A component analysis of the electroretinogram

Enhancing Electroretinogram Classification with Multi-Wavelet Analysis and Visual Transformer

The electroretinogram (ERG) is a clinical test that records the retina's electrical response to light. Analysis of the ERG signal offers a promising way to study different retinal diseases and disorders. Machine learning-based methods are expected to play a pivotal role in achieving the goals of retinal diagnostics and treatment control. This study aims to improve the classification accuracy of the previous work using the combination of three optimal mother wavelet functions. We apply Continuous Wavelet Transform (CWT) on a dataset of mixed pediatric and adult ERG signals and show the possibility of simultaneous analysis of the signals. The modern Visual Transformer-based architectures are tested on a time-frequency representation of the signals. The method provides 88% classification accuracy for Maximum 2.0 ERG, 85% for Scotopic 2.0, and 91% for Photopic 2.0 protocols, which on average improves the result by 7.6% compared to previous work.

Flicker cone function in normal and day blind sheep: a large animal model for human achromatopsia caused by CNGA3 mutation

PURPOSE

Recently we reported on day blindness in sheep caused by a mutation in the CNGA3 gene, thus making affected sheep a naturally occurring large animal model for therapeutic intervention in CNGA3 achromatopsia patients. The purpose of this study was to characterize flicker cone function in normal and day blind sheep, with the aim of generating a normative data base for ongoing gene therapy studies.

METHODS

Electoretinographic (ERG) cone responses were evoked with full-field conditions in 10 normal, 6 heterozygous carriers and 36 day blind sheep. Following light adaptation (10 min, 30 cd/m(2)), responses were recorded at four increasing light intensities (1, 2.5, 5 and 10 cd s/m(2)). At each of these intensities, a single photopic flash response followed by 8 cone flicker responses (10-80 Hz) was recorded. Results were used to generate a normative data base for the three groups. Differences between day blind and normal control animals were tested in two age-matched groups (n = 10 per group).

RESULTS

The normal sheep cone ERG wave is bipartite in nature, with critical flicker fusion frequency (CFF) >80 Hz. In all four flash intensities, the single photopic flash a-wave and b-wave amplitudes were significantly lower (p < 0.005), and implicit times significantly delayed (p < 0.0001), in day blind animals. In all four flash intensities, CFF values were significantly lower (p < 0.0001) in day blind sheep.

CONCLUSIONS

Cone function is severely depressed in day blind sheep. Our results will provide a normative data base for ongoing gene therapy studies.

Structural and functional maturation of the retina of the albino Hartley guinea pig

PURPOSE

Altricial animals, such as rats and mice, are born with their eyes closed, compared to precocial animals, such as guinea pigs and humans, which have their eyes opened at birth. The purpose of this study was to investigate if the retina of guinea pigs (precocial animal) is subjected to a postnatal maturation process similar to that previously reported for rodents.

METHODS

Photopic and scotopic electroretinograms (ERG) and retinal histology were obtained from albino guinea pigs aged P1 to P75.

RESULTS

Photopic ERG responses reached maximal amplitudes at P5 (a-and b-waves), that is 5 days (b-wave) to 10 days (a-wave) earlier than scotopic responses. However, the postnatal gain in b-wave amplitude was significantly (P < 0.05) more important for the cone (73.38 +/- 4.4%) signal than for the rod (15.23 +/- 3.96%), suggesting that the rod function is more mature at birth. Similarly, the short latency photopic oscillatory potential (ie: OP2) reached its maximal value 5 days (P10) earlier than its scotopic equivalent (P15), while the long latency OPs (ie: OP3, OP4), reached their maximal values nearly 20 days sooner in scotopic condition. Finally retinal histology revealed a thinning of the retina with age, the latter being most pronounced at the level of the ganglion cell layer (GCL).

CONCLUSION

Our results thus confirm that despite its relative maturity at birth (compared to rodents), the retina of newborn albino guinea pigs undergoes significant postnatal maturation modifying its structure as well as its function, albeit not as extensive as that previously documented for altricial animals.

Two signals in the human rod visual system: a model based on electrophysiological data

In the human rod visual system, self-cancellation of flicker signals is observed at high rod intensity levels near 15 Hz, both perceptually and in the electroetinogram (ERG). This and other evidence suggests that two rod signals are transmitted through the human retina with different speeds of transmission. Here we report a series of flicker ERG recordings from a normal observer and an observer who lacks cone vision. From these results, we propose a quantitative model of the two rod signals, which assumes (1) that the amplitude of the slow signal grows linearly with log intensity but then saturates at approximately 1 scot. td; (2) that the amplitude of the fast signal grows linearly with intensity; (3) that there is a difference in time delay of approximately 33 ms between two rod signals of the same polarity (or of approximately 67 ms if the signals are of inverted polarity); and (4) that the time delay of both signals declines linearly with log intensity (by approximately 10 ms per log scot. td). These simple assumptions provide a remarkably good account of the experimental data. Our results and model are relevant to current anatomical theories of the mammalian rod visual system. We speculate that the slower signal in the human ERG may reflect the transmission of the rod response via the rod bipolars and the AII amacrine cells, while the faster signal may reflect its transmission via the rod-cone gap junctions and the cone bipolars. There are, however, several objections to this simple correspondence.

The late positive retinal potential in dogs

The late positive potential of the mammalian electroretinogram has been called the 'PI' or the 'c-wave' potential. It is unusual among retinal potentials because its peak implicit time increases in response to increasing stimulus intensity and because it cannot be demonstrated consistently in small samples of normal humans or normal dogs. We recorded wideband (DC-1 kHz) responses from 34 normal Beagles or dogs of similar size. Of the 34, 11 produced a late positive potential set that satisfied the criteria for c-waves. Multiple aspartate injections always increased c-wave amplitude and stimulus-response linearity in all 'producers'. Non-producers were never converted to producer status by aspartate blocking of the inner retina. Interaction of late positive and negative potentials and the possible influence of normal individual variations in the trans-epithelial potential are discussed. Individual mammal c-wave production is controlled by outer retinal phenomena which vary between individuals.

Contributions to the electroretinogram of currents originating in proximal retina

We have investigated responses in proximal retina of the cat that contribute to two kinds of electroretinogram (ERG) recordings: (1) the pattern ERG, a light-adapted response and (2) the threshold and near threshold ERG, a dark-adapted response (Sieving et al., 1986a, 1986b; Sieving & Steinberg, 1985). In intraretinal, extracellular recordings, two negative-going responses were identified that are maximal around the inner plexiform layer, and distinct from PII, which is maximal in distal retina: under light-adapted conditions, a spatially tuned response at light and light offset, the "M-wave" (previously described in cold-blooded animals by Karwoski & Proenza (1977, 1980)), and under dark-adapted conditions, the scotopic threshold response, or "STR," a response at light onset. The results under dark-adapted conditions are examined in more detail here. The STR is a very sensitive response whose threshold is 1.5-2.0 log units below that of the dc-component of PII and therefore well below the threshold of the a-, b-, and c-waves. It saturates about 2.4 log units below rod saturation. The STR contributes a negative-going potential to the dark-adapted ERG that is dominant near threshold; while PII (dc-component and b-wave) contributes a positive-going potential that is dominant at higher intensities (Sieving et al., 1986b). Investigation of the mechanism of the proximal retinal responses that contribute to the ERG supports of K(+)-Müller cell hypothesis of their origin.

Origin of the electroretinogram in the intact macaque eye--I. Principal component analysis

Responses to 400 msec flashes of white light were recorded at various depths in the retina of the intact macaque eye. A statistical technique, principal component analysis (PCA), was used to isolate independent components from the LERG, using the changing contribution of a cell's response to the LERG with retinal depth. Two approaches were evaluated: first, PCA was performed on the complete LERG waveforms and, second, on small sections of the LERG. The first method yielded a component strongly resembling the receptor response, apart from some inconsistencies in the proximal layers, and a component in which the b-wave was the most prominent feature, but which still contained some other responses. Hence, the components were not suitable to describe the LERG in terms of responses of specific cell types. The second method uses PCA to determine whether or not the potential change within a small time window is accounted for by a single component. This method, which essentially uses the knowledge that different cell types respond with different delay times, yielded three components. These were identified as the receptor component, the b-wave and the d.c.-component. The voltage-depth profiles of these components were assessed.

Effect of short-term intraocular pressure increase on cat retinal ganglion cell activity

By means of microelectrode recordings receptive field properties and flicker responses of cat retinal ganglion cells were studied during elevated intraocular pressure (IOP). At moderate IOP elevation (perfusion pressure, PP, approx. 50 mm Hg), a slight increase in the maintained activity and the neuronal activation obtained from the receptive field (RF)-center or RF-periphery occurred in some of the neurons. At critical perfusion pressure (PP 30-10 mm Hg), the RF-periphery mechanisms were less sensitive to ischemia than the RF-center mechanisms. The spontaneous activity was completely suppressed at an average PP of 36.6 +/- 9.0 mm Hg, the flicker-evoked responses at 17.7 +/- 9.6 mm Hg. At critical perfusion pressures, the stimulus-response latency increased by 20 ms at stimulation frequencies of 10-20 Hz. The critical flicker frequency was already slightly reduced at a PP of 50 mm Hg and decreased further with a further decrease in PP. The threshold values of the intensity function shifted to higher luminance levels when PP was reduced to less than 40 mm Hg, but the slope remained constant. A close interneuronal correlation was found in simultaneous recordings of pairs of ganglion cells during different levels of increased IOP. This finding indicates that the interneuronal response variability to IOP increase appearing in successively recorded neurons was predominantly caused by experimental factors. The main impairment in retinal neuronal function during increased IOP seems to occur proximal to the receptors but distal to the retinal ganglion cell somata.

[X-linked night-blindness: clinical and electrophysiological investigations including de-ERG in two families (author's transl)]

Patients from two families with X-linked recessive nyctalopia and myopia were investigated. The functional tests included clinical examinations (visual acuity, color sense, visual fields, dark adaptation) and electrophysiological examinations (ac coupled ERG, dc-coupled ERG, EOG). The dc-coupled ERG registrations were done without general anesthesia. In one case, a comparison with results obtained in 1967 was possible. There was no evidence of increase in functional impairment. On the basis of the electrophysiological results, the nature and localization of the defect causing nyctalopia are discussed. The postreceptoral retinal structures appear not to be involved exclusively.

A comparative study of b- and d-waves of the electroretinogram, in frogs (Rana catesbeiana)

The aim of this study was to compare the b- and d-waves of the ERG, evoked respectively by ON and OFF stimuli. The experiments were carried out on frogs. The d-wave is facilitated by light adaptation and relatively high stimulus intensity. The depth profiles of b- and d-waves are different. Aspartic acid abolished only the b-wave whereas nembutal affected the d-wave. These results suggested that b- and d-waves were produced by separate neuronal mechanisms, even though both appeared as positive deflection in the corneal ERG and were sensitive to K+ ion.

Changes in ultrastructure and function of the sheep pigment epithelium and retina induced by sodium iodate. II. Early effects

The present investigation shows that the membrane properties of the sheep pigment epithelial cells were very rapidly and severely affected by sodium iodate, whereas the effects concerning the neuroretina were delayed. The c-wave of the ERG was immediately abolished and replaced by a cornea-negative potential, but the a- and b-waves were preserved for about 80-100 min. Ultrastructurally the plasma membranes (particularly the basal plasma membrane) of the pigment epithelial cells were destroyed or less distinct than normally. The cell organelles were swollen and ruptured. There were indications that the pigment epithelium could no longer participate in the receptor outer segment turnover. The photoreceptor cells were morphologically undamaged, and few or no signs of injury were observed in the inner layers of the retina. The effects upon the neuroretinal functions seen after 80-100 min, consisting of a reduction of alpha- and beta-wave amplitudes, were most likely caused by an inability of the pigment epithelium to maintain in the long run its metabolic and barrier properties. It appears that at an early stage after sodium iodate injection, the present preparation may be useful for the study of the effects on the neuroretina proper of drugs and other agents.

A new wave (2nd c-wave) on corneoretinal potential

The 2nd c-wave is a new wave of corneoretinal potential which is an on-response with a long latency (65-98 sec), and appears following the end of the c-wave of ERG. It suggested that the 2nd c-wave is based on the tail of the late receptor potential of the retina.

ERG of humans without C-wave

Three out of 18 healthy persons investigated with the DC-ERG do not show a c-wave. By varying the adaptation levels and intensities of the stimuli four components in the human ERG can be isolated, the cone and rod late receptor potentials, the positive DC response, and a c-wave in most subjects. It is suggested that investigation of these components reveals a more detailed understanding of the pathophysiologic mechanisms in some retinal diseases.

Experiments concerning the human C-wave

The dependence of the human c-wave from the step amplitude was studied qualitatively. The maximum of the c-wave is achieved with light stimuli longer than 10 sec. After a dark period of 10 sec a fully developed c-wave appears. Also, in response to a brilliant photoflash a c-wave could be recorded. The influence of a mobile pupil on the response in the DC-ERG is demonstrated. Responses of the retinal potential to square wave and sinusoidal stimuli of different period times have been registered.

Melanotropic drugs and retinal functions. I. Effects of quinine and chloroquine on the sheep ERG

The present work is the first in a series of investigations on the effects of melanotropic drugs on retinal function. Teh changes in the conventional ERG of the dark-adapted, intact sheep eye were studied after iv administrations of quinine and chloroquine. Both drugs influenced the ERG in principally the same way. Small doses changed only the c-wave amplitude which began to oscillate after a delay of about 1 hour. Larger doses had an immediate effect on both the b- and c-waves. The b-wave amplitude dropped and the c-wave showed an initial peak followed by oscillatory changes about 1 h later. These late changes were similar to those observed after administration of small doses of the drugs. The present results indicate that small doses of quinine and chlorquine have a delayed effect on the pigment epithelial cells, while larger doses exert a more immediate and general retinotoxic effect.

Melanotropic drugs and retinal functions. II. Effects of phenothiazine and rifampicin on the sheep ERG

The acute effects of chlorpromazine, promethazine and rifampicin on the a-, b- and c-waves of the conventional electroretinogram (ERG) were studied in sheep. Iv administration of chlorpromazine and promethazine resulted in a b-wave amplitude decrease and an initial c-wave amplitude decrease, followed by cyclic amplitude changes resembling damped oscillations. Iv injections of rifampicin, however, resulted in cyclic changes of the c-wave amplitude without initial concomitant b-wave changes. The results of the present study indicate that rifampicin has a selective influence on the pigment epithelial cells, while chlorpromazine and promethazine seem to have more generalized retinal effects on both the neuroretina and the pigment epithelial cells.