser., 1984, 40:30 48.

17 Regan, D. Steady state evoked potentials. J. Opt. Soc. Am., 1977, 67: 1475-1489.

18 Spekreijse, H., Estevez, O. and Reits, D. Visual evoked potentials and the physiological analysis of visual processes in man. In: J.E. Desmedt (Ed.), Visual Evoked Potentials in Man. New Develop- ments. Clarendon Press, Oxford, t977:16 89.

19 Strasburger, H. The analysis of steady state evoked potentials revisited. Clin. Vision Sci., 1987, 3:245 256.

20 Tobimatsu, S., Tashima-Kurita, S., Nakayama-Hiromatsu, M. and Kato, M. Clinical relevance of phase of steady-state VEPs to PI00 latency of transient VEPs. Electroenceph. clin. Neurophys- iol., 1991, 80: 89-93.

21 Tomoda, H., Celesia, G.G. and Cone Toleikis, S. Effect of spatial frequency on s imultaneous recorded steady-state pattern elec- troret inograms and visual evoked potentials. Electroenceph. ctin. Neurophysiol., 1991, 80: 81-88.

22 Kirk, R.E. Experimental Design: Procedures for the Behavioral Sciences. Brooks/Cole , New York, 1982: 91] pp.

23 Rosner, B. Statistical methods in ophthalmology: an adjustment for the intraclass correlation between eyes. Biometrics, 1982, 38: 105-114.

24 Cohen, J. and Cohen, P. Applied Multiple Regress ion / Correlation Analysis for the Behavioral Sciences. John Wiley and Sons, New York, 1973:490 pp.

25 Mardia, K.V. Statistics of Directional Data. Academic Press, London, 1972:357 pp.

26 Zar, J.H. Biostatistical Analysis. Prentice-Hall, Englewood Cliffs, NJ, 1974: 310-328.

27 Celesia, G.G. Visual evoked potentials in clinical neurology. In: M.J. Aminoff (Ed.), Electrodiagnosis in Clinical Neurology. Churchill Livingstone, New York, 1992: 467-489.

28 Zrenner , E. The physiological basis of the pattern electroretino- gram. In: N. Osborne and G. Chader (Eds.), Retinal Research, Vol. 9. Pergamon Press, Oxford, 1990: 427-464.

29 Celesia, G.G. and Kaufmann, D. Pattern ERGs and VEPs in

436 IFCN COMMITTEE

maculopathies and optic nerve diseases. Invest. Ophthalmol. Vis. Sci., 1985.26: 726-735.

30 Maffei, L. and Fiorentini, A. ERG responses to alternating gratings in the cat. Exp. Brain Res., 1982, 48: 327-334.

31 Tobimatsu, S., Celesia, G.G., Cone, S.B. and Gujurati, M. Elec- troretinograms to checkerboard pattern reversal in cats; physio- logical characteristics and effect of retrograde degeneration of ganglion cells. Electroenceph. clin. Neurophysiol., 1989, 73: 344- 352.

32 Arden, G.B., Carter, R.M., Hogg, C. et al. A gold foil electrode: extending the horizons for clinical electroretinography. Invest. Ophthalmol. Vis. Sci., 1979, 18: 421-426.

33 Prager, T.C., Saad, N., Schweitzer, C., Garcia, C.A. and Arden, G.B. Electrode comparison in pattern electroretinography. In- vest. Ophthalmol. Vis. Sci., 1992, 33: 390-394.

34 Marmor, M.F., Arden, G.B., Nilsson, S.E.G. and Zrenner, E. An international standard for electroretinography. Doc. Ophthalmol., 1990, 73: 299-302.

35 Berson, E.L., Rosen, J.B. and Simonoff, E.A. Electroretino- graphic testing as an aid in detection of carrier of X-chro-

mosome-linked retinitis pigmentosa. Am. J. Ophthalmol., 1979, 87: 460-468.

36 Berson, E.L., Sandberg, M.A., Maguire, A., Bromley, W.C. and Roderick, T.M. Electroretinograms in carriers of blue cone monochromatism. Am. J. Ophthalmol., 1986, 102: 254-261.

37 Pagon, R.A., Chatrian, G.E., Hamer, R.D. and Lindberg, K.A. Heterozygote detection in X-linked recessive incomplete achro- matopsia. Ophthal. Paediat. Genet., 1988, 9: 43-56.

38 Sokol, S., Moskowitz, A. and Towle, V.L. Age related changes in the latency of visual evoked potentials: influence of check size. Electroenceph. clin. Neurophysiol., 1981, 51: 559-562.

39 Moskowitz, A. and Sokot, S. Developmental changes in the human visual system as reflected by the latency of the pattern reversal VEP. Electroenceph. clin. Neurophysiol., 1983, 56: 1-15.

40 Sokol, S. The visual evoked cortical potential (VECP) in optic nerve and visual pathway disorders. In: G.A. Fishman and S. Sokol (Eds.), The ERG, EOG and VECP in Retinal, Optic Nerve and Visual Pathway Disorders. American Academy of Ophthal- mology, San Francisco, CA, 1991: 105-141.