Waveform Changes of the First-Order Multifocal Electroretinogram in Patients with Glaucoma

Abstract

purpose. To investigate the relationship between the components of the first-order multifocal electroretinogram (M-ERG) and glaucomatous visual field loss.

methods. Twenty-six eyes of 14 patients with primary open-angle glaucoma (POAG) were evaluated with the M-ERG techniques. Twenty-six eyes of 26 normal subjects also were tested as control subjects. To record the M-ERG, a stimulus matrix of 103 scaled hexagonal elements was displayed on a monitor driven at a 75-Hz frame rate according to a binary m-sequence. The M-ERG responses were averaged in each quadrant of the stimulus field and the peak-to-trough amplitudes and peak implicit times of the first trough (N1), the first peak (P1), and the second trough (N2) of the M-ERG were compared with the mean sensitivity values (dB) of the corresponding quadrant of the Humphrey static perimetric field.

results. The changes in the peak latencies of P1 and N2 in the POAG group were small but significant compared with those in the normal group (P < 0.01). However, no significant differences in the amplitudes of (P1–N1) and (P1–N2) between the two groups were found. Significant negative correlations between the peak implicit times of N1, P1, and N2 and the mean sensitivity values (dB) of static perimetry were observed. The correlation coefficients were −0.20 (P < 0.05) for the N1, −0.41 (P < 0.001) for the P1, and −0.59 (P < 0.001) for the N2. No significant correlations were observed between the amplitudes (P1–N1 and P1–N2) and the mean sensitivity values.

conclusions. The present study findings suggest that the peak implicit times, but not the amplitudes, of the M-ERG increase as the glaucomatous visual field deteriorates. The amplitudes of the M-ERG did not decrease as the glaucomatous optic nerve dysfunction progressed.

In a recent study, Sutter and Tran ¹ detailed a method for recording the multifocal electroretinogram (M-ERG). In this method, many retinal areas are independently stimulated according to a binary m-sequence, and local ERG responses are extracted from a continuous ERG recording using a cross-correlation technique. The multifocal technique allows a comparison between local retinal activity and perimetric sensitivity. Hood et al. ² compared the M-ERG and full-field ERGs. They concluded that the negative wave of the M-ERG was comprised of the same components as the a-wave of the full-field ERG and further that the positive wave of the M-ERG is a combination of the positive components of the full-field ERG. Although the M-ERG reportedly is useful in the detection of subclinical diabetic retinopathy, ³ little is known about the M-ERG components derived from patients with optic nerve disease. 

Retinal ganglion cells are lost in glaucoma as optic nerve damage progresses. ^{4 5} Goldmann’s perimetry and other visual field tests are commonly used to detect glaucomatous optic nerve damage. These methods are effective to a degree, but not sufficient for the detection of nerve dysfunction in the early stages of glaucoma. Pattern electroretinograms (P-ERGs) are reportedly a sensitive tool for the detection of ganglion cell dysfunction in patients with glaucoma. ⁶ However, the origins of the P-ERG are still being debated. The P-ERG was suggested in a few studies to be a multioriginated response. ^{7 8 9 10} P-ERGs may not yet be a sensitive enough tool to detect glaucomatous optic nerve dysfunction in the early stages of glaucoma. 

Bearse et al. ¹¹ reported that the first-order M-ERG from patients with advanced glaucoma decreased throughout much of the tested area. However, the sample in this study was very small. Further, there have been few reports as to quantitative relationship between the first-order M-ERG and sensitivity loss measured behaviorally with perimetry in patients with glaucoma. In the present study we investigate the relationship between the waveform changes of the first-order M-ERG and perimetric field loss. 

Methods

Subjects

Twenty-six eyes of 14 glaucoma patients (mean age = 47.2 ± 10.2 years) were analyzed in the present study. For the 12 patients in whom two eyes were tested, only one eye was chosen on the basis of the mean deviation (MD) value. Eyes with relatively slight damage were grouped as primary open-angle glaucoma (POAG)(A) and those with severe damage as POAG(B). The mean sensitivity was 24.1 ± 7.2 dB (mean ± SD) for POAG(A) and 15.6 dB (MD = 10.5) for POAG(B). Twenty-six eyes of 26 normal subjects (mean age = 46.1 ± 13.5 years) were tested as control subjects. Six eyes of three normal individuals also were entered into the study to compare the M-ERGs between the left and right eye. The mean ages of the two groups were not significantly different. All patients were diagnosed as having POAG. The mean period after establishment of the diagnosis was 25.6 ± 14.8 months (mean ± SD). The patients had clear medias and did not have any eye diseases other than glaucoma. Their corrected visual acuities were greater than 20/20. Patients who had previous eye surgery were excluded from the study. Pupils were maximally dilated more than 8 mm with tropicamide when the M-ERG was recorded. Refractive errors were within ±4 diopters. The MDs in the static perimetry ranged from 0 to −30 dB. The intraocular pressure of all patients with glaucoma was held at less than 22 mm Hg. 

Recordings

The Visual Evoked Response Imaging System (Tomey, Nagoya, Japan) was used to record the M-ERG. The stimulus matrix consisted of 103 scaled hexagonal elements displayed on a monochrome monitor (MD-B1700; Chu-ou musen, Tokyo, Japan) driven at a 75-Hz frame rate. At a viewing distance of 27 cm, the radius of the stimulus array subtended 25 × 30°. Each element was independently alternated between black (20 candelas [cd]/m²) and white (200 cd/m²) according to a binary m-sequence. ¹ Luminance of the surrounding region and of the fixation cross was 100 cd/m². The ERGs were recorded with a bipolar contact lens electrode (Kyoto Contact Lens, Kyoto, Japan) that was inserted after the cornea was anesthetized with oxybuprocaine hydrochloride. A ground electrode was attached to the earlobe. An individual recording session consisted of 8 or 16 segments separated by rests for a total recording time of 8 or 16 minutes. The quality of the recording was verified by monitoring the row data. After completion of each segment, subjects also were asked about the state of fixation during the recording. Inappropriate segments were discarded and rerecorded. The M-ERGs were recorded with low- and high-frequency cutoffs of 10 to 300 Hz, respectively. Subjects’ eyes were refracted to achieve the best visual acuity for the viewing distance, which was adjusted to compensate for changes in the retinal image size caused by the corrective lenses. A Humphrey static perimetry (program 30-2) also was performed to compare to the M-ERGs. The visual field subtended 30 × 30°. Background illumination of the perimeter was set at 31.5 apostilb (Asb). 

Analysis

We examined the first-order components of the M-ERGs using VERIS Science 3.0.1 (Electro Diagnostic Imaging, San Mateo, CA). The first-order component is defined as the difference between the mean response to all white and black frames in the sequence as described in detail in a previous study. ¹ Twenty-two focal responses for each quadrant (superior temporal, ST; superior nasal, SN; inferior temporal, IT; inferior nasal, IN) were averaged to improve the signal-to-noise ratio (Fig. 1C ), and four waveforms (M-ERGq) were obtained (Fig. 1D) . The first negative trough, the first positive peak, and the second negative trough were called N1, P1, and N2, respectively (Fig. 1D) . We measured implicit times of N1, P1, and N2 and peak-to-trough amplitudes of N1–P1 and P1–N2 of the M-ERGq (Fig. 1D) . 

The M-ERGq are displayed as the amplitude density (amplitude per unit of retinal area, nV/deg²). To compare with the M-ERGs, the mapping data of the visual fields were divided into four areas by a horizontal and vertical axis that passed through a central point of origin. Nineteen sensitivity values (dB) for each quadrant of the visual field were averaged and four mean sensitivity values were calculated (Fig. 1A) . 

Our investigation followed the tenets of the Declaration of Helsinki, and informed consent was obtained from all participants once the nature of the study had been clearly explained. 

Results

An Example of the M-ERG Derived from a Patient with Glaucoma

Figure 1 shows an example of the M-ERG and static perimetry obtained from a patient with unilateral glaucoma. Figure 1A shows the sensitivity map obtained with Humphrey static perimetry (program 30- to 2). There is a marked depression in the sensitivity value (dB, reciprocal of threshold) for the right eye (MD = −12.25), particularly in the upper half of the visual field. The mean sensitivity values (dB) for each quadrant are shown in Figure 1A (ST, 12.4; SN, 0; IT, 28.4; IN, +24.6). In contrast, the values obtained from the left eye were almost within the normal range (MD =+ 1.25). Fundus examination revealed that the right optic disc was severely damaged by glaucoma. Figure 1B shows the M-ERG trace arrays of 103 first-order M-ERGs from the right and left eyes of the patient shown in Figure 1A . No focal reductions of the M-ERG were found in spite of the sensitivity loss in the right visual field. Figure 1D shows the M-ERGq (nV/deg²) in each quadrant of the visual field (ST, SN, IT, and IN) from the patient in Figure 1A . The amplitudes did not decrease in the right eye in comparison with the left eye. The peak (P1 and N2) implicit time values of the M-ERGq in the right eye were high, particularly in those areas where the visual field was severely damaged (ST and SN quadrant). The latencies also were slightly increased in the IT quadrant. 

Statistical Difference of the Parameters of the M-ERG between POAG Patients and Normal Control Subjects

The mean values and the SDs of the parameters of the M-ERGq derived from each of the POAG groups were compared with those of normal control subjects over all quadrants (Figs. 2A 2B , Table 1 ). The mean implicit time values in the POAG(B) group were 15.5 ± 1.0 msec (mean ± SD) for N1, 29.1 ± 1.4 msec for P1, and 43.5 ± 1.8 msec for N2. In the POAG(A) group, the mean values of P1 (28.8 ms) and N2 (42.6 ms) were smaller than those in the POAG(B) group. Significant differences in the mean values were observed between the normal and POAG(B) groups for N2 (P < 0.001), P1 (P < 0.001), and N1 (P < 0.005). Significant differences in the values also were observed between the normal and POAG(A) groups for N2 (P < 0.01), P1 (P < 0.01), and N1 (P < 0.01). 

Mean values of the P1–N1 and P1–N2 in the POAG(A) group were smaller than those in normal individuals (Fig. 2B) . The mean values in the POAG(B) group were larger than those in the POAG(A) group. However, we found no significant differences in the mean amplitudes of P1–N1 and P1–N2 between the POAG groups and the normal control subjects. 

Statistical difference of the M-ERG parameters in each quadrant between the POAG groups and normal subjects are also shown in the Table 1 . In the latency measure, a significant difference was found in each quadrant for both N2 and P1. No significant difference in the amplitudes of (P1–N1) or (P1–N2) was found in any quadrant. 

Correlation between the M-ERG and Static Perimetry

Significant negative correlations were found between the peak latencies of the M-ERGq and the mean sensitivity values (dB) of static perimetry over all quadrants (Table 2 , Fig. 3 ). However, there are seven eyes with normal N2 latencies (<1 SD) with losses of more than 10 dB. The correlation coefficient was largest for the N2 latency (r = −0.590, n = 104, P < 0.001), followed by the P1 (r = −0.414, n = 104, P < 0.001) and the N1 latency (r = −0.200, n = 104, P < 0.05). The slope of the regression line was greatest for the N2 latency (slope = −0.101) followed by the P1 latency (slope =− 0.061). The intercepts (msec) of the regression lines were 15.9 for the N1 latency, 30.2 for the P1 latency, and 45.1 for the N2 latency. 

The N2 latencies also significantly correlated with the sensitivity values in each quadrant (ST: r = −0.627, n = 26, P < 0.001; SN: r = −0.745, n = 26, P < 0.001; IT: r = −0.540, n = 26, P < 0.01; and IN: r = −0.423, n = 26, P < 0.05). The P1 latencies correlated with the sensitivity values in only two quadrants (ST: r = −0.544, n = 26, P < 0.01; and SN: r = −0.575, n = 26, P < 0.01). There were no significant correlations observed between the N1 peak latencies and the sensitivity values in any quadrant (Table 2) . 

No significant correlations between the amplitudes (P1–N1 and P1–N2) and the mean sensitivity values (dB) of the static perimetry over all quadrants were found (Table 2) . We also did not find any significant correlation with the mean sensitivity values in any quadrant of ST, SN, IT and IN (Table 2) . Similarly, no significant correlation between the amplitude ratio (P1–N2/P1–N1) and the mean sensitivity value was observed (Table 2) . 

The Subtraction of the M-ERGq between the Right and Left Eye of Patients with Glaucoma and Normal Control Subjects

Figure 4C shows components (M-ERGd) obtained by subtracting the M-ERGx (Fig. 4A) from the M-ERGy (Fig. 4B) in ST quadrant, where M-ERGx is defined as the response from the ST quadrant with severe damage and M-ERGy defined as the response from the corresponding area in the contralateral eye with relative slight damage from the glaucoma. We selected those areas with interocular differences in the mean sensitivity of the visual fields that were larger than 10 dB. Eight patients met this criterion for quadrant ST. Figure 4D shows the same subtraction of the left M-ERGq from the right M-ERGq for three normal control subjects. The waveforms could not be distinguished from the noise in normal subjects (Fig. 4D) . 

In Figures 4A 4B 4C , one subject (P8) was recorded from on two occasions and good reproducibility was shown for each of the M-ERGy, M-ERGx, and M-ERGd waveforms. In glaucoma patients, the waveforms of the M-ERGd with polarities inverted against the M-ERGy were distinct (Fig. 4C) . The waveforms produced a very small response (<5 nV/deg²) and were a smaller delay compared to the M-ERGy. 

Subtraction of the M-ERGq between the SN and IN Quadrant

Figure 5 (lower two rows) shows the M-ERG waveforms from the SN and IN quadrant from patients (5 eyes of 4 patients) for whom the interquadrant differences (IN − SN) in the mean sensitivity of the visual fields was larger than 15 dB. (One subject was tested on two occasions.) The differences between the M-ERG from the SN and IN quadrant also are superimposed in the figure. The difference (M-ERGd) were obtained by subtracting the M-ERG of the SN quadrant from that of the IN quadrant in each case. The results from three normal control subjects are shown in the top row. 

There is a small difference in the M-ERGs from the SN and IN quadrants in the POAG patients. The changes were more distinct near N2. Although small, the amplitudes of the difference were larger in the POAG patients than in the normal subjects. The M-ERGd in the glaucoma patients resembled those of the difference between left and right eyes shown in the Figure 4 . The amplitude of the M-ERGd is also very small (<5 nV/deg²) in contrast with the large difference in the perimetric sensitivity value between the SN and IN quadrant. 

Discussion

The possibility that the ERGs from patients with optic nerve diseases exhibit an abnormal pattern has long been debated. Components of the ERG are commonly believed to be a product of the photoreceptor and middle retinal layers, suggesting that the flash ERG would be normal in cases of glaucoma that are not advanced and in which the outer retinal layers are not involved. On the other hand, abnormalities in oscillatory potentials ¹² have been reported in patients with glaucoma. In contrast, Wanger et al. ¹³ reported that the amplitudes of the flash ERGs and oscillatory potentials were not reduced in patients with unilateral glaucoma. 

Bearse et al. ^{14 15} reported an abnormal second-order kernel of the M-ERG in glaucoma. Sutter and Bearse ¹⁶ detailed a method for extracting an optic nerve head component (ONHC) from the MERG using a latency-adjusted algorithm and the assumption that a retinal component was invariant over retinal location. They hypothesized that the ONHC was generated by retinal ganglion cells and suggested that it was either reduced in amplitude or completely extinguished in patients with glaucoma and optic atrophy. However, their method for extracting the ONHC is not easily implemented and the assumption on which it rests has yet to be validated. 

In the present study, we compared the M-ERG with visual fields quantitatively, using 14 glaucoma patients and normal control subjects. The peak implicit times of N1, P1, and N2 of the first-order M-ERG increased with increasing glaucomatous visual field loss. The slope of the regression line was larger for the N2 latency (slope = −0.16) than for the P1 latency (slope = −0.10). In retinitis pigmentosa, timing changes appeared to be an early indication of local retinal damage of the cone system, with latency changes appearing before the local region has lost 5 dB of sensitivity. ¹⁷ In our study, a loss of more than 10 dB was necessary before the points fell out of the normal range, even in the case of the N2 latency. Thus, peak latency is not a sensitive enough tool for detecting optic nerve dysfunction. We hypothesized that the mechanism involved in the timing changes of optic nerve dysfunction was different from that of diseases of the outer retina. 

The amplitudes of the M-ERGs did not correlate significantly with mean sensitivity values of the static perimetry. At any rate, we did not perceive any tendency for the M-ERG amplitude to decrease as the visual field deteriorated. The amplitudes obtained for sensitivity values were widely divergent. This may be due to large interindividual variation in amplitude. Comparing interocular or intraocular differences in unilateral glaucoma subjects may allow for clearer results. 

By definition, subtracting the M-ERGx from the M-ERGy should produce the component M-ERGd, which changes with glaucomatous optic nerve dysfunction. The M-ERGd was a small component and was inverted in polarity against the M-ERGy. The trough and peak of the M-ERGd is slightly delayed compared with the peak (P1) and trough (N2) of the M-ERGy. Bearse et al. ¹⁵ previously reported that first-order kernel contains an ONHC that is inverted in polarity relative to the retinal component. Their finding is consistent with that of the present study. Thus, optic atrophy theoretically should lead to an increase in M-ERGd and a subtle change in the first-order M-ERG. However, the large interindividual variations in amplitudes and the small size of the M-ERGd probably obscure this change in the amplitude of the M-ERG. The interocular and interquadrant comparison in unilateral glaucoma reveals that the amplitude was not significantly decreased by glaucoma (Figs. 4 5) . Recently, Hood et al.²² described a tetrodotoxin (TTX)-sensitive component in the primate M-ERG. In particular, they showed that the M-ERG after TTX injection became more invariant in waveform and larger in amplitude. Their results are consistent with the present study. We hypothesize that the M-ERGd is an average response of the ganglion cell activity and optic nerve potential. Further studies are required to clarify the properties and origin of the response. 

In conclusion, the latency and amplitude of the M-ERG changed with glaucomatous optic nerve dysfunction. A subtle change in the M-ERG of glaucoma patients can be detected by subtraction of the M-ERG with severe pathology from that with normal one. This subtle change may provide more sensitive tool to detect optic nerve dysfunction. 

 Submitted for publication September 14, 1999; revised November 22, 1999; accepted December 1, 1999.

 Commercial relationships policy: N.

 

Figure 1.

 

(A) Sensitivity maps of the static perimetry from a patient with unilateral glaucoma. MD, mean deviation (dB); fovea, sensitivity value (dB) at fovea. Visual field was divided into four quadrants (ST, SN, IT, and IN). Mean sensitivity values of the right eye show marked decrease in SN (0 dB) and ST (12.4 dB). (B) The trace array of the M-ERG from the same patient in Figure 1A . No changes are perceived in contrast to the visual field loss as is shown in Figure 1A . (C) The trace array of the M-ERG was divided into four quadrants by the horizontal and vertical axes. Fifteen focal traces on the axis were excluded from the area, so that each of the quadrants contains 22 local responses. (D) The M-ERG averaged response in each quadrant from the same patient in Figure 1A . The peak latencies of N1, P1, and N2 are displayed in the right column of the figure.

Figure 2.

 

(A) The mean and SD of the peak latencies of N1, P1, and N2 in the normal, POAG(A), and POAG(B) groups. *P < 0.01, **P < 0.005, ***P < 0.001. (B) The mean and SD of the amplitudes of P1–N1 and P1–N2 in the group of normal subjects, POAG(A), and POAG(B).

Table 1.

 

Parameters of the M-ERGq of POAG Groups Compared with Normal Control Subjects

Quadrant	M-ERG	Normal, †	POAG(A), †	POAG(B), †	P *
										N–A	N–B
All	N1	14.9 ± 1.1	15.5 ± 1.3	15.5 ± 1.0	<0.01	<0.005
	P1	27.6 ± 1.2	28.8 ± 1.5	29.1 ± 1.4	<0.001	<0.001
	N2	41.4 ± 1.1	42.6 ± 1.7	43.5 ± 1.7	<0.001	<0.001
ST	N1	15.0 ± 1.2	15.3 ± 1.4	15.3 ± 0.9	NS	NS
	P1	27.7 ± 1.2	28.6 ± 1.5	29.2 ± 1.7	NS	<0.01
	N2	41.4 ± 1.1	42.7 ± 1.7	43.6 ± 1.8	<0.05	<0.001
SN	N1	14.8 ± 0.9	15.4 ± 1.2	15.1 ± 1.1	NS	NS
	P1	27.3 ± 1.2	28.5 ± 1.4	28.9 ± 1.3	<0.02	<0.002
	N2	41.1 ± 1.1	42.7 ± 1.9	43.5 ± 2.1	<0.02	<0.002
IT	N1	15.0 ± 1.1	15.6 ± 1.2	15.4 ± 1.1	NS	NS
	P1	27.8 ± 1.2	29.2 ± 1.6	29.2 ± 1.1	<0.02	<0.02
	N2	41.6 ± 1.1	42.8 ± 1.7	43.5 ± 1.6	<0.05	<0.002
IN	N1	14.9 ± 1.1	15.6 ± 1.5	16.0 ± 0.9	NS	0.005
	P1	27.4 ± 1.2	29.0 ± 1.7	29.0 ± 1.6	<0.01	<0.01
	N2	41.3 ± 0.9	42.1 ± 1.6	43.3 ± 1.6	NS	<0.001
All	P1–N1	16.5 ± 4.5	15.5 ± 4.6	16.8 ± 4.1	NS	NS
	P1–N2	16.5 ± 5.2	15.1 ± 4.8	15.9 ± 5.6	NS	NS
ST	P1–N1	15.6 ± 4.2	14.3 ± 4.5	16.0 ± 4.2	NS	NS
	P1–N2	15.4 ± 4.6	13.8 ± 4.7	14.9 ± 5.5	NS	NS
SN	P1–N1	17.7 ± 4.5	15.7 ± 4.7	17.6 ± 4.2	NS	NS
	P1–N2	17.7 ± 5.4	15.7 ± 4.9	16.9 ± 5.6	NS	NS
IT	P1–N1	15.5 ± 4.5	15.3 ± 4.9	16.2 ± 3.8	NS	NS
	P1–N2	15.4 ± 5.1	14.7 ± 5.3	14.8 ± 5.3	NS	NS
IN	P1–N1	17.2 ± 4.7	16.6 ± 4.7	17.3 ± 4.6	NS	NS
	P1–N2	17.5 ± 5.4	16.0 ± 4.5	16.9 ± 6.1	NS	NS

 Values are means ± SD. ;cell>NS, not significant; N, Normal; A, POAG(A); B, POAG(B).

*  Determined by unpaired t-test.

†  See the Methods section for details.

Table 2.

 

Correlations between Peak Latencies of the M-ERG-q and Mean Sensitivity Values

	Quadrant	M-ERG	Correlation
						Regression Line			Statistics
						Slope	Intercept	Coefficient	P	n
Latency	All	N1	−0.02	15.9	−0.20	<0.05	104
		P1	−0.06	30.2	−0.41	<0.001	104
		N2	−0.11	45.1	−0.59	<0.001	104
	ST	N1	−0.08	16.8	−0.18	NS	26
		P1	−0.09	30.7	−0.54	<0.01	26
		N2	−0.12	45.5	−0.63	<0.001	26
	SN	N1	0.01	15.5	−0.14	NS	26
		P1	−0.07	29.9	−0.58	<0.01	26
		N2	−0.13	45.5	−0.75	<0.001	26
	IT	N1	−0.03	16.2	−0.24	NS	26
		P1	−0.04	30.0	−0.24	NS	26
		N2	−0.10	45.2	−0.54	<0.01	26
	IN	N1	−0.02	16.3	−0.20	NS	26
		P1	−0.06	30.1	−0.35	NS	26
		N2	−0.07	44.1	−0.42	<0.05	26
Amplitude	All	P1–N1	−0.03	16.7	−0.06	NS	104
		P1–N2	−0.01	15.7	−0.02	NS	104
	ST	P1–N1	−0.08	16.8	−0.18	NS	26
		P1–N2	−0.08	15.9	−0.14	NS	26
	SN	P1–N1	−0.05	17.6	−0.14	NS	26
		P1–N2	−0.02	16.6	−0.04	NS	26
	IT	P1–N1	−0.03	16.3	−0.06	NS	26
		P1–N2	0.02	14.6	0.02	NS	26
	IN	P1–N1	0.07	15.6	0.16	NS	26
		P1–N2	0.06	15.2	0.12	NS	26
Ratio*	All	N/R	0.002	0.09	0.10	NS	104

 Ratio = P1–N1/P1–N2 (N = P1–N2, R = P1–N1). The most significant correlation is observed between N2 peak latency and the mean sensitivity value in each area.

Figure 3.

 

Correlations between the peak latencies of N1, P1, and N2 of the M-ERG and the mean sensitivity values over all quadrants of visual field. Negative correlations are observed.

Figure 4.

 

(A) M-ERGy: averaged responses of the M-ERG derived from the ST quadrant of visual field with relative slight damage. (B) M-ERGx: averaged responses of the M-ERG derived from the ST quadrant with severe damage in the visual field (contralateral eye of the M-ERGy). These quadrants were selected whose interocular differences of the MD were more than 10 dB. (C) The traces after subtracting the M-ERGx from the M-ERGy. The waveforms inverted in polarity and slightly delayed in peaks compared with the M-ERGy. (D) Three traces produced by subtraction of the right M-ERGq from the left M-ERGq in three normal subjects. MS, mean sensitivity (dB); RE, right eye; LE, left eye; epoch time = 80 msec; 1 div = 10 nV/deg².

Figure 5.

 

The subtraction of the M-ERG between the SN and IN quadrants. Top row: M-ERGs from 3 normal control subjects. Bottom two rows: M-ERGs from 5 eyes of 4 glaucoma patients. 1 div = 5 nV/deg². A good reproducibility of the M-ERG waveform can be perceived in one subject (P8). RE, right eye; LE, left eye; epoch time = 80 msec. Top two rows: 1 div = 5 nV/deg²; bottom row: 1 div = 2 nV/deg².