Electrodiagnostic Evaluation of Visual Function

Michael H. Sims, PhD

Behavioral Responses to Visual Stimulation
When electromagnetic waves in the range of visable light enter the eye, the result is an alteration of electrical potentials in millions of retinal cells. These electrical changes are propagated through several layers of the retina and then on to the visual cortex via the optic nerve, optic tract, and thalamus. Once in the visual cortex, these impulses recreate the visual scene, and with the involvement of selected visual association areas result in conscious perception. An assessment of vision in animals without the use of verbal communication is difficult at best. This is further complicated by an animal's ability to adapt to lose of vision, even to the extent that its owner may be unaware of its disability. An animal's adaptation to partial or complete loss of vision is species dependent and related to its ability to compensate by increased reliance on auditory and tactile sensations. What an animal can see, i.e., its conscious awareness of visual stimulation, is not always the immediate focus of a clinical assessment. The goal is sometimes predicting what the animal "should" be able to see given the functional status of component parts of the visual pathway. An animal's response to menace testing, maze testing, and the cotton-ball test as described elsewhere are integral parts of the visual exam. The results of these procedures and the owner's history may still be the best estimate of what the animal can actually see.

Electrodiagnostic Evaluation of Vision
Unlike behavioral evaluations of vision, electrodiagnostic assessments are based on objective analyses of changes in electrical potentials in various parts of the visual system. Therefore, none of the electrodiagnostic procedures should be thought of as being measures of vision per se. The primary pathway for conscious vision consists of receptor cells (rods and cones), bipolar cells, ganglion cells, and cells in the lateral geniculate body, and the visual cortex. The receptor and bipolar cells are neurons that are totally contained within the retina whereas the axons of the ganglion cells exit the eye and comprise the optic nerve and optic tract. Each of these components and others are directly or indirectly evaluated by electrodiagnostic tests. The responses obtained, however, are dependent upon the type of visual stimulation, the condition of light adaptation, type and frequency bandwidth of the amplifier, the time window for display, and the use of signal averaging. Multichannel recorders even allow different types of responses to be recorded simultaneously as long as the time window is the same.

This chapter is concerned with a general description of electrodiagnostic procedures and their application in clinical visual testing. Some of the procedures have not been proven to be as useful in Veterinary Medicine as others. The electroretinogram (ERG), for example, is still the most widely reported procedure in the veterinary literature. It is used quite often for the diagnosis of retinal diseases such as progressive retinal atrophy (PRA) and for an evaluation of general retinal function prior to cataract surgery. Among the other procedures that are available are the visual evoked potential (VEP), the pattern electroretinogram (PERG), and oscillatory potentials (OP). These potentials have received less attention than the ERG, but certainly could offer additional information about retinal function in animals that might otherwise be missing from clinical evaluations. The overwhelming majority of the veterinary literature in clinical electrodiagnosis is concerned with dogs and less to cats, large animals and exotic animals.

Electroretinography
As early as the middle of the nineteenth century, researchers showed that the vertebrate eye responded to light with electrical changes and that the retina was the source of these new potentials. The ERG is an intricate and complex response initiated by light-sensitive cells in the retina. The ERG is both multiphasic and transient but unfortunately not all of its subcomponents are apparent with a single stimulus protocol. Although the precise nature of the ERG continues to unfold with further research, it provides a unique assessment of receptor and early central nervous system components. The unique aspect of the ERG is that it assesses function in structures that can also be directly visualized with an ophthalmoscope, a privilege not possible with other electrodiagnostic techniques. In dogs, the ERG has been useful in the diagnosis or evaluation of retinal function in cases of cataracts, glaucoma, hemeralopia, retinal dysplasia, toxicology screening, degenerative retinopathies, optic nerve hypoplasia, sudden acquired retinal degeneration, or cortical blindness.2-13 The ERG can be used in cats to diagnose diseases of the retina such as hereditary retinal degeneration14, non-inflammatory retinophathy15, and central retinal degeneration due to dietary taurine deficiency16. For some retinal diseases, the ERG provides a much earlier diagnosis than an ophthalmoscopic or behavioral exam and even provides a characterization of the function of specific cell types in the retina.

In its simplest form, (Fig. 1) the typical vertebrate ERG consists of a early small corneal negative potential that occurs within 10-12 msec after a bright light stimulus. The cornea then becomes more intensely positive only to be followed by a slower negative trough. After a period of several hundred milliseconds, the cornea once again becomes positive. The one negative and two positive peaks were named the a-, b-, and c-, waves, respectively, by Eithoven and Jolly in 1908.17 The A, B, and C waves of the ERG, which originally referred to "substances" in the eye and the potentials they produced, are still used to refer to the three peaks of the ERG.18

A typical canine ERG, which was recorded by a contact lens electrode referenced to a needle electrode inserted subcutaneously just posterior to the lateral canthus, is shown in Figure 2. The stimulus was a white stroboscopic flash delivered to a dog that had been dark-adapted for approximately 5 minutes. The same stimulus presented to a light-adapted dog is shown in Figure 3.

Unless special provisions are made in the recording technique, the ERG can be thought of as various combinations of rod and cone responses to light. The a-wave is unquestionably caused by photoreceptor potentials. Although the origin of the b-wave may be open to some question, it is generally accepted that it is generated by Muller cells. Current source density analysis and intracellular recording electrodes support the hypothsis of radial current flow through Muller cells. Exactly how Muller cells produce the b-wave is explained by the movement of potassium into and out of certain portions of the cell. It seems that light stimulation of the retina results in an increase in extracellular potassium in both the inner and outer plexiform layers which then promotes an influx of potassium into the portions of the Muller cells in those layers. The potassium influx results in depolarization of the cells and a radial current flow within the cell. The return current through the extracellular fluid, especially in the outer layers, provides much of the b-wave. Apparently, depolarizing bipolar cells are the source for the potassium in the outer layers. These currents promote a vitreal-positive potential that accounts for the polarity of the b-wave. Thus, the b-wave is directly related to Muller cell activity and indirectly related to activity in bipolar cells. It is interesting to note that occlusion of the central retinal artery abolishes the b-wave but spares the photo receptor response.

In most protocols used in veterinary ophthalmology, the c-wave (also called the late positive potential or PI) is not apparent. The c-wave has been defined as the positive potential after the b-wave with a latency or implicit time greater than 500 msec.19 The c-wave may be studied by using longer and more intense stimuli and a DC-coupled amplifier (bandwidth = DC-1000 Hz). The c-wave is unusual because its implicit time increases as the stimulus intensity increases, and because it only occurs in about one-third of adult dogs.19 The c-wave has not been reported to be characteristically altered in disease states in most veterinary publications.

One might reasonably conceptualize the ERG as a number of different physiologic potentials involving different cell types, but sharing the same time domain. Granit's graphic separation of the "retinal action potential" into the PIII, PII, and PI processes illustrated the relationships of the c-, b-, and a-waves, respectively, in time.20-21 This analysis continues to command both historic and fundamental physiologic interests. Granit's analysis was based on a variety of experiments, one of which involved the effects of ether on the waveform of the ERG. For both a bright stimulus and a dim stimulus, the ERG was a composite of three independent waveforms referred to as PI, PII, and PIII. These potentials roughly correspond to the a-, b-, and c-wave, respectively. The order of disappearance in response to ether administration was from PI to PII. Once PIII disappeared, the ERG was incapable of recovery.

Light that enters the eye may be light that participates in the formation of an image, or it may be stray light.22 This is important to remember because the amplitude of the ERG may or may not relate to the visual image quite the way that some may think. One might mistakenly assume that an ERG with a large amplitude is a "good" ERG in that it is evidence of good retinal function. While this is generally true, an ERG elicited by an intense stroboscopic flash may be good because many cells in the retina are being stimulated via stray light and not because there is a good image being formed on the retina. The physiologic difference between stray light and image light may actually complicate the recorded response because these two types of light cause additional peaks as a result of differences in their respective latencies.

The clinical ERG is usually reported by providing values for the implicit times (or latencies) and amplitudes for the a- and b-waves. In the broadest sense, a latency is a period of time between any two events of interest. However, the terms implicit time have been used to refer to the time between stimulus onset and the peak of a wave of interest. The amplitude of the a-wave is measured from the prestimulus baseline to the nadir of the peak and the b-wave is measured from the nadir of the a-wave to the peak of the b-wave. The effects of retinal lesions on the amplitude or implicit time of the ERG are somewhat predictable. A retinopathy, which involves large areas of the retina will result in decreased amplitudes and prolonged implicit times, whereas focal lesions may result in decreased amplitudes, but the implicit times are not altered.23

Additional activities in the retina may be recorded during the same period of time that is occupied by the ERG. These include a differentiation of the a- and b-waves into photopic and scotopic components, the early receptor component (ERP), the scotopic response threshold (STR), oscillatory potentials (OP), the late negative potential, and the d-wave. Some of these activities are apparent when the visual stimulus is presented in a special way, and others have more to do with how the activity is recorded or the preparation and/or the light adaptation of the animal. In animals such as dogs that have high concentrations of rods in their retinas, the c-wave may show a negative off deflection if the stimulus has a long duration.

Recording the Flash Electroretinogram.
The flash ERG, or FERG, is a mass response of the retina to full-field stimulation of the retina with a flash of light. Like other modern electrodiagnostic procedures, electroretinography can involve advanced techniques and sophisticated recording equipment that one might expect to find in a specialty practice or veterinary teaching hospital, or it can be a very simple procedure that can be recorded with uncomplicated instruments. A simple strobe light capable of triggering an oscilloscope equipped with good physiologic amplifiers that can provide low-noise amplification in the range of 104 will provide valuable information about visual function. If the clinician is attempting to assess the integrity of the retina prior to cataract surgery, an ERG can be reported as pass/fail based on previously established criteria. However, if there is an attempt to gather objective data for diagnosing specific retinal diseases, or for following the course of a disease through serial examinations, then the protocol and the equipment should be more sophisticated.24

Recording Equipment
Electrodes. The recording electrode may be one of several different types including contact lens, cotton wick, gold foil, and subdermal needle electrodes. The primary considerations in the selection of electrodes are accuracy and reliability, cost, potential of trauma to the patient, fit, and ease of use. Most commercially available electrodes are made of silver, gold, or platnium, but they can also be constructed from 38 AWG teflon-insulated stainless steel wire25 or silver-impregnated microfibers.26 Contact lens electrodes are probably used more often than any other because they are non-traumatic and provide reliable and reproducable results. With contact lens electrodes, an ionic medium such as hydroxyproply methylcellulose solution, is used to insure good electrical contact between the electrode and the cornea. Needle electrodes have been used in animals, but ERGs recorded with contact lens electrodes have significantly greater amplitudes than those recorded with subdermal electrodes.27-28

Amplifier. Retinal potentials are small (usually 1 mV or less) and in order for them to be properly visualized, a system of preamplifiers and amplifiers are necessary. The three inputs (positive, negative and ground) are connected to a high quality differential amplifier so that a positive potential occuring at the active electrode results in an upward deflection in the response. The amplifier should have band-pass filters that allows an attenuation of frequencies below and above the low and high settings, respectively. The bandwidth of the amplifier may vary, depending on the purpose of the procedure. For ERGs, some have used a bandwidth as narrow as 0.8-250 Hz29, While others have used wider bandwidths such as 1-3000 Hz9. Using a bandwidth of DC-1000 Hz will permit a recording of the c-wave19. The choice of bandwidths is dependent upon which visual potential is being analyzed. Suggested bandwidths for visual potentials are as follows: electro-oculogram (dc-100 Hz); ERG a- and b-waves (0.2-200 Hz); oscillatory potentials (90-300 Hz); ERG c-wave (dc-200 Hz); pattern ERG (0.1-200 Hz); visual evoked potentials (1-300 Hz).30

Many of the commercially available electrodiagnostic systems include analog-to-digital converters early in the stages of amplification. A response converted to a digital format then allows for ease of storage and subsequent analysis with a variety of computer programs.

Stimulator. A photostimulator such as a xenon flash tube provides short-duration flashes of light to the light- or dark-adapted eye and is probably the most popular light source for clinical testing. The strobe is small and simple to use, but the disadvantages include difficulty in calibration, variation in output, and electromagnetic artifacts. The Tungsten/Quartz/Halogen light source, as a part of an optical stimulator, is more exactly calibrated, and can give repeatable stimulation for longer periods of time, but is much larger and more complicated to construct. Strobe units are usually placed 20-30 cm from the test eye whereas a fiber optic light guide can be used to deliver light from a tungsten halide lamp.31 The intensity of the stimulus will vary greatly with the type of flash used, but it is recommended that a minimum intensity of at least 4.0 log units greater than the absolute dark adapted threshold be used. This will mean that the stimulus is only 80-90% of that required for saturation2. Ganzfeld stimulators have also been used successfully in dogs, and provide indirect and homogeneous illumination of the visual field.32 This device also provides adapting background illumination in addition to stroboscopic flashes of light.

Signal Averager. Signal averagers may be used to enhance the signal-to-noise ratio of a response33-34. A signal averager is basically a computer that stores voltage values in bins at fixed times after the stimulus application. After the next stimulus application, the new voltage in each bin is added to the previous voltage and the sum divided by two. This process is repeated for a predetermined number of stimuli and the result is an averaged response that will look more or less like any single trace depending on the amount of background noise. This technique is particularly helpful when anesthesia is not in the best interest of the patient and some movement must be tolerated. The most common source of physiologic noise is muscle activity from the eyelids, which is enhanced if these muscles are stretched with a speculum. Signal averagers reduce electromyographic noise because this type of activity is truly random with respect to the stimulus trigger. In some instances, noise is the result of a faulty protocol and lengthy averaging is certainly not a an acceptable remedy.

Anesthetic Protocols
Recording ERGs from awake human beings or animals is not a painful process, but the procedure does require cooperation on the part of the patient. Because this level of cooperation is usually not attainable in conscious animals, anesthesia is often used to prevent movement artifacts, reduce patient stress, and allows the examiner to fix the position of the eye relative to the light source and recording electrodes. Most clinical protocols for anesthetized animals are designed to produce minimal effects that are then incorporated into a baseline. The most common effect of anesthesia on the ERG in dogs is general depression, i.e., a reduction in amplitude and an increase in implicit times.35 The effects of gaseous anesthetics, such as halothane2 or isoflurane,9 have some effect on the ERG, but this is offset by the elimination of much larger more consequential artifacts.2 To unequivocally demonstrate the specific effects of anesthetic agents in animals would require data before and after drug administration to be equally free of artifacts. This was accomplished in a study in dogs by comparing ERGs recorded under halothane anesthesia to ERGs recorded under succinylcholine akinesia.36 There were differences in most recorded variables with the most notable being a 36-77% reduction in the b-wave amplitude. In most ERGs under halothane, the latencies were prolonged with the exception of the dark-adapted series. It was further reported that the halothane-induced changes were independent of light and dark adaptation and independent of cone- and rod-mediated responses. The overall effects of halothane may simply be related to its ability to interfere with membrane transport.36 An anesthetic protocol that has given good results in dogs is the use of sodium thiopentone and halothane.37 An isoflurane-nitrous oxide protocol has been used in dogs because of its low glood:gas partition coefficent for rapid induction, low biotransformation, decreased myocardial sensitivity to catecholamines and positive tidal volume trend.9

Protocols for recording ERG in awake and freely moving animals have been developed38-39, but they are not ideal for clinical evaluations. Although signal averaging can appear to reduce artifacts encountered when recording from awake animals, this is probably not the best alternative. In rhesus monkeys, cone adaptation curves (response to 40-Hz flicker) were used to assess the influence of inhalant anesthetics on the ERG during the course of dark adaptation.40 The volatile anesthetics that were tested were methoxyflurane, halothane, enflurane, ether, and choloroform. All of the agents were shown to retard dark adaptation, perhaps by reducing the rate of photopigment regeneration. The ERG in rabbits was used to test for the effects of the volatile anesthetics methoxyflurane, halothane, and enflurane.41 The amplitudes of the a-wave and OP peak O1 showed dose-dependent decreases in amplitude. Even more striking were the change in O1 latencies. Also in rabbits, it was shown that alpha blockers reduced ERG amplitudes and beta agonists increased the retinal response to blue, red and white light. Adrenalin revealed mixed responses.42 Urethane anesthesia did not have an effect on the ERGs in mice, but dopamine blockade using haloperidol reduced the amplitude. However, that may have been due to a slight toxic retinitis in the inner retinal layer.43 A study in rats revealed that rod saturation is anesthetic dependent when comparing urethrane and pentobarbitol.44

The combination of ketamine and xylazine in dogs was reported to be preferable to thiopentane, halothane and nitrous oxide.45 Thirty one ophthalmascopically normal Labrador Retrievers that were given varying dosages of xylazine and ketamine did not show any significant changes in the ERG elicited by single flashes that were 1.60 log relative units above b-wave threshold.46 The dosages of xylazine ranged from 0.96-1.23 mg/kg body weight IM and ketamine ranged from 7.8-10.1 mg/kg body weight, IM. The addition of acepromazine (0.5 mg/kg IV) to ketamine hydrochloride (10 mg/kg IM) was reported to provide good immobilization in dogs for electroretinography.7 In another study, a combination of glycopyrrolate (0.01 mg/kg) and medetomidine hjydrochloride (15 ug/kg), was given IM to dogs. Twenty minutes later, ketamine hydrochloride (10 mg/kg) was given IV followed by vicuroniuim bromide (0.2 mg/kg) IV. This anesthetic regime, which was designed to prevent interference with receptor hyperpolarization during stimulation and to provide adequeate anesthesia and muscle relaxation, provided the 10-12 minutes necessary for an ERG exam.

In a recent study, dogs were sedated with IV ketamine (5 mg/kg) and then given 40 ug/kg of vecuronium bromide. Throughout the recording period, anesthesia was maintained with 25% oxygen and 75% nitrous oxide and muscle paralysis was sustained with vecuronium bromide at a rate of 400 ug/kg/hr.47 A similar procedure consisted of glycopyrrolate (0.01 mg/kg, IM), induction with thiamylal sodium (15 mg/kg, IV), and after incubation, pancuronium bromide (0.1 mg/kg, IV) to promote muscle relaxation.19 General anesthesia in cats undergoing electroretinographic examination has been achieved with sodium thiopentone and maintained with halothane14 or a mixture of ketamine hydrochloride (25 mg/kg, IM) and xylazine HCl (2.5 mg/kg, IM).48 One of the potential consequences of anesthesia is a depression of respiratory activity and possibly hypoxia. At a PaO2 of 45 mm Hg, the ERG begins to show the effects of hypoxia. The b-wave is affected first as seen in a decrease of the amplitude and an increase in the latency.8 The a-wave is apparently more resistent to decreased oxygen tension and remains after the b-wave is extinguished. The same trend is true when perfusion pressure is decreased. In experiments designed to look at c-wave changes in the face of increasing intraocular pressure (IOP) and decreased choroidalal oxygen tensions, it was observed that the ERG b-wave was less sensitive to changes than the c-wave.49 It was hypothesized that increasing IOP decreased choroidal blood flow which caused hypoxia and early changes in the c-wave, primarily an increase in amplitude. This may have been due to the inability of the choroidal circulation to autoregulate the way the retinal circulation does. In addition to the direct effects of anesthetic agents or hypoxia, the ERG in dogs has been shown to be adversely affected by CO2. Hyperventilation in anesthetized dogs enhanced the b-wave, and it was proposed that the mechanism was the reduction of CO2.50

Separation of Rod and Cone Function
The ERG provides clinicians with the opportunity of examining not only the retina in general, but also specific classes of photoreceptors. It should be the aim of any clinical protocol for recording ERGs to include a method for separating rod and cone components in the recordings. This is achieved primarily by varying the wavelength, intensity, duration, and position of light on the retina and by the frequency of stimulations. Manipulation of these variables allows one to take advantage of the characteristic ways that rods and cones respond to different types of light stimulation at different levels of light or dark adaptation as described in the duplicity theory. This theory states that the retina is characterized by two active systems, one that is based on rod function and one based on cone function. It has become common practice to refer to rod and cone systems as scotopic and photopic, respectively. Actually, the terms scotopic and photopic refer to the ambient light conditions during the test, i.e., a scotopic ERG is one recorded in the dark and a photopic ERG is one recorded in the light. Rods and cones contribute their signals to the ERG independently of each other and under certain conditions, their activities can be separated. Cones respond best to more intense light in the longer (red) wavelengths and can discharge at a more rapid rate, whereas rods respond best to shorter (blue) wavelengths at lower intensities and lower rates of discharge. Red flashes are produced by placing Kodak Wratten filter 25 in front of white light and blue flashes are produced by using Kokak Wratten filters 47, 47A, and 47B. Cones maintain sensitivity in the presence of background illumination whereas rod sensitivity decreases. A stimulus may activate both systems simultaneously or accentuate one or the other system. The exact nature of the response may also depend on the area of the retina that is stimulated. Cones are more numerous in the area centralis whereas rods increase in the peripheral zones in the retina.51

Some clinical electrodiagnosticians have proposed five ERG responses to be recorded as a minimum standard protocol:
1) a mixed rod and cone response obtained by a high intensity stimulus in a dark adapted eye,
2) a rod response by using a low intensity stimulus in a dark-adapted eye,
3) oscillatory potentials elicited by a high intensity stimulus in a dark-adapted eye,
4) cone response by high intensity stimulus in the light adapted eye ( or other techniques), and
5) responses obtained by a rapidly repeated stimulus (flicker).52

The length of time that an animal has been dark-adapted can also be used to further accentuate rod/cone separation. During the early phase of dark adaptation, a red flash can be used to demonstrate cone function. The development of the photopic b-wave occurs first with a low amplitude and short implicit time.53 This even occurs in animals such as owls, in which the retina is a predominantly populated with rods.54 As dark adaptation continues, however, the scotopic b-wave masks the photopic b-wave with its greater amplitude and longer implicit time. This sequential masking of the photopic b-wave with a scotopic b-wave does not occur in all animals. In sheep,55 deer,56 and swine,57 the photopic (bp) and scotopic (bs) b waves can be seen together as different peaks during light adaptation.

