14.1 Multifocal Electroretinogram (mfERG): Topographic Retinal Mapping

Multifocal Electroretinogram (mfERG): Topographic Retinal Mapping

Core Clinical Mandate: Full-field electroretinography (ffERG) measures the summed, mass electrical potential of the entire retina. Because the macula accounts for less than 5% of the total retinal surface area, extensive localized macular lesions—including dense geographic atrophy, severe vitelliform dystrophies, or early toxic retinopathies—can exist in the presence of a completely normal full-field ERG. The multifocal electroretinogram (mfERG) resolves this diagnostic blind spot by mathematically mapping local cone-system electrophysiologic responses across the central 40° to 50° of the retina.

---

Biophysical Foundations and Limitations of Full-Field ERG

To appreciate the clinical utility of the multifocal ERG, the ophthalmic medical technologist must first understand why conventional full-field (Ganzfeld) electroretinography fails to diagnose focal macular pathology:

• Mass Response Physics: Full-field ERG records a global electrical response. A pathological process must damage at least 15% to 20% of the total retinal area before a measurable reduction in a-wave or b-wave amplitude becomes detectable on a Ganzfeld recording.
• Foveal Surface Area Disproportion: The fovea centralis covers approximately 1.5 mm² of the retina (~0.1% of the total 1,200 mm² retinal area), and the entire anatomical macula (5.5 mm diameter) represents only approximately 2% to 4% of the total retinal surface. Consequently, a patient with absolute foveal destruction, legally blind with 20/400 visual acuity, will demonstrate an entirely normal full-field photopic and scotopic ERG.
• The Solution: Developed by Erich Sutter in the early 1990s, the multifocal electroretinogram uses a single continuous recording channel to extract hundreds of discrete, localized retinal electrical responses simultaneously, generating a detailed topographic map of central retinal function.

Mathematical Architecture: Binary m-Sequences and Hexagonal Arrays

The m-Sequence (Maximum-Length Sequence)

Rather than stimulating each retinal location sequentially—which would require hours of testing—mfERG presents all stimulus elements simultaneously using a specialized mathematical algorithm known as a maximum-length sequence (m-sequence):

• Binary Pseudorandom Sequence: An m-sequence is a deterministic, pseudorandom binary code generated by shift registers with feedback. Each state in the sequence is either 0 (black hexagon) or 1 (white hexagon).
• Orthogonality and Balance: Over the course of the sequence, each hexagon is white exactly $(2^n - 1) / 2$ times and black exactly $(2^n - 1) / 2$ times. Furthermore, the cross-correlation between any two distinct time-shifted versions of the sequence is mathematically zero (orthogonality).
• Stimulus Frame Rate: Hexagons alternate between black and white at the video frame rate, typically 75 Hz (one frame every 13.3 milliseconds). The sequence order is usually set to $n = 14$ or $n = 15$, resulting in $2^{14} - 1 = 16,383$ or $2^{15} - 1 = 32,767$ frames, corresponding to a total recording time of approximately 3.6 to 7.3 minutes.

Hexagonal Stimulus Scaling with Eccentricity

The stimulus array consists of a mosaic of abutting regular hexagons displayed on a calibrated video monitor:

• Array Density: The International Society for Clinical Electrophysiology of Vision (ISCEV) standard specifies arrays containing either 61 hexagons (standard clinical screening) or 103 hexagons (higher spatial resolution for subtle parafoveal defects). Research applications may utilize 241 hexagons.
• Eccentricity Scaling (Cortical/Retinal Magnification): The hexagons are not uniform in physical size. Hexagons in the central fovea are very small, while hexagons in the peripheral field are progressively larger.
  • Physiological Rationale: Cone photoreceptor density drops precipitously with increasing distance from the fovea (from ~150,000 cones/mm² at the foveal center to ~5,000 cones/mm² in the periphery). Scaling hexagon size inversely with cone density ensures that each individual hexagon stimulates approximately the same absolute number of cone photoreceptors.
  • Signal Equilibrium: This geometric scaling produces local electrical responses of approximately equal amplitude across the entire healthy array, preventing large peripheral responses from drowning out tiny central foveal potentials.

Mathematical Kernel Deconvolution

The raw electrical signal recorded at the cornea is a continuous, complex voltage fluctuation representing the summed response of all 61 or 103 flickering hexagons. To isolate the individual response of a single hexagon:

1. Cross-Correlation: The continuous corneal signal is cross-correlated with the unique binary m-sequence of that specific hexagon using a Fast Walsh-Hadamard Transform (FWHT).
2. First-Order Kernel (K1): Represents the basic linear impulse response of the local retinal patch to a single flash. It reflects the average difference between the electrical response when the hexagon was white versus when it was black.
3. Second-Order Kernels (K2): Represent non-linear adaptive interactions, reflecting how the response to a given flash is modulated by the preceding flash (local retinal adaptation and recovery kinetics).

---

Waveform Morphology and Cellular Neurophysiology

The extracted first-order mfERG response waveform closely resembles a miniature conventional photopic flash ERG, but its underlying cellular generators exhibit key electrophysiologic distinctions.

   P1 Peak (~30-40 ms, Depolarizing/Hyperpolarizing Bipolar Cells)
       /\
      /  \
     /    \
____/      \          /
\  /        \        /
 \/          \      /
 N1 Trough    \____/
 (~15-20 ms)    N2 Trough (~45-55 ms)
 (Cones & OFF-Bipolar)

Waveform Components and Latencies

1. N1 Trough:
   • Polarity and Timing: Initial negative deflection occurring at 15 to 20 milliseconds post-stimulus.
   • Cellular Origin: Generated predominantly by hyperpolarizing cone photoreceptors (outer segments) with a substantial post-receptoral contribution from OFF-bipolar cells.
   • Functional Analogy: Analogous, but not identical, to the photopic a-wave of full-field ERG.
2. P1 Peak:
   • Polarity and Timing: Prominent positive peak occurring at 30 to 40 milliseconds post-stimulus.
   • Cellular Origin: Driven primarily by post-receptoral inner nuclear layer elements, specifically depolarizing ON-bipolar cells and hyperpolarizing OFF-bipolar cells. Unlike the full-field b-wave, Müller glial cells contribute minimally to the mfERG P1 peak.
   • Functional Analogy: Serves as the primary clinical marker for local retinal response density and signal delay.
3. N2 Trough:
   • Polarity and Timing: Second negative trough following P1, occurring at 45 to 55 milliseconds.
   • Cellular Origin: Complex post-receptoral contributions involving OFF-bipolar cells, horizontal cells, and third-order inner plexiform layer elements.

Quantitative Parameters: Amplitude Density and Implicit Time

• Response Density (nV/deg²): Because hexagon areas vary by eccentricity, absolute microvolt amplitude cannot be compared directly between center and periphery. Instead, amplitude is normalized to the solid visual angle subtended by each hexagon, yielding response density expressed in nanovolts per square degree (nV/deg²).
  • Normal central foveal response density: ~80 to 120 nV/deg².
  • Normal peripheral response density: ~15 to 25 nV/deg².
• Implicit Time (Latency): The time in milliseconds from the onset of the stimulus frame to the crest of the P1 peak (or trough of N1). Prolongation of P1 implicit time reflects delayed synaptic transmission, outer retinal ischemia, or cellular stress.

Clinical Topographic Displays and Concentric Ring Analysis

Standard clinical mfERG software formats the extracted responses into three complementary visual representations:

1. Hexagonal Trace Array

Displays the individual first-order response waveforms arranged in their exact spatial geometry across the fundus. The technologist and clinician inspect this array to localize discrete, focal scotomas, quadrantic defects, or sectorial dropouts (e.g., matching a branch retinal vein occlusion or laser photocoagulation scar).

2. Three-Dimensional (3D) Response Density Topography

A continuous, pseudocolor surface plot representing local response density across the central 40° to 50° visual field:

• Resembles the classical "hill of vision" seen in static automated perimetry.
• In a healthy eye, it demonstrates a towering central peak corresponding to the foveola, sloping down smoothly into a symmetric, surrounding plateau.
• Pathologies present as craters, trenches, or collapsed central volcanoes.

3. Concentric Ring (Averaged Ring) Analysis

To improve the signal-to-noise ratio and quantify subtle paracentral defects, responses are mathematically averaged into 5 concentric annular rings centered on the fovea:

• Ring 1 (Fovea): Central 0° to 2° (central single hexagon).
• Ring 2 (Parafovea): Approximately 2° to 5° eccentricity.
• Ring 3 (Perifovea): Approximately 5° to 10° eccentricity.
• Ring 4 (Near Periphery): Approximately 10° to 15° eccentricity.
• Ring 5 (Mid Periphery): Approximately 15° to 20°+ eccentricity.

Clinicians evaluate the Ring Ratios (e.g., Ring 1 / Ring 2, or Ring 1 / Ring 4). A healthy eye maintains a steep fovea-to-periphery amplitude ratio (>3:1). A decrease in Ring 2 and Ring 3 relative to Ring 1 signifies paracentral ring depression.

---

Clinical Applications and Pathological Patterns

Hydroxychloroquine (Plaquenil) Retinal Toxicity Screening

Hydroxychloroquine (HCQ) and chloroquine (CQ) are widely prescribed for systemic lupus erythematosus, rheumatoid arthritis, and dermatological conditions. Retinal toxicity is irreversible and can progress even after drug cessation, making early objective screening imperative.

According to the American Academy of Ophthalmology (AAO) Recommendations on Screening for Chloroquine and Hydroxychloroquine Retinopathy:

• Baseline and Annual Screening: Annual screening should begin after 5 years of use for patients without major risk factors (daily dose >5.0 mg/kg actual body weight, renal disease, concomitant tamoxifen use, or pre-existing maculopathy).
• Primary Objective Modalities: mfERG is designated as one of the primary objective screening modalities alongside Spectral-Domain OCT (SD-OCT), Fundus Autofluorescence (FAF), and 10-2 Automated Visual Fields.
• Ethnic Divergence in Toxicity Patterns:
  • Non-Asian (Caucasian) Patients: Hydroxychloroquine binds to melanin in the RPE and accumulates in the parafoveal region. The classic early toxicity pattern is paracentral ring depression, characterized by selective amplitude reduction and implicit time delay in Ring 2 and Ring 3 (2° to 6° from fixation), with relative preservation of the central foveal peak (Ring 1) and peripheral rings. This creates a "flying saucer" sign on OCT and a ring scotoma on mfERG.
  • Asian Patients: Demonstrates a distinct anatomical predisposition where toxicity manifests predominantly in a pericentral / extramacular distribution. Cone dysfunction develops along the vascular arcades in Ring 4 and Ring 5 (10° to 20° from fixation), sparing the central fovea and parafovea until late stages. In Asian patients, screening requires widefield testing (e.g., 24-2/30-2 fields and wider mfERG arrays).

Other Major Clinical Indications

1. Stargardt Disease (ABCA4 Macular Dystrophy): Shows marked depression or complete extinction of central foveal and parafoveal mfERG responses (Rings 1–3), while peripheral responses and full-field ERG remain normal in early stages.
2. Best Vitelliform Macular Dystrophy: Demonstrates focal depression of central mfERG responses corresponding to the vitelliform yolk lesion, contrasting sharply with a completely normal full-field ERG and an extinguished electrooculogram (EOG).
3. Central Serous Chorioretinopathy (CSCR): Characterized by prolonged P1 implicit times and reduced response density over areas of neurosensory retinal detachment. Prolonged implicit times often persist for months to years after complete subretinal fluid reabsorption, explaining persistent subjective visual complaints despite 20/20 visual acuity.
4. Branch Retinal Vein Occlusion (BRVO): Shows sectorial reduction in P1 amplitude and marked implicit time delay confined strictly to the affected vascular quadrant, documenting the degree of capillary non-perfusion and ischemia.
5. Retinal Dystrophies (Cone Dystrophy): Generalized reduction in response density across all 5 concentric rings with marked P1 latency prolongation, often detecting cone dysfunction years before funduscopic changes appear.

---

Technologist Protocol, Patient Preparation, and Technical Artifacts

Achieving diagnostically valid mfERG recordings requires meticulous attention to patient alignment, electrode application, and optical clarity. Because mfERG relies on mathematical cross-correlation, technical errors destroy the recording.

Patient Preparation Checklist

1. Pupillary Mydriasis: Pupils must be maximally dilated (minimum 7 to 8 mm diameter) using 1% tropicamide and 2.5% phenylephrine. Constricted pupils reduce retinal illumination, artificially depressing amplitude densities and prolonging implicit times.
2. Photopic Adaptation: The patient must be light-adapted to standard clinical room luminance for at least 10 to 15 minutes before testing. Testing must never be performed immediately following dark adaptation or fundus photography (which bleaches retinal photopigments).
3. Refractive Correction at Testing Distance:
   • The patient must wear their exact optical refractive correction calibrated for the specific eye-to-monitor distance (typically 30 to 40 cm).
   • The Danger of Refractive Blur: Uncorrected refractive error blurs the edges of the flickering hexagons on the retina. Optical blur severely attenuates the high-spatial-frequency contrast of the stimulus, leading to a marked, artificial reduction in central response density.
   • Incorporate the patient's distance prescription plus an appropriate reading addition (typically +2.50 to +3.00 D for 33 cm working distance).

Electrode Selection: DTL Fiber vs. Contact Lens Electrodes

• DTL (Dawson-Trick-Litzkow) Fiber Electrode: The preferred gold standard for mfERG. A fine, conductive silver-impregnated nylon thread draped across the inferior conjunctival fornix. It provides excellent patient comfort, causes no corneal abrasion, and—crucially—does not induce optical distortion or corneal surface haze during the 10-minute testing session.
• Contact Lens Electrodes (e.g., Burian-Allen, ERG-Jet): While providing high signal amplitude, contact lenses can induce corneal epithelial edema, tear film disruption, and optical aberrations that blur the hexagonal array, compromising topographic resolution.
• Skin Electrodes: Placed at the outer canthus (reference) and forehead/earlobe (ground). Ensure skin impedances are balanced and below 5 kΩ.

Fixation Stability and Artifact Rejection

• Continuous Fixation Monitoring: The technologist must continuously monitor the patient's eye via the real-time infrared fixation camera. If the patient makes a saccade or loses central fixation, the stimulus sequence must be paused.
• Saccadic Smearing: If a patient fixates erratically, the small foveal hexagons will stimulate peripheral retina, and large peripheral hexagons will stimulate the fovea, completely smoothing out and invalidating the topographic map.
• Blink and Muscle Artifacts: Software artifact rejection thresholds discard segments contaminated by blinks (large voltage spikes) or myogenic noise (clenched jaw, squinting). Segments with >10% artifact rejection must be repeated.