Skip to content
Pass Duke Elder
Textbook / Chapter 15 / Clinical examination, investigations and imaging

Electrodiagnostic tests and neuroimaging

Electrodiagnostic tests record the electrical signals of the visual system, from the pigment epithelium to the occipital cortex. They show function where imaging shows only structure, and they localise a fault to one layer of cells. This lesson explains the electroretinogram, electro-oculogram and visual evoked potential, then sets out when to ask for CT and when for MRI.

11 minute read3 linked questions

One pathway, three tests

Light is converted to an electrical signal in the photoreceptors, passed to bipolar cells, then to ganglion cells, whose axons form the optic nerve and carry it to the brain. Behind the photoreceptors the retinal pigment epithelium maintains a steady voltage across itself. Each stage can be recorded separately from outside the eye, so the tests divide the pathway into parts.

TestStimulusSource of the signal
Electro-oculogram (EOG)Change from dark to lightRetinal pigment epithelium
Full-field ERGFlash filling the whole fieldPhotoreceptors and bipolar cells across the whole retina
Multifocal ERGFlickering hexagonsCones in many small areas of the central retina
Pattern ERGReversing chequerboardMacula and ganglion cells
Visual evoked potential (VEP)Reversing chequerboard or flashWhole pathway from macula to occipital cortex
What each electrodiagnostic test examines

The tests are used when symptoms and examination do not match: poor vision with a normal-looking fundus, night blindness, unexplained visual loss in a child, suspected inherited retinal disease, drug toxicity, or visual loss thought not to be organic.

3D atlasThe generators: pigment epithelium (EOG), photoreceptors (a-wave), bipolar cells in the inner nuclear layer (b-wave) and ganglion cells (pattern ERG).

The full-field electroretinogram

The electroretinogram (ERG) is the retina's electrical response to a flash of light, recorded with a fine electrode touching the cornea or lower lid and a reference electrode on the skin. For the full-field test the pupils are dilated and the flash is delivered inside a dome so that it lights the entire retina evenly. The recording is the summed response of all of it.

A bright flash in a dark-adapted eye gives a waveform with two main parts. The a-wave is an initial downward (negative) deflection. It comes from the photoreceptors, which hyperpolarise when light strikes them. The b-wave is the larger upward (positive) wave that follows. It comes from the bipolar cells of the inner retina, with a contribution from the Muller cells around them. Small ripples on the rising edge of the b-wave are the oscillatory potentials, attributed mainly to amacrine cells. They are reduced early in diseases of the retinal circulation such as diabetic retinopathy.

Rods and cones are separated by changing the conditions. After 20 minutes in the dark, a dim flash stimulates only rods. A bright flash in the dark gives a mixed response dominated by rods. After 10 minutes of light adaptation the rods are saturated, and a flash on a lit background, or a rapid flicker at about 30 Hz that rods cannot follow, isolates the cones.

PatternMeaningExamples
Rod responses lost first, cones later; eventually nothing recordableRod-cone dystrophyRetinitis pigmentosa
Cone responses reduced, rod responses preservedCone dysfunctionCone dystrophy; achromatopsia, in which cone responses are absent from birth
Normal a-wave with a b-wave smaller than the a-wave (a negative ERG)Photoreceptors work but the signal is not passed on to the inner retinaCongenital stationary night blindness, X-linked retinoschisis, melanoma-associated retinopathy
Reduced b-wave and oscillatory potentialsInner retinal ischaemiaCentral retinal artery occlusion
NormalDisease confined to a small area, or lying beyond the bipolar cellsMacular dystrophies, optic nerve disease
Full-field ERG patterns

Pattern and multifocal ERG

Two variants fill the gaps left by the full-field test. The pattern ERG uses a chequerboard whose black and white squares swap places while the overall brightness stays constant. The eye is not dilated and must be in focus. Only the central retina responds. The waveform has a positive peak at about 50 milliseconds, the P50, and a later negative trough, the N95. Both depend on working macular cones. The N95 arises from the ganglion cells. A reduced P50 therefore indicates macular dysfunction, and a selective loss of the N95 with a preserved P50 indicates ganglion cell or optic nerve disease. This makes the pattern ERG the bridge between a retinal and an optic nerve cause when the VEP is abnormal.

The multifocal ERG maps cone function across the central 30 degrees or so. The screen shows an array of 61 or 103 hexagons, each flickering in its own sequence, and the computer extracts a small local response for each one. The result is a topographic map with a peak at the fovea. It detects localised macular and paracentral loss that the full-field test cannot. In hydroxychloroquine retinopathy the responses in the ring around the fovea fall first.

The electro-oculogram and the Arden ratio

The eye carries a standing voltage of a few millivolts, with the cornea positive relative to the back of the eye. It is generated across the retinal pigment epithelium. The electro-oculogram (EOG) measures it indirectly. Electrodes are placed on the skin at the inner and outer corners of each eye. When the eye turns towards an electrode, the positive cornea comes closer and the recorded voltage rises. The patient looks from side to side between two fixed lights, and the size of the swing in voltage tracks the size of the standing potential.

The standing potential changes with lighting. During 15 minutes in the dark it falls to a minimum, the dark trough. When the lights are turned on it climbs over several minutes to a maximum, the light peak. The rise depends on a signal from illuminated photoreceptors reaching a healthy pigment epithelium.

Arden ratio = light peak / dark troughA ratio of 1.8 or more is normal. Below 1.65 is clearly subnormal.

Worked example: a dark trough of 300 microvolts and a light peak of 390 microvolts give an Arden ratio of 390 / 300 = 1.3, which is abnormal.

Most diseases that reduce the EOG also reduce the ERG, because damaged photoreceptors send no signal to the pigment epithelium. The EOG earns its place in one condition. Best disease (Best vitelliform macular dystrophy) is caused by mutation of BEST1, whose protein, bestrophin, is needed in the pigment epithelium for the light rise. The fault affects the pigment epithelium across the whole fundus although the visible lesion is only at the macula. The result is the classic pairing of a reduced Arden ratio with a normal full-field ERG. Carriers with a normal fundus also show the abnormal EOG.

Visual evoked potentials

The visual evoked potential (VEP) is the response of the visual cortex, recorded from a scalp electrode over the occiput. It is tiny against the background brain activity, so the stimulus is repeated many times and the responses are averaged. Each eye is tested alone, with glasses on and pupils undilated. Because the macula has a very large share of the visual cortex, the VEP is chiefly a test of central vision and of the pathway that carries it.

The standard stimulus is a reversing chequerboard. The waveform has a negative, positive and negative sequence named by polarity and typical timing: N75, P100 and N135. The P100, a positive peak about 100 milliseconds after each reversal, is the one measured. Two features are read: its latency (how late it arrives) and its amplitude (how big it is).

Latency reflects conduction speed. Myelin allows fast conduction, so demyelination delays the P100, often while its amplitude is fairly well preserved. This is the signature of optic neuritis. The delay commonly persists long after vision has recovered, so a VEP can show evidence of a previous, even unnoticed, attack and support a diagnosis of multiple sclerosis. Amplitude reflects the number of working axons. Compression, ischaemia and toxic damage mainly reduce amplitude.

  • Infants and people who cannot cooperate: a flash VEP needs no fixation and shows whether any signal reaches the cortex.
  • Suspected non-organic visual loss: a normal pattern VEP to small squares is hard to reconcile with a claim of very poor acuity.
  • Albinism: too many fibres cross at the chiasm, so stimulating one eye gives a response that is lopsided towards the opposite hemisphere when recorded with several electrodes across the occiput.

An abnormal VEP does not by itself locate the fault. Uncorrected refractive error, cataract and macular disease all degrade it. A pattern ERG done alongside separates macular causes from optic nerve causes.

CT or MRI

Computed tomography (CT) uses X-rays. It is fast, widely available out of hours, and excellent for bone, fresh blood, calcification and metal. It exposes the lens to radiation and shows soft tissue contrast less well. Magnetic resonance imaging (MRI) uses a strong magnetic field and radio waves. It gives far better soft tissue detail, particularly of the optic nerve, orbital apex, cavernous sinus, chiasm and brain, with no ionising radiation. It is slower, needs the patient to lie still, and is unsafe with some implants and with metallic foreign bodies. In the orbit, bright fat hides disease unless the sequence suppresses the fat signal, and gadolinium contrast shows inflammation and tumour.

Clinical problemFirst choiceReason
Orbital trauma, suspected blow-out fractureCT of the orbits with thin slices and coronal reconstructionsShows bone, and muscle or fat herniated into the sinus
Suspected intraocular or orbital foreign bodyCTLocates metal accurately. MRI can move a magnetic fragment and must not be used unless metal has been excluded
Optic neuritisMRI of brain and orbits with gadolinium and fat suppressionShows enhancement of the nerve and white matter lesions that predict multiple sclerosis
Thyroid eye diseaseCT for bony anatomy and the orbital apex, especially before decompression; MRI to show active muscle inflammationBoth show enlarged muscle bellies with spared tendons
Painful or pupil-involving third nerve palsyUrgent CT angiography or MR angiographyExcludes a posterior communicating artery aneurysm
Chiasmal field defect or unexplained optic atrophyMRI with contrastBest view of the pituitary region and the optic nerves
Which scan to request

Some details behind the table are worth knowing. Wood and other low-density organic material may be invisible on CT, and plain X-rays show only radio-opaque fragments. In optic neuritis, the Optic Neuritis Treatment Trial found that about 25% of patients with a normal baseline brain MRI developed multiple sclerosis within 15 years, against about 72% of those with one or more white matter lesions. In thyroid eye disease the inferior rectus is affected most often, then the medial rectus, and crowding of swollen muscles at the apex is what compresses the optic nerve.

The fibres that constrict the pupil travel on the outside of the third nerve, where an aneurysm pressing on the nerve reaches them first. A microvascular palsy in a person with diabetes tends to spare them because they have a better collateral blood supply. A new third nerve palsy with a dilated pupil is therefore treated as an aneurysm until imaging proves otherwise. Compression can occasionally spare the pupil, so many authorities advise imaging every isolated third nerve palsy.

Summary

  • The ERG a-wave comes from photoreceptors and the b-wave from bipolar cells with Muller cells.
  • Dark adaptation isolates rods, and light adaptation and 30 Hz flicker isolate cones.
  • A negative ERG, a normal a-wave with a reduced b-wave, means the fault lies after the photoreceptors.
  • The full-field ERG is normal in disease confined to the macula and in optic nerve disease.
  • The pattern ERG separates macular dysfunction (P50) from ganglion cell dysfunction (N95); the multifocal ERG maps macular cone function.
  • The EOG measures the standing potential of the pigment epithelium; the Arden ratio is light peak over dark trough and is normal at 1.8 or more.
  • Best disease gives an abnormal EOG with a normal full-field ERG.
  • The VEP P100 is delayed by demyelination, classically after optic neuritis, and the delay persists after recovery.
  • CT is the scan for trauma, fractures and metallic foreign bodies; MRI is contraindicated when a metallic foreign body is possible.
  • MRI is the scan for optic neuritis and soft tissue; a pupil-involving third nerve palsy needs urgent angiography for an aneurysm.

Check yourself

Quick recall on this lesson. These are easier than the question bank.

1. Which cells generate the a-wave of the electroretinogram?
2. Which result is typical of Best disease?
3. What is the typical VEP finding after optic neuritis?
4. A man hammering metal may have an intraocular foreign body. Which imaging should be requested?
Practise 3 questions on this lesson

Written for revision. It is not a source of clinical advice, and guidance changes.