A rapid rate of light stimulation, or flicker, can also be used to separate rod and cone function. Flicker fusion is the frequency of stimulation at which individual responses can no longer be discerned. A bipartite flicker fusion response curve (FFRC) is a graph in which the log-luminance of the stimulus is plotted against the maximum frequency (flashes/sec) at which individual responses occur. The curve is characteristic for most dark-adapted mammals in that there is both a rod slope and a cone slope (Fig. 4). The shallow portion of the curve represents rod fusion at low frequencies and low intensity stimulation whereas the steeper portion of the curve represents cone fusion at greater intensities. The FFRC for the dark-adapted dog is similar to that recorded in human beings except for slight shifts along the log-luminance axis.58 The electrical activity from cones can be selectively recorded because only cones are able to keep pace with a flash rate of 30/sec.

Flicker photometry has also been used to examine the photopigment complements in dogs and foxes.59 These canids were shown to have a cone pigment with a peak sensitivity at 555 nm and a second one in the 430-435 nm range. The rod pigment had a peak at 508 nm or in a range of 500-510 nm.47 These results support the prediction that dichromatic color vision is characteristic of both dogs and foxes. A similar procedure was used in squirrel monkeys in which color-vision had been previously established using behavioral tests.60 In 12 monkeys, three different varieties of dichromacy were demonstrated. Prairie dogs also have two distinct photopigments but only one morphologically distinct receptor type (cones).61

Scotopically balanced red and blue light can be used to characterize diseases that have a predilection for receptors in one or the other system. If a long-wavelength (red) light produces an ERG with the same amplitude as does a short-wavelength (blue) light in a dark-adapted animal, then the stimuli are said to be scotopically balanced. The clinical use of these stimuli can reveal underlying causes of abnormal ERG responses resulting from monochromatic (white) light stimulation. For example, low-amplitude ERGs elicited by white light in dark-adapted Norwegian elkhounds was shown to consist primarily of the cone response. Rod dysplasia in these dogs basically eliminated the large rod response normally evoked by blue light.2

Even when red and blue lights are balanced, rod activity may mask cone activity in the resultant records at certain stages of dark adaptation.47 This may require further means for isolating the smaller cone activities by the use of a background light that eliminates rod components. This can be shown by the use of balanced red and blue stimuli in the presence of blue-green background light of increasing intensity. Figure 5 illustrates the similarity of the waveform produced by scotopically balanced red and blue light in dark-adapted dogs. Notice, however, the diminishing rod b-waves that were evoked by both blue and red stimuli as the relative intensity of the background light increased; the blue light no longer evoked a rod or cone response, but the red light produced a small cone b-wave with a short implicit time and low amplitude (left column). Furthermore, as soon as the scoptic b-wave appears during dark adaptation, the photopic b-wave will be masked and will no longer be apparent.

Normal ERG Data
The values for ERG implicit times and amplitudes in normal animals will vary according to the type and intensity of the flash, the type of recording electrodes, the bandwidth of the amplifier, and the length of time in the light or dark. Because no two labs use exactly the same protocol, a comparison of normal values will be difficult. A high intensity white flash presented to dogs dark-adapted for three minutes had mean a- and b-wave implicit times of 14.3 and 36 msec, respectively and amplitudes of 44 and 108 mV, respectively.33 The amplitudes of the a- and b-wave increase linearly as the stimulus intensity is increased logarithmically and as the period of dark adaptation increases up to about 60 minutes. After 60 minutes of dark adaptation in adult dogs, scotopic ERGs had a- and b-wave amplitudes of 99.3 and 316.7 mV, respectively,9 and a- and b-wave implicit times of 13.6 and 34 msec, respectively. Using a slightly different protocol that included 15 minutes of dark adaptation, a-waves were reported to have a mean amplitude of 41.56 mV and b-waves had a mean of 148.21 mV.62

Flash ERGs recorded from anesthetized cats, which were dark-adapted for 60 minutes, were characterized by an a-wave with an amplitude of 160.6 mV and an implicit time of 13.2 msec.63 The b-wave had an amplitude of 521.5 mV and an implicit time of 34.8 msec.

Flash ERGs recorded from Japanese Black and Holstein-friesian cattle were used to assess retinal function in visually deficient calves.64 After 15 minutes of dark adaptation, the amplitudes of a-waves ranged from 573-695 mV and the b-wave amplitudes ranged from 1,075-1,305 mV. The latencies of the a- and b-waves ranged from 9-12 msec and 36-41 msec, respectively. In light-adapted Holstein cows, mean a- and b-wave latencies were 13.9 and 30.2 msec, respectively and mean b-wave amplitudes were 43.24 mV.65

Scotopic Threshold Response (STR)
There are several negative components that contribute to the early phase of the cat ERG, some of which are strongly rod-dependent.66 These include a sustained negative potential, a negative going off response, and the M-wave. A rod-driven response in dark-adapted cats was reported to occur in the proximal retina at a 17% depth that is clearly different from the b-wave that occurs at a depth of 48%.67 This activity, which is called the scotopic threshold response (STR) contributes a negative potential to the diffuse ERG at threshold. At low stimulus intensities, the STR may actually resemble the a-wave that occurs with more intense stimulation.68 The STR appears to approximate the threshold of the most sensitive ganglion cells in cats and may be caused by proximal retinal neurons with an indirect participation by Muller cells. The contribution of the STR to the vitreal ERG at low stimulus intensities closely follows its presence in the proximal retina.

The SRT in the canine is a corneal negative response elicited by dim light stimulation even below that required for the scotopic b-wave (Fig. 6).69 This response, with amplitudes as high as 120 mV and implicit times that range from 75-96 msec, has a spectral sensitivity of that corresponding to the canine rod visual pigment. Further evidence of the scotopic nature of this response is evidenced by its disappearance in the presence of a dim background light that has no effect on the a- and b-wave components.

Pattern Electroretinogram
Much of the research devoted to identifying the structural contributors to the ERG has been based on those responses elicited by flashes of light. The flash stimulus produces a distinct response and as presented earlier, many refer to the flash ERG as the FERG. Cortical neurons are known to be rather insensitive to homogeneous light stimulation and there is reason to believe that patterned stimulation excites different retinal cells than non-patterned stimulation. The pattern-evoked electroretinogram or PERG is a type of ERG elicited by phase-reversal gratings.70 The pattern consists of checks or bars that are periodically reversed such that the light areas become dark and vice versa. In this way, the overall mean luminance remains constant and the stimulus is characterized by overall brightness, contrast between light and dark areas, the size of the gratings, and the rate of reversal. A further condition of this procedure requires that the stimulus be focused and stable on the retina. Accurate refraction and patient cooperation are essential but the response permits an examination of pattern-processing functions of the retina.

The source of the PERG is retinal, but different from the uniform-field FERG. Unlike the FERG, which is influenced by the luminance of the stimulus, the PERG relates more to the pattern of the stimulus and is dependent upon bandpass spatial characteristics. That is, the response is highly influenced by size of the pattern elements in such a way that increasing or decreasing spatial resolution increases or decreases components of the PERG. The PERG has been useful in human beings for the clinical diagnosis of retinal ischemia, macular disorders, optic nerve dysfunction, glaucoma, diabetes, amblyopia, Alzheimer's disease, increased intraocular pressure, and multiple sclerosis.71

Another important difference between PERG and FERG is that only the former is significantly suppressed by degeneration of ganglion cell axons.72-73 Several weeks after the optic nerve was transsected in cats, the amplitude of the PERG decreased while the amplitude of the FERG remained constant. Because the reduction in PERG amplitude paralleled the degeneration of ganglion cells, it was concluded that the source of response was in the inner layers of the retina. While in primates, the PERG is predominantly photopic, it is not surprising that the PERG is a mixed response in cats.74 The PERG has been used in dogs75-76 and cats77 to assess the limits of visual resolution. When similar techniques are performed in human beings, the correlation between the PERG estimate of resolution and verbal communication of visual resolution is significant.

The PERG has also been used in dogs to assess primary open angle glaucoma (POAG).76 The dogs tested were those with moderate signs of glaucoma including open iridocorneal angles, elevated intraocular pressure, cupping of the optic disc and disentergration of zonular ligaments. The visual resolution of the central 15° of the retina in normal dogs was 6.9 +2.6 minutes of arc/phase whereas the resolution in the toroidal 15° was 11.8+2.3 minutes of arc/phase. There were no significant differences between the normal and the glaucomatous dogs. Apparently, young dogs do not experience loss of visual resolution in the early stages of glaucoma. Using this procedure, the visual resolution in dogs is greater than that in cats, which correlates with a greater number of ganglion cells and optic nerve fibers.76

PERG have been reported in dogs using square-wave vertical grating patterns presented at spatial frequencies ranging from 0.015-1.92 cyles/degree.78 Signal averaging of responses elicited by low spatial frequencies produced a PERG characterized by a small negative peak (N1), a large positive peak (P1), and a much larger negative peak (N2) (Fig. 7). Increases in spatial frequencies caused a decrease in the amplitudes of P1 and N2 and an increase in the peak times of N1 and P1. PERG have been recorded in cats and unlike those recorded from human beings and dogs do not show a b-wave.79 The feline PERG consisted primarily of a vitreous negative after potential peaking about 120-200 msec after pattern reversal. The response amplitudes decreased with increasing spatial frequency and showed low-frequency attenuation with sinusoidal phase reversal but not with square wave reversal. This observation is important in identifying a pattern-specific component involving lateral interactions in the retina.

Oscillatory Potentials
Superimposed on the a- and b-waves of the ERG are four to ten peaks or wavelets collectively referred to as oscillatory potentials (OP). Oscillatory potentials have been reported in several animal species80-81 including dogs,82-83 and cats63-84. Although the name may imply a periodic activity, the OP are probably not true reverberations. The ERG and OP occur during the same time period but they represent different physiologic activities and have different cellular generators. Because OP are smaller than the ERG in which they are embedded, special recording procedures are used to better visualize the individual components.63, 82, 85

Direct retinal recordings and the use of pharmacologic agents that specifically interfere with selected cellular components have provided further information about OP. Intraretinal depth recordings of OP reveal that they probably arise from the interplexiform layer of the retina distal to ganglion cells and proximal to receptor cells.86 Furthermore, OP appear to represent a radial current flow involving inhibitory feedback circuits that are initiated by amacrine cells87 and each peak probably has a different origin.88 A recent report has provided evidence that the renin-angiotensin system has a direct effect on retinal neurons that is not related to its systemic effects.89 Such evidence supports the hypothesis of an antiotensinergic process involving amacrine cells in the inner retina. Reserpine, caused a selective suppression of later OP components in cats without a significant change in the a- or b-waves or early OP components.90 These observations appear to indicate a definite link between the activity of dopaminergic cells and generation of at least some of the OP peaks.87 Although the cones have been favored as generators of OP, some investigators have reported that both rods and cones participate.85

Like the ERG, characteristics of OP depend upon the intensity, rate, and spectral characteristics of the stimulus and the state of retinal adaptation. Selected aspects of the stimulus or recording amplifier can be manipulated to accentuate the OP and also reduce variability include bandpass filtering, dark adaptation, stimulus conditioning, stimulus duration, and signal averaging. The OP are best recorded with an intense white light stimulus. The dominant frequency of OPs is in the 100-160 Hz range depending on the state of adaptation. The OP can be selectively enhanced in dogs and cats by settting the bandwidth of a recording amplifier to a low frequency of 100 Hz and a high frequency of 500 Hz, (Fig. 8). Because OP are quite large in dark adapted animals, they are also visible in ERGs recorded with wider amplifier bandwidths (Fig. 9). Those OP peaks that occur on the peak of the b-wave are problematic because they obscure the exact point at which the b-wave is measured. To eliminate this, the high frequency setting can be reduced, but this artificially prolongs the latencies of both the a- and b-waves.91

Studies in human beings have demonstrated what is called a conditioning flash effect or CFE. Because the OP responses to flashes after the first of a series are larger and less variable than those elicited by the first flash, it is thought that the first response is somehow confounded by rod participation and subsequent responses are basically cone responses85. This would appear to indicate that OP are best recorded in the mesopic range of retinal sensitivity, i.e., somewhere between scotopic and photopic conditions.

Oscillatory potentials have been utilized clinically to assess several ophthalmologic disorders, such as diabetic retinopathy, retinal ischemia, pigmentary retinal degenerations, chloroquine retinopathy, central retinal artery occlusion, glaucoma, Alzheimer's disease, and other diffuse chorioretinal degenerations.92-94 Because OP are dependent on good retinal circulation, acute increases in intraocular pressure (IOP) cause changes in both latencies and amplitudes of OP peaks in dogs. 83 A 300% increase in IOP for 3-5 minutes completely obliterated both OP and ERG. Both responses return to normal when pressures are reduced to baseline. In this study, there was an average increase of 89.4% in the amplitude of OP peaks and an 81.68% increase in the b-wave amplitude by the end of 60 minutes of dark adaptation.

The effects of stimulus conditioning and intensity on OP peaks 01-04 were studied in anesthetized cats.63 In this study, there were mean decreases in peak time that ranged from 17.81 to 29.47 msec and mean amplitude increases of 19.72 to 57.03 lV for peaks O1-O4 across 3 log units of stimulus increase. Clearly then, OP peak time and amplitude should be considered as functions of stimulus intensity. The curves for peak time and amplitude reveal a distinct difference between O1 and O2-O4; one that is more dramatic for amplitude than peak time. This supports the claims of others that OP peaks probably have different generators. Studies have emphazised the uniqueness of O4 in humans, but the present study in cats seems to accentuate the activities of O1.

Visual Evoked Potentials
An electrodiagnostic evaluation of the visual system should include recordings of retinal activity, but also recordings of postretinal activity. Visual evoked potentials (VEP) are averaged postretinal potentials that arise from areas of the brain in response to visual stimulation. Possible generators of the VEP include rostral colliculi or their brachia, pretectal areas, and specific cortical visual areas. The clinical rationale for recording VEP is simple; a normal response indicates that the visual pathways from the retina to the visual cortex are intact. The involvement of the retina in the VEP is somewhat different from its role in the clinical ERG. The ERG is generated by large areas of the retina, but the VEP provides an indirect assessment of signals that originate primarily in the area centralis. The independence of retinal and postretinal visual potentials was recently demonstrated in a study in cats.48 Long-term (3-4 years) monocular visual deprivation was accomplished by lid sutures. At the end of the deprivation period, VEP were virtually eliminated in the test eye whereas PERG and flash ERG were unaffected. Several techniques have been described for recording VEP in dogs using a stroboscopic white flash for visual stimulation.39, 95-101 Others use a different type of stimulation such as LED light sources102 or square-wave vertical gratings.75 In most of the reports, VEP were recorded from anesthetized dogs,96-100 but some were recorded from awake dogs lightly restrained in a body sling.101 In some of the studies, dogs were dark adapted39, 95-97 and some were light adapted.100 A canine VEP can be recorded from subcutaneous electrodes overlying the brain in response to a flash or pattern-reversal stimulation. An example of a VEP recorded by a positive electrode placed subcutaneously about 3 cm. rostral to the bregma and referenced to a negative electrode placed under the chin is shown in Figure 10. These responses were recorded from the left side of the brain in response to stimulation of the right eye. Light stimulation was achieved with a 3x5 matrix of light-emitting diodes. In light-adapted cats, VEP were characterized by a polyphasic potential with latencies in the 0-100 ms range (Fig. 11).103

In human beings, VEP have been helpful in diagnosing demyelination, optic nerve damage as a consequence of glaucoma, axonal neuropathy, maculopathy and neurotransmitter deficiencies.104 Pattern-reversal stimulation is produced by a video display connected to a pattern generator or computer. Human patients are asked to fix the eyes at the center of the grating and maintain a focus throughout the procedure. Contrast of a stimulus projected onto the retina is important because of the center-surround organization of retina cells.104 The organization of receptive fields are not uniform across the entire retina and thus the pattern that best stimulates retinal components may vary in different species.

In addition to its role in disease diagnosis, the VEP has been used in humans and animals to predict subjective psychophysical contrast sensitivity or limits of spatial resolution.105 The differences between VEPs from different animals are more striking than the differences between ERGs. This is due not only the greater complexity of the VEP, but to a wider variety of recording protocols. As with other electrodiagnostic procedures that are not particularly painful but do require patient cooperation, VEP are usually recorded in anesthetized animals.

The exact waveform of the response will depend on the type and level of anesthesia, placement of electrodes, method of stimulation, and body temperature of the animal.100 White stroboscopic flashes were used to record VEP from rats, guinea pigs, cats, monkeys, and human beings using similar technique.106 The responses were similar among the species and consisted of 3-4 positive peaks with latencies ranging from 15-274 msec. Cats were particularly noteworthy because of fast P-N components during the first 45 msec. The interhemispheric variability in the VEP was shown to relate to the percentage of fibers that cross the midline in the chiasma. In cats, about 65% of the fibers in the optic nerve cross in the chiasma and in dogs, about 75% cross.107 In response to white light flash in dogs, VEP were characterized by a single positive peak with an average latency of 58 msec.99 In a later study in dogs also using a white flash stimulus, the VEP consisted of three positive peaks, P1, P2, and P3 with mean latencies of 14.3, 54.5, and 98.1 msec, respectively (Strain et al., 1990). In those studies that used stroboscopic flashes of white light, caution was advised because of the likelihood of volume conducted ERGs comprising much of the VEP activity recorded from scalp electrodes.96, 108

Using a 3 x 5 matrix of light-emitting diodes as a stimulator, VEP and ERG were recorded from 10 normal light-adapted adult dogs.103 The LED are quiet and preclude the VEP from containing stray auditory-evoked potentials. Four scalp electrodes were used to record VEP and ERG were recorded by 2 scalp electrodes near the eye and by a conjunctival electrode. The VEP consisted of 3 major positive peaks (P1 to P3) with peak latencies ranging from 20 to 70 msec (Fig. 12). There was no differences between the latencies of peaks P1-P3 when comparing VEP recorded 1 week apart in the same dog. The likelihood of volume-conducted activity from the retina was thus reduced in two ways. First, because the total illuminance of the LED matrix was much less than that of the average stroboscopic flash, the amplitude of the ERG was substantially reduced. What little remained of the ERG could be easily masked, even by low-amplitude cortical activity. Second, the amplitude of the ERG was low because the VEP were recorded from light-adapted dogs. The ERG was a low amplitude cone response.

The postretinal origin of the VEP was substantiated by recording VEP before and after unilateral optic nerve transsections in 4 dogs.102 Electroretinograms were also measured before and after surgery to assess the integrity of the retina. Postsurgically, the VEP was absent when the eye on the operated side was stimulated. Stimulation of the contralateral eye produced VEP with the same waveform shape, but the latencies were slightly prolonged when compared to presurgical recordings. Electroretinograms recorded from the remaining eye also had slightly prolonged b wave implicit times, but when these times were compared to ERG from the intact eye after surgery, there were no differences.

The VEP waveforms in animals are usually polyphasic and consist of positive and negative peaks in the 0-100 msec range. The nomenclature most commonly used is a designation of "P" for positive and "N" for negative followed by a number indicative of the order of the peak. In some cases, the number following the peak gives the expected mean value of the latency. For example, the P100 that is characteristic of VEP elicited by pattern-reversal stimulation in human beings is the positive peak with a mean latency of 100 msec. The primary clinical analysis of VEP consists of measuring the latencies and amplitudes of resultant peaks, especially the most prominent and/or the most stable potential. In human beings, this would be the P100.

The VEP, together with other tests, can provide data from different portions of the visual pathway. The simultaneous recording of PERG and VEP could potentially permit a differentiation of demyelinating conditions of the optic nerve for example, and lesions causing complete axonal loss and degeneration of ganglion cells.92 The VEP is particularly helpful in diagnosing blindness that may be post-retinal in origin. Optic nerve hypoplasia is an inherited disease that commonly causes blindness in dogs.109 Opthalmoscopically, the optic nerve head is small and this is accompanied by fixed and dilated pupils. The histopathologic changes include hypoplasia of ganglion cells and nerve fiber layers. One would reasonably expect the ERG associated with this condition to be normal, but the VEP is dramatically altered. Sheep and goats with thiamine-responsive polioencephalomalacia (PEM) were reported to have abnormal VEP but normal ERG.108 The sudden onset of blindness in PEM is reported to be cortical as a result of neuronal degeneration in gray matter.

Postnatal Development of Visual Potentials
Electrodiagnostic procedures have become very important for the early diagnosis of selected eye diseases in animals. Unaffected littermates may provide age-matched controls, but data collected during the maturation of normal animals is also needed if entire litters are affected. The maturation of the two receptor types may differ within the same animal and differ among species. Most animals have retinas in which both rods and cones develop with age, even though they develop at different rates. Exceptions to this general rule are the tree shrew (Tupaia) and some Sciuridae, which have retinas that contain only cones.61

At birth, the photoreceptor layer in the canine retina is not formed; i.e., the two granular layers have not separated. The ERG is absent at birth, but begins to appear in puppies at about 10-15 days of age and is completely formed by 15-28 days of age.62, 97, 110 The first response to occur is a small negative wave that appears at the beginning of the a-wave. This small negative response occurrs at the time of differentiation of the rods and cones. During development, the amplitudes increase and latencies decrease until they approximate adult values between the 5th and 8th weeks of age.62, 110 The greatest changes in ERG amplitude are expected to occur between the 3rd and 4th weeks of age at a time that corresponds to maturation of the inner and outer nuclear layers.62 Maturation is evidenced by an increase in the amplitude of the b-wave and a decrease in the latency. Using a stimulus equal to 4.0 log foot lamberts, the amplitudes for a- and b-waves in three-week-old pups were on the order of 0-15 mV and 25-50 mV, respectively.7 One week later, the amplitudes had increased to 0-90 mV and 40-135 mV, respectively. In newborn lambs, the implicit times for a- and b-waves decreased during from birth to 30 days of age, but the interval between the respective peaks did not change.111 This was interpreted as meaning that the maturation was primarily occurring in the photoreceptors.

In Red Irish Setters, the b-wave reaches its full size between 40 and 50 days.10 High amplitude ERGs occur during maturation at a time when the retina is reaching maturity. In vitro recordings of transretinal mass receptor potentials in 9 to 10-week-old dogs point to potential changes across photoreceptors as the primary cause of the amplitude increase.112 Clearly, the cone system in dogs develops more rapidly than the rod system.113 The amplitudes of the ERG components increase and the implicit times decrease with age. The waveform of the ERG changes from being a-wave dominated to b-wave dominated, events that are paralleled by the development of the outer plexiform layer.

Kittens are born with an immature visual system that requires about 4 months for maturation. Photoreceptors are mature by the end of the first month and the remainder of the maturational process occurs in more proximal elements such as ganglion cells. Development of the electroretinogram in kittens has also been reported to occur in phases. The first stage is marked by the b-wave appearance during the first 10 days after birth, and the second phase corresponds to the increase in amplitude of the b-wave (10-49 days). The b-wave evoked by high intensity stimulation matures by 5-7 weeks of age whereas maturation of the b-wave evoked by low-level stimulation requires an additional 3-5 weeks.114 The third phase is required for attainment of adult implicit time values and oscillatory potential development.114 A study of cone b-wave maturation in kittens revealed that between 25 and 94 days of age, the amplitudes were either normal or greater than normal.115 Adults values for the b-wave were attained after 100 days, implicit times reached adult levels by 80 days, and flicker fusion matured at about 74 days of age. As in dogs, maturation in the cat includes ages at which the b-wave amplitudes are greater than those expected in adults. Some have suggested that there may actually be more functional cells in the retina than in adults during those periods and that the decline to adult values is a reflection of normal degeneration.115

The manner in which the canine retina responds to a flickering stimulus also changes with maturation and permits an evaluation of rod and cone function.113 The cone branch of the FFRC is present first at 16 days of age and by 30 days, the FFRC was clearly bipartite. The break point between rod and cone branches occurred at lower intensity levels and by 60 days, the values were essentially the same as adults. Oscillatory potentials appear on the b-wave of the ERG at about 3 weeks of age with high light intensities.116 By 11 weeks of age, the potentials were well-developed and became apparent with lower intensities.

The development of VEP in puppies was reported to begin as early as the second day after birth with a negative wave characterized by a latency that exceeded 500 msec.98 By day 3 or 4, the peak had shortened and this was followed by the addition of another negative peak. The number of peaks increased between 2 and 3 weeks of age, which was accompanied by a decrease in peak latencies. By the 5th week, the response simplified somewhat with considerably shortened latencies. In a subsequent study conducted in puppies from 7 to 100 days of age, it was shown that the latency for P1 of the VEP was mature as early as 11 days and 38 days was required for the latency of N1 and P2.117 The latencies of N2 and P3 were not mature by the end of the study. During the first 5 weeks after birth, the maturation of the VEP closely parallels neuronal development in the visual cortex.95 Generally, VEPs in maturing animals are more variable between individuals and more fatigable than in adults.97 Long latency visual evoked responses appear in the feline visual cortex on or about the second day of life and short latency responses at 5 days to 2 weeks of age.118 The first potnetial reported to appear was a negative wave that would become the N2 in the adult.119 The second potential was N1, which occurred between 5-8 days of age, and then P1 preceeded N1 by 9-14 days.

The long latency response resembles that in an adult cat between 30-40 days. Between 14 and 20 days, kittens begin to visually locate or avoid objects.120 Postnatal development of OP in kittens follow the same general course as ERG and VEP. Prior to 7 weeks of age, the OP differed from adults in numbers, amplitude and time between peaks.114 By 18 weeks of age, values were similar to those of adults. Sequential maturation of VEP components was reported in lambs during the first 30 days after birth.111 The early component, N1, did not show a significant decrease in latency whereas the later potential, P2, did. This was taken to mean that the primary nerve tracts (N1) were mature at birth, but increasing myelination and integration in the cortex accounted for shorter latencies for P2.

Use of the ERG in The Diagnosis of Retinal Diseases


Progressive Retinal Atrophy
Electroretinography has became an essential diagnostic aid for several retinal diseases in veterinary species. Among those that have been documented are rod-cone dysplasia, pigment epithelial dystrophy or central progressive retinal atrophy, rod dysplasia, and hemeralopia. In most of these, except pigment epithelial dystrophy, the photoreceptor layer is affected first with secondary degeneration of remaining retinal layers. Progressive retinal atrophy (PRA) refers to inherited diseases that are characterized by degeneration or dysplasia of the retina. These abnormalities are progressive and result in more or less predictable visual deficits much like retinitis pigmentosa in human beings.29, 121 The different varieties of PRA differ in onset and severity of ophthalmic, behavioral, and electroretinographic changes, and the layers of the retina that are involved.

Progressive retinal atrophy has been classified as generalized PRA (GPRA) in which the photoreceptors are affected and central PRA (CPRA), in which the neural retina degenerates secondary to the pigment epithelial layer. Generally, the early ophthalmoscopic signs of GPRA include an increase in pupil size, increased reflectivity of the tapetal fundus, retinal vascular attenuation. Nyctalopia is usually the first behavioral sign and vision may progress towards total loss. As the visual defect progresses, the pupillary light reflex is depressed with no accompanying anisocoria. Many types of PRA are inherited as simple autosomal recessive traits.37 The specific characteristics of PRA vary between breeds, and even within breeds different types have been reported. In contrast to GPRA, CPRA is characterized by hypertrophy of pigment epithelial cells.122 What begins as irregular pigment foci in the tapetal region just above the optic nerve near the area centralis progresses to include the entire pigment epithelium. Central PRA is actually a type of retinal pigment epithelial dystrophy.

Another classification of PRA is based on age of onset and progression of the disease.113 The photoreceptor dysplasias are those with an early onset and rapid progression. The second group consists of those diseases characterized by abnormal photoreceptor development which occurs early but progress slowly. Those retinal degenerations with a late onset and slow progression constitute the third group. Clinical onset can be as early as 6 weeks in the Norwegian Elkhound.113 6-8 weeks in the Irish Setter123, 6 months in the longhair Dachshund124, ten months in the Tibetan Terrier37, 2-3 years in Samoyeds, and 3-5 years in Miniature Poodles.125

Because ophthalmic signs of disease are often undetectable in puppies affected with PRA, the ERG is the only method of early diagnosis. In fact, the majority of the reports in the veterinary literature that pertain to the clinical use of the ERG deal with some form of PRA. The age at which PRA is electroretinographically confirmed varies with the type of PRA and the breed in which it occurs (Table I). In the Dachshund, the early development of the ERG is normal, but changes appeared as early as 17 weeks. By nine months of age, ERG were characterized by reduced amplitudes and prolonged latencies, or extinguished altogether. Histopathologic changes, which precede clinical or electrodiagnostic changes, appear as early as 9 weeks of age in Tibetan Terriers.

Different electrodiagnostic procedures have been developed to take advantage of the way that PRA affects certain species and breeds of animals. PRA may initially involve both rods and cones, but in many cases, the progression in rods is greater than in cones.37 An electroretinographic technique was proposed for use in diagnosing rod dysplasia (RD) in 6-week-old Norwegian Elkhound pups.141 After light adapting for 5 minutes against a background intensity of 10 ft.-lamberts, the background was discontinued and red flashes presented at 0, 1,2,3,10,15, and 20 minutes of dark adaptation. Last, a white flash was presented after 20 minutes of adaptation. In normal pups the initial photophic ERG was followed by the gradual development of the scotopic ERG after 1 minute. As dark adaptation progressed, the scotopic b-wave overwhelmed the photopic b-wave. In the RD-affected puppies, the scotpic b-wave was absent and the photopic b-wave did not change in form or amplitude during dark adaptation.

The electroretinograhic responses to flickering stimuli has also been used to diagnose early retinal degeneration (ERD) in Norwegian Elkhounds.113 In affected dogs, the responses were low in amplitude and failed to follow the stimulus in a 1:1 manner. The responses were initially corneal-positive (b-wave dominated), but gradually became more corneal negative (a-wave dominated) as maturation proceeded. In ERD-affected animals, the overall sensitivity was reduced (curve shifted to the right), the rod branch of the curve was lost, and the cone branch eventually failed (Fig. 13).113

Full-field stimulation was also useful in diagnosing progressive rod-cone degeneration (PRCD) in miniature poodles.127 From 4-8 months of age, affected animals had normal ERGs, but a reduction in rod responses were obvious between 12-16 months while cone responses were normal. Eventually cone responses showed abnormalities even though implicit times remained normal.

In miniature poodles and English cockers, the ERG is characterized by a loss of rod function with preservation of cone function. The changes, which consisted of reduced amplitudes and prolonged implicit times, were obvious in dogs ranging in age from 9-28 weeks. In Tibetan terriers, ERG changes were not apparent until approximately 10 months and consisted of changes in the amplitude only.37 The same pattern is seen in American and English cocker spaniels in which the onset may be seen between 2-3 years. The onset in American cockers may be slightly earlier than in English cockers.142

Retinopathies in cats have also been reported which, like dogs, are characterized by loss of rods and cones and by thinning of the outer nuclear layer of the retina.15 The fundus appeared devoid of blood vessels but without optic atrophy. The cause of some concern was the coincident absence of a recordable ERG but the presence of a pupillary light reflex using bright light. Retinal degeneration has also been reported in Abyssinian cats in which the electrophysiologic characteristics of the disease were similar to retinitis pigmentosa in human beings.14 There was no change in the waveform or implicit times in affected cats, but the amplitudes of the a- and b-waves were diminished. In kittens that were 8-16 weeks of age, there was a decrease in the maximum dark adapted (45 minutes) b-wave amplitude, but not the a-wave. This occurs at a time when retinal sensitivity is also slightly decreased.

Electroretinography has also been performed in Rdy cats which are affected with rod-cone dysplasia (Rdy is the gene symbol).116 In 4.5 week-old kittens, the ERG was absent or just barely recordable using contact lens electrodes, but were recordable using intravitreal needle electrodes. Using this technique, the ERG was devoid of oscillatory potentials, the a-wave dominated the ERG, both a- and b-waves had reduced amplitudes and prolonged implicit times, the b-wave threshold was higher than the a-wave threshold, and photopic ERG could not be recorded. The a-wave domination seemed to point to defective synaptogenesis in the outer plexiform layer or dystropic retinas. Even using the intravitreal technique, ERG could not be recorded at 7 months of age. In displaying an a-wave dominated ERG, the Rdy cats were like Norwegian Elkhounds affected with early retinal degeneration, but unlike them in having prolonged implicit times.

Hemeralopia
Hemeralopia is a retinal disease characterized by the progressive loss of cone photoreceptors. The behavioral sign is obviously day-blindness or inability to see in areas where there is a high level of artificial light. The electroretinogram reveals the absence of cone function while rods continue to function normally. Scotopic flicker fusion was used to identify hemeralopia in Alaskan Malamutes as early as 6 weeks of age.12 Hemeralopia was confirmed if flicker ERGs were not recordable at 20 flashes/sec (fps). In Malamutes that were behaviorally day-blind, flicker responses were not obtainable at a frequency above 10 flashes per second.126 Non-hemeralopic dogs could still respond at frequencies as high as 75 fps. There were no differences between affected and non-affected dogs when using single flashes in dark adapted dogs. In 8-week-old Malamute pups, flicker fusion frequency (FFF) was 5-15 fps and in the non-hemeralopes, the fusion was between 44 and 75 fps.127 By 24 weeks of age, FFF was 76 fps in non-hemeralopes and 10-18 in the hemeralopes. In some pups, FFF was high at 8 weeks of age, but fell to much lower rates by 24 weeks, thus revealing its progressive nature.

Cataracts, Glaucoma and Sudden Blindness
Cataracts are often associated clinically with retinal degeneration or retinal detachment. ERGs are usually recommended prior to cataract removal because an opaque lens often prevents visualization of the fundus. The only benefit in removing a cloudy lens from an animal with advanced retinal dysfunction would be cosmetic. Cataracts are common in miniature and toy poodles and mature quite rapidly. This rapid maturation prevents an ophthalmoscopic confirmation of PRA. Cataracts alone will not necessarily eliminate or even depress the ERG.12 In some cases, the presence of a cataract will actually enhance the amplitude of the a- and b-waves. The cataractous lens will scatter light more than a clear lens and the end result is that more of the peripheral retina is stimulated.. A depressed ERG in an animal with cataracts is usually a sign that there is some form of retinal disease present. Because as many as 13.4% of the dogs with cataracts also have some form of retinopathy, pre-surgical electroretinographic evaluation of the retina is advised. Glaucoma has been shown to reduce the amplitude of the b-wave and cause a prolongation of the a-wave,12 especially during the end stages of the disease when retinal damage is the result of chronically high pressures. In Basset Hounds, Beagles, and American Cocker Spaniel, glaucoma causes a depression in the a- and b-waves of the ERG, which is detectable in pups that are 6-18 months of age.7 Most of the reported diseases that cause blindness in dogs have a gradual onset and in many breeds follow a predictable pattern that includes electroretinographic changes. There is one form of blindness, however, that has a sudden onset and does not appear to be breed-specific. This acute form of blindness, which is called sudden acquired retinal degeneration (SARD), is accompanied by an ophthalmoscopically normal fundus and abnormal pupillary light reflexes.13 Electrodiagnostically, this disease is characterized by bilaterally extinguished ERGs. Histopathologic examinations of retinae from affected dogs reveal that retinal changes are restricted to the photoreceptor layer whereas the inner layers of the retinas are normal.143 There is some suspicion that the disease has an autoimmune component in which affected dogs become sensitized by their own retinal antigens.

Optic Nerve Hypoplasia and Retinal Detachment
Optic nerve hypoplasia is not uncommon in dogs. This condition may be inherited and does not always occur bilaterally. In dogs with hypoplastic optic nerves, the principle histopathologic finding is a reduction in the diameter of the optic nerves, chiasma, and optic tracts.6 The retina is characterized by a reduced number of ganglion cells and decreased thickness of the optic-fiber layer, but other layers appeared to be normal. The ERG may be normal because ganglion cells and optic nerve fibers do not directly contribute to the response.109 In some cases however, ERGs were reported to have inconsistent a-waves, and there was a progressive reduction in the a- and b-waves and the flicker-fusion was low. These findings were taken as evidence of additional retinal disease that was not described by light microscopy.6 The ERG can be combined with ultrasonography to detect retinal detachments. In most cases, the ERG waveforms are diminished early, but disappear altogether in the later stages.7 In some cases, the ERG may be recordable in cases of retinal detachment, but the b wave may be absent leaving only the a wave.12

Vitamin E Deficiency Retinopathy
Puppies fed a vitamin E deficient diet have been reported to show ophthalmoscopic signs of retinopathy beginning as early as 3 months after weaning.144 Early signs began around the central artery with mottling or stippling in the deeper layers of the retina. After dogs had been vitamin E deficient for 4 months, the mottling worsened and ERGs were not recordable. Dogs were clinically blind in dim light, but not in bright light. Vitamin E supplementation for the next 2 months resulted in the return of ERGs with diminished a- and b-wave amplitudes. Prolonged vitamin E deficiency did result in permanently extinguished ERGs.

Equine Night Blindness
Equine night blindness (ENB) has been reported to occur in Appaloosas with an onset as early as one month of age.145 The ERG in a foal that showed obvious night blindness was reported to consist of a long, slow negative wave that consumed approximately 400 msec. This pattern has been called a "negative ERG," and taken as a physiologic indicator of scotopic function. The proposed cause of ENB is a neural transmission defect in the region of the inner segments of the photoreceptors or in the region of the bipolar cells.

Retinal Degeneration in Cattle
Blindness in Hostein-Friesian cattle was reported to be related to diffuse hydropic degeneration of ganglion cells accompanied by the invasion of melanin in the nerve fiber layer, loss of photoreceptors, and thinning of nuclear layers.64 Affected calves also showed evidence of internal hydrocephalus and cerebellar hypoplasia at necropsy. In the same study, blindness in Japanese Black cattle was apparently related to thinning of retinal nuclear layers, vacuolar degeneration of Muller cells and absence of photoreceptors. The ERGs and OPs in blind calves were either non-recordable or had reduced amplitudes.

Bibliography
1. Moore, C.P.: Visual disturbance in the dog. Part I. The visual process, ptient assessment, and nonneurologic causes. Comp. Cont. Ed. Small Anim. Pract., 6:465, 1983.
2. Acland, G.M.: Diagnosis and differentiation of retinal diseases in small animals by electroretinography. Semin. Vet. Med. Surg. (Small Anim.), 3:15, 1988.
3. Aguirre, G.D., and Rubin, L.F.: The electroretinogram in dogs with inherited cone degeneration. Invest. Ophthalmol., 11:840-847, 1975.
4. Buyukmihci, N., Aguirre, G., and Marshall, J.: Retinal degenerations in the dog. II. Development of the retina in rod-cone dysplasia. Exp. Eye Res., 30:575, 1980.
5. Carter, J.D.: Clinical use of electroretinography. Iowa St. Univ. Vet., 3:86, 1969.
6. Gelatt, K.N., and Leipold, H.W.: Bilateral optic nerve hypoplasia in two dogs. Can. Vet. J., 12:91, 1971.
7. Gum, G.G.: Electrophysiology in veterinary ophthalmology. Vet. Clin. N. Amer. Small Anim. Pract., 10:437, 1980.
8. Howard, D.R., and Sawyer, D.C.: Electroretinography of acute hypoxic and increased intraocular pressure status in the dog. Am. J. Vet. Res., 36:81, 1975.
9. Jones, R.D., et al.: An electroretinogram protocol for toxcicological screening in the canine model. Toxicol. Let., ____:223, 1994.
10. Parry, H.B., Tansley, K., and Thomson, L.C.: Electroretinogram during development of hereditary retinal degeneration in the dog. Br. J. Ophthalmol., 39:349, 1955.
11. Rubin, L.F.: Clinical electroretinography in dogs. J. Am. Vet. Med. Assoc., 151:1456, 1967. 12. Rubin, L.F., Bourns, T.K.D., and Lord, L.H.: Hemeralopia in Dogs: Heredity of hemeralopia in Alaskan Malamutes. Am. J. Vet. Res., 28:355, 1967.
13. Venter, I.J., Petrick, S.W.: Akute blindheid in 'n hond veroorsaak deur skielike verworwe retinale degenerasie. Tydskr. S. Afr. Vet. Ver., 66:32-34, 1995.
14. Narfstrom, K., Arden, G,B., and Nilsson, S.E.G.: Retinal sensitivity in hereditary retinal degeneration in Abyssinian cats: Electrophysiological similarities between man and cat. Br. J. Ophthalmol., 73:516, 1989.
15. Rubin, L.F.: Atrophy of rods and cones in the cat retina. J. Am. Vet. Med. Assoc., 142:1415, 1963.
16. Schmidt, S.Y., et al.: Retinal degeneration in cats fed casein. III. Taurine deficiency and ERG amplitudes. Invest. Ophthal. Vis. Sci., 16:673, 1977.
17. Eithoven, W., and Jolly, W.A.: The form and magnitude of the electrical response of the eye to stimulation by light at various intensities. Q. J. Exp. Physiol., 1:373, 1908.
18. Armington, John C.: The Electroretinogram. New York, Academic Press, 1974.
19. Dawson W.W., and Kommonen, B.: The late positive retinal potential in dogs. Exp. Eye Res., 60:173, 1995.
20. Granit, R.: The components of the retinal action potential in mammals and their relation to the discharge in the optic nerve. J. Phjysiol., 77:207, 1933.
21. De Rouck, A.F.: History of the electroretinogram. In Principles and Practice of Clinical Electrophysiology of Vision. Edited by J.R. Heckenlively and A.B. Geoffrey. St. Louis, Mosby Year Book, 1991, pp. 5-13.
22. Armington, John C.: Electroretinography. In Electrodiagnosis in Clinical Neurology. 3rd Ed. Edited by M.J. Aminoff. New York, Churchill Livingstone, 1992, pp. 433-466.
23. Gouras, P.: Electroretinography: Some basic principles. Invest. Ophthalmol., 9:557, 1970.
24. Acland, G.M.: Electroretinography: Choosing and using the appropriate hardware. Trans. Am. Coll. Vet. Ophthalmol., 10:141, 1979.
25. Strain, G.M., and Carter, J.D.: Atraumatic loop electrodes for electroretinography. Vet. Res. Comm., 11:9, 1987. 26. Dawson, W.W., Tricke, G.L., and Litzkow, C.A.: Improved electrode for electroretinography. Invest. Ophthalmol. Vis. Sci., 18:988, 1979.
27. Steiss, J.E., Storrs, D.P., and Wright, J.C.: Comparisons of electroretinograms recorded with a contact lens versus needle electrode in clinically normal dogs. Prog. Vet. Comp. Ophthalmol., 2:143, 1992.
28. Witzel, D.A., Johnson, J.H., Pitts, D.G., and Smith III, E.L.: Scleral contact lens electrodes for electroretinography in domesticated animals. Am. J. Vet. Res., 37:983, 1976.
29. Aguirre, G.D., and Rubin, L.F.: An electrophysiologic approach for early diagnosis of progressive retinal atrophy in the Norwegian Elkhound. J. Am. Anim. Hosp., 7:136, 1971.
30. Odom, J.V.: Amplifiers and special-purpose data acquisition systems. In Principles and Practice of Clinical Electrophysiology of Vision. Edited by J.R. Heckenlively and A.B. Geoffrey. St. Louis, Mosby Year Book, 1991, pp. 183-192.
31. Aguirre, G.D., and Rubin, L.F.: Progressive retinal atrophy (rod dysplasia) in the Norwegian Elkhound. J. Am. Vet. Med. Assoc., 158:208, 1971.
32. Schaeppi, U., and Liverani, F.: Procedures for routine clinical electroretinography (ERG) in dogs. Ag. Act., 7:347, 1977.
33. Howard, D.R., Keller, W.F., and Blanchard, G.L.: Clinical electroretinography. A protocol for testing the retina. J. Amer. Anim. Hosp., 9:219, 1973.
34. Coulter, D.B., and Martin, C.L.: Signal averaging of electroretinograms. Ga. Vet., 30:14, 1978.
35. Kommonen, B.: The DC-recorded dog electroretinogram in ketamine medetommidine anesthesia. Acta Vet. Scand., 29:35, 1988.
36. Acland, G., Forte, S., and Aguirre, G.: Halothane effect on the canine electroretinogram. Trans. Am. Coll. Vet. Ophthalmol., 12:66, 1981.
37. Millichamp, N.J., Curtis, R., and Barnett, K.C.: Progressive retinal atrophy in Tibetan Terriers. J. Am. Vet. Med. Assoc., 192:769, 1988.
38. Imai, R., et al.: A procedure for recording electroretinogram and visual evoked potential in freely moving cats. J. Toxicol. Sci., 15:263, 1990.
39. Sato, S., Sugimoto, S., and Chiba, S.: A procedure for recording electroretinogram and visual evoked potential in conscious dogs. J. Pharmacol. Meth., 8:173, 1982.
40. Norren, D.V., and Padmos, P.: Influence of anesthetics, ethyl alcohol, and freon on dark adaptation of monkey cone ERG. Invest. Ophthalmol. Vis. Sci., 16:80, 1977.
41. Tashiro, C., Muranishi, R., Gomyo, I., Mashimo, T., Tomi, K., and Yoshiya, I.: Electroretinogram as a possible monitor of anesthetic depth. Arch. Clin. Exp. Ophthalmol., 224:473, 1986.
42. Czepita, D.: Influence of a- and b-adrenergic stimulators and blockers on the electroretinogram and visually evoked potentials of the rabbit. Biomed. Biochim. Acta, 49:509, 1990.
43. Adachi-Usami, E., Ikeda, H., and Satoh, H.: Haloperidol delays visually evoked cortical potentials but not electroretinograms in mice. J. Ocular. Pharm., 6:203, 1990.
44. Brown, C.H., and Green, D.G.: Rod saturation in b-wave of the rat lectroretinogram under two different anesthetics. Vision Res., 24:87, 1984.
45. Kommonen, B., Karhunen, U., and Raitta, C.: Effects of thiopentone halothane-nitrous oxide anaesthesia compared to ketamine-xylazine anaesthesia on the DC recorded dog electroretinogram. Acta Vet. Scand., 29:23, 1988.
46. Kommonen, B., and Raitta, C.: Electroretinography in Labrador Retrievers given ketamine-xylazine anesthesia. Am. J. Vet. Res., 48:1325, 1987.
47. Yanase, J., Ogawa, H., and Ohtsuka, H.: Rod and cone components in the dog electroretinogram during and after dark adaptation. J. Vet. Med. Sci., 57:877, 1995. 48. Baro, J.A., Lehmkuhle, S., and Dratz, K.E.: Electroretinograms and visual evoked potentials in long-term monocularly deprived cats. Invest. Ophthalmol. Vis. Sci., 31:1405, 1990. 49. Yancey, C.H., and Linsenmeier, A.: The electroretinogram and choroidal PO2 in the cat during elevated intraocular pressure. Invest. Ophthalmol. Vis. Sci., 29:700, 1988. 50. Murray, M.J., and Borda, R.P.: Physiologic correlates of the ERG hyperventilatory response in dogs. Acta Ophthalmol. Copenh., 62:808, 1984. 51. Koch, S.A., and Rubin, L.F.: Distribution of cones in retina of the normal dog. Am. J. Vet. Res., 33:361, 1972. 52. Marmor, M.F., Arden, G.B., Nilsson, S., and Zrenner, E.: Standard for clinical electroretinography. In Principles and Practice of Clinical Electrophysiology of Vision. Edited by J.R. Heckenlively and A.B. Geoffrey. St. Louis, Mosby Year Book, 1991, pp. 283-290. 53. Aguirre, G.D., and Rubin, L.F.: Progressive retinal atrophy in the miniature poodle: An electrophysiologic study. J. Am. Vet. Med. Assoc., 160:191, 1972. 54. Ault, S.J.: Electroretinograms and retinal structure of the eastern screech owl (Otus asio) and great horned owl (Bubo virginianus). Raptor Res., 18:62, 1984. 55. Smith, E.L., Witzel, D.A., and Pitts, D.G.: The waveform and scotopic CFF of the sheep electroretinogram. Vision Res., 16:1241, 1976. 56. Witzel, D.A., Springer, M.D., and Mollenhauer, H.H.: Cone and rod photoreceptors in the White-tailed Deer Odocoileus virginianus. Am. J. Vet. Res., 39:699, 1978. 57. Witzel, D.A., Smith, E.L., Beerwinkle, K.R., and Johnson, J.H.: Arsanilic acid-induced blindness in swine: Electroretinographic and visually evoked responses. Am. J. Vet. Res., 37:521, 1976. 58. Liverani, F., and Schaeppi, U.: Electroretinography as an indicator of toxic retinopathy in dogs. Pharmac. Ther., 5:599 1979. 59. Jacobs, G.H., et al.: Photopigments of dogs and foxes and their implications for canid vision. Vis. Neurosci., 10:173, 1993. 60. Neitz, J., and Jacobs, G.H.: Electroretinogram measurements of cone spectral sensitivity in dichromatic monkeys. J. Ophthalmol. Soc. Am., 12:1175, 1984. 61. Jacob, G.H., and Yolton, R.L.: Some characteristics of the eye and the electroretinogram of the prairie dog. Exp. Neurol., 37:538, 1972. 62. Gum, G.G., Gelatt, K.N., and Samuelson, D.A.: Maturation of the retina of the canine neonate as determined by electroretinography and histology. Am. J. Vet. Res., 45:1166, 1984. 63. Sims, M.H., Sackman, J.E., McLean, R.A., and Slaymaker, C.: Effects of stimulus intensity and conditioning on the electroretinogram and oscillatory potentials in dark-adapted cats. Prog. Vet. Comp. Ophthalmol., 3:177-185, 1991. 64. Kotani, T., et al.: The normal electroretinogram in cattle and its clinical application in calves with visual defects. Prog. Vet. Comp. Ophthalmol., 3:37, 1993. 65. Strain, G.M., Olcott, B.M., and Hokett, L.D.: Electroretinogram and visual-evoked potential measurements in Holstein cows. Am. J. Vet. Res., 47:1079-1081, 1986. 66. Frishman, L.J., and Steinberg, R.H.: Origin of negative potentials in the light-adapted ERG of cat retina. J. Neurophysiol., 63:1333, 1990. 67. Sieving, P.A., Frishman, L.J., and Steinberg, R.H.: Scotopic threshold response of proximal retina in cat. J. Neurophysiol., 56:1049, 1986. 68. Frishman, L.J., Sieving, P.A., and Steinberg, R.H.: Contributions to the electroretinogram of currents originating in proximal retina. Vis. Neurosci., 1:307, 1988. 69. Yanase, J., Ogawa, H., and Ohtsuka, H.: Scotopic threshold response of the electroretinogram of dogs. Am. J. Vet. Res., 57:361, 1996. 70. Riggs, L.A., Johnson, E.P., and Schrick, A.M.L.: Electrical responses of the human eye to moving stimulus patterns. Science, 144:567, 1964. 71. Hull, B.M., and Thompson, D.A.: Review of the clinical applications of the pattern electroretinogram. Ophthalmol. Physiol. Opt., 9:143, 1989. 72. Maffei, L., and Fiorentini, A.: Electroretinographic responses to alternating gratings before and after section of the optic nerve. Science, 211:953, 1981. 73. Maffei, L., et al.: Pattern ERG in the monkey after section of the optic nerve. Exp. Brain Res., 59:423, 1985. 74. Maffei, L., and Fiorentini, A.: Pattern visual evoked potentials and electroretinograms in man and animals. In Visual Evoked Potentials. Edited by J.E. Desmedt. New York, Elsevier Science, 1990, pp. 25-33. 75. Odom, J.V., Bromberg, N.M., and Dawson, W.W.: Canine visual acuity: Retinal and cortical field potentials evoked by pattern stimulation. Am. J. Physiol., 245:R637, 1983. 76. Ofri, R., Dawson, W.W., and Gelatt, K.N.: Visual resolution in normal and glaucomatous dogs determined by pattern electroretinogram. Prog. Vet. Comp. Ophthalmol., 3:111, 1993. 77. Schuurmans, R.P., and Berninger, T.: Pattern reversal responses in man and cat: A comparison. Ophthalmic Res., 16:67, 1984.
78. Sims, M.H., and Ward, D.A.: Response of pattern-electroretinograms (PERG) in dogs to alterations in the spatial frequency of the stimulus. Prog. Vet. Comp. Ophthalmol., 3:106-112, 1993. 79. Ohzawa, I., and Freeman, R.D.: Pattern evoked potentials from the cat's retina. J. Neurophysiol., 54:691, 1985. 80. Masuda, Y.: Studies on "the oscillatory potential" of the rabbit's ERG. Acta Soc. Ophthalmol. Jpn., 65:1481, 1961.
81. Kadona, H., Kitamura, M., and Iwano, Y.: Oscillatory potential in the chicken ERG. Jpn. J. Vet. Sci., 49:801, 1987. 82. Sims, M.H., and Brooks, D.E.: Changes in oscillatory potentials in the canine electroretinogram during dark adaptation. Am. J. Vet. Res., 51:1580, 1990. 83. Brooks, D.E., Sims, M.H., and Gum, G.G.: Changes in oscillatory potentials of the canine electroretinogram during acute sequential elevations in intraocular pressure. Prog. Vet. Comp. Ophthalmol., 2:80, 1992. 84. Doty, R.W., and Kimura, D.S.: Oscillatory potentials in the visual system of cats and monkeys. J. Physiol., 168:205, 1963. 85. Peachey, N.S., Alexander, K.R., and Fishman, G.A.: Rod and cone system contributions to oscillatory potentials: An explanation for the conditioning flash effect. Vision Res., 27:859, 1987. 86. Wachtmeister, L., and Dowling, J.E.: The oscillatory potentials of the mudpuppy retina. Invest. Ophthal. Vis. Sci., 17:1176, 1978.
87. Heynsen, H., Wachtmeister, L., and Van Norren, D.: Origin of the oscillatory potentials in the primate retina. Vis. Res., 25:1365, 1985.
88. Yanagida, T., et al.: Microelectrode depth study of the electroretinographic oscillatory potentials in the frog retina. Doc. Ophthalmol., 67:355, 1988. 89. Jacobi, P.C., Osswald, H., Jurklies, B., and Zrenner, E.: Neuromodulatory effects of the renin-angiotensis system on the cat electroretinogram. Invest. Ophthalmol. Vis. Sci., 35:973, 1994. 90. Gutierrez, C.O., and Spiguel, R.D.: Oscillatory potentials of the cat retina. Effects of adrenergic drugs. Life Sci., 13:991, 1973. 91. Sims, M.H.: Partial masking of the canine electroretinogram by oscillatory potentials. J. Am. Vet. Med. Assoc., 4:40, 1990.
92. Celesia, G.G., and Tobimatsu, S.: Electroretinograms to flash and to patterned visual stimuli in retinal and optic nerve disorders. In Visual Evoked Potentials. Edited by J.E. Desmedt. Amsterdam, Elsevier, 1990, pp. 45-55.
93. Bresnick, G.H., et al.: Electroretinographic oscillatory potentials predict progression of diabetic retinopathy. Preliminary report. Arch. Ophthal., 102:1307, 1984. 94. Nesher, R., and Trick, G.L.: The pattern electroretinogram in retinal and optic nerve disease. Doc. Ophthalmol., 77:225, 1991. 95. Fox, M.W.: Neuronal development and ontogeny of evoked potentials in auditory and visual cortex of the dog. Electroenceph. Clin. Neurophysiol., 24:213, 1968.
96. Malnati, G.A., Marshall, A.E., and Coulter, D.B.: Electroretinographic components of the canine visual evoked response. Am. J. Vet. Res., 42:159, 1981. 97. Boyer, S., and Kirk, G.R.: Maturation of the visual evoked response in the dog. Exp. Neurol., 38:449, 1973. 98. Myslivecek, J.: The development of the response to light flash in the visual cortex of the dog. Brain Res., 10:418, 1968. 99. Howard, D.R., and Breazile, J.E.: Normal visual cortical-evoked response in the dog. Am. J. Vet. Res., 33:2155, 1972. 100. Aunon, J.I., Weirich, W.E., and Nyholm, R.: Effects of hypothermia on the visual-evoked brain potential in dogs. Am. J. Vet. Res., 38:383, 1977. 101. Strain, G.M., Jackson, R.M., and Tedford, B.L.: Visual evoked potentials in the clinically normal dog. J. Vet. Int. Med., 4:222, 1990.
102. Sims, M.H., Laratta, L.J., Bubb, W.J., and Morgan, R.V.: Waveform analysis and reproducibility of visual-evoked potentials in dogs. Am. J. Vet. Res., 50:1823, 1989. 103. Sims, M.H., and Laratta, L.J.: Visual-evoked potentials in cats, using a light-emitting diode stimulator. Am. J. Vet. Res., 49:1876, 1988. 104. Bodis-Wollner, I., Brannan, J.R., Ghilardi, M.F., et al.: The importance of physiology to visual evoked potentials. In Visual Evoked Potentials. Edited by J.E. Desmedt. New York, Elsevier Science, 1990, pp. 1-24. 105. Maffei, L., and Fiorentini, A.: The pattern electroretinogram in animals and humans: Physiological and clinical. In Vision and the Brain. Edited by B. Cohen and I. Bodis-Wollner. New York, Raven Press, Ltd., 1990, pp. 289-296.
106. Creel, D.J., Dustman, R.E., and Beck, E.C.: Visually evoked responses in the rat, guinea pig, cat, monkey, and man. Exp. Neurol., 40:351, 1973.
107. Kornegay, J.N.: Small animal neuroophthalmology. Comp. Cont. Ed., 11:923, 1980. 108. Strain, G.M., Claxton, M.S., Olcott, B.M., and Turnquist, S.E.: Visual-evoked potentials and electroretinograms in ruminants with thiamine-responsive polioencephalomalacia or suspected listeriosis. Am. J. Vet. Res., 51:1513, 1990. 109. Gwin, R.M., and Wyatt, R.L.: Diagnosis and treatment of retinopathies in the dog and cat. Proceedings, Kal Kan symposium for the treatment of dog and cat diseases, Columbus, 1979, pp. 9-17. 110. Kirk, G.R, and Boyer, S.F.: Maturation of the electroretinogram in the dog. Exp. Neurol., 38:252, 1973. 111. Woods, J.R., Prisi, V., Coppes, V., and Brooks, D.E.: Maturational sequence of the visual system: Serial measurements of visual evoked potential and electroretinogram in the healthy neonatal lamb. J. Obstet. Gynecol., 145:738, 1983. 112. Millichamp, N.J., and Arden, G.B.: Transretinal mass receptor potentials recorded from the canine retina in vitro. Am. J. Vet. Res., 50:1710, 1989. 113. Acland, G.M., and Aguirre, G.D.: Retinal degenerations in the dog: IV. Early retinal degeneration (erd) in Norwegian Elkhounds. Exp. Eye Res., 44:491, 1987. 114. Hamasaki, D.I., and Maguire, G.W.: Physiological development of the kitten's retina: An ERG study. Vision Res., 25:1537, 1985.
115. Jacobson, S.G., Ikeda, H., and Ruddock, K.: Cone-mediated retinal function in cats during development. Documenta. Ophthalmol., 65:7, 1987. 116. Leon, A., Hussain, A.A., and Curtis, R.: Autosomal dominant rod-cone dysplasia in the Rdy Cat. 2. Electrophysiological findings. Exp. Eye Res., 53:489, 1991. 117. Strain, G.M., Jackson, R.M., and Tedford, B.L.: Postnatal development of the visual-evoked potential in dogs. Am. J. Vet. Res., 52:231-235, 1991. 118. Ellingson, R.J., and Wilcott, R.C.: Development of evoked responses in visual and auditory cortices of kittens. J. Neurophysiol., 23:363, 1960. 119. Rose, G.H., Gruenau, S.P., and Spencer, J.W.: Maturation of visual electrocortical responses in unanesthetized kittens: Effects of barbiturate anesthesia. Electroenceph. Clin. Neurophysiol., 33:141, 1972.
120. Rose, G.H., and Lindsley, D.B.: Development of visually evoked potentials in kittens: Specific and nonspecific responses. J. Neurophysiol., 31:607, 1968. 121. Wiechmann, A.F., Font, R.L., and Hollyfield, J.G.: Idiopathid retinal degeneration in the dog: Differential patterns of [3H]Uridine incorporation and HIOMT-like immunoreactivity in surviving photoreceptors. Exp. Eye Res., 52:311, 1991.
122. Aguirre, G.D., and Laties, A.: Pigment epithelial dystrophy in the dog. Exp. Eye Res., 23:247, 1976. 123. Aguirre, G.D., and Rubin, L.F.: Rod-cone dysplasia (progressive retinal atrophy) in Irish Setters. J. Am. Vet. Med. Assoc., 166:157, 1975. 124. Curtis, R., and Barnett, K.C.: Progressive retinal atrophy in miniature longhaired dachshund dogs. Br. Vet. J., 149:71, 1993. 125. Aguirre, G., et al.: Pathogenesis of progressive rod-cone degeneration in miniature poodles. Invest. Ophthalmol. Vis. Sci., 23:610, 1982. 126. Rubin, L.F.: Clinical features of hemeralopia in the adult Alaskan Malamute. J. Am. Vet. Med. Assoc., 158:1696, 1971. 127. Rubin, L.F.: Hemeralopia in alaskan Malamute pups. J. Am. Vet. Med. Assoc., 158:1699, 1971. 128. Curtis, R., Barnett, K.C., and Leon, A.: Diseases of the canine posterior segment. In Veterinary Ophthalmology. Edited by K.N. Gelatt. Philadelphia, Lea & Febiger, 1991, pp. 461-525. 129. Wolf, E.D., Vainisi, S.J., and Scantos-Anderson, R.S.: Rod-cone dysplasia in the Collie. J. Am. Vet. Med. Assoc., 173:1331, 1978. 130. Slatter, D.H., Blogg, J.R., and Constable, I.J.: Retinal degeneration in greyhounds. Aust. Vet. J., 56:106, 1980.
131. Hodgman, S.F.J., et al.: Progressive retinal atrophy in dogs. I. The disease in Irish Setters (Red). Vet. Rec., 61:185, 1949. 132. Aguirre, G.D., et al.: Progressive retinal atrophy in the Miniature Schnauzer. Trans. Am. Coll. Vet. Ophthalmol., 16:226, 1985.
133. Acland, G.M., et al.: A new early onset inherited retinal degeneration in the Norwegian Elkhound. Trans. Am. Coll. Vet. Oph