A Temporal White Noise Analysis for Extracting the Impulse Response Function of the Human Electroretinogram

Melanopsin-mediated amplification of cone signals in the human visual cortex

The ambient daylight variation is coded by melanopsin photoreceptors and their luxotonic activity increases towards midday when colour temperatures are cooler, and irradiances are higher. Although melanopsin and cone photoresponses can be mediated via separate pathways, the connectivity of melanopsin cells across all levels of the retina enables them to modify cone signals. The downstream effects of melanopsin-cone interactions on human vision are however, incompletely understood. Here, we determined how the change in daytime melanopsin activation affects the human cone pathway signals in the visual cortex. A 5-primary silent-substitution method was developed to evaluate the dependence of cone-mediated signals on melanopsin activation by spectrally tuning the lights and stabilizing the rhodopsin activation under a constant cone photometric luminance. The retinal (white noise electroretinogram) and cortical responses (visual evoked potential) were simultaneously recorded with the photoreceptor-directed lights in 10 observers. By increasing the melanopsin activation, a reverse response pattern was observed with cone signals being supressed in the retina by 27% (p = 0.03) and subsequently amplified by 16% (p = 0.01) as they reach the cortex. We infer that melanopsin activity can amplify cone signals at sites distal to retinal bipolar cells to cause a decrease in the psychophysical Weber fraction for cone vision.

Single opsin driven white noise ERGs in mice

Purpose

Electroretinograms elicited by photopigment isolating white noise stimuli (wnERGs) in mice were measured. The dependency of rod- and cone-opsin-driven wnERGs on mean luminance was studied.

Methods

Temporal white noise stimuli (containing all frequencies up to 20 Hz, equal amplitudes, random phases) that modulated either rhodopsin, S-opsin or L*-opsin, using the double silent substitution technique, were used to record wnERGs in mice expressing a human L*-opsin instead of the native murine M-opsin. Responses were recorded at 4 mean luminances (MLs).Impulse response functions (IRFs) were obtained by cross-correlating the wnERG recordings with the corresponding modulation of the photopigment excitation elicited by the stimulus. So-called modulation transfer functions (MTFs) were obtained by performing a Fourier transform on the IRFs.Potentials of two repeated wnERG recordings at corresponding time points were plotted against each other. The correlation coefficient (r2repr) of the linear regression through these data was used to quantify reproducibility. Another correlation coefficient (r2ML) was used to quantify the correlations of the wnERGs obtained at different MLs with those at the highest (for cone isolating stimuli) or lowest (for rod isolating stimuli) ML.

Results

IRFs showed an initial negative (a-wave like) trough N1 and a subsequent positive (b-wave like) peak P1. No oscillatory potential-like components were observed. At 0.4 and 1.0 log cd/m2 ML robust L*- and S-opsin-driven IRFs were obtained that displayed similar latencies and dependencies on ML. L*-opsin-driven IRFs were 2.5-3 times larger than S-opsin-driven IRFs. Rhodopsin-driven IRFs were observed at -0.8 and - 0.2 log cd/m2 and decreased in amplitude with increasing ML. They displayed an additional pronounced late negativity (N2), which may be a correlate of retinal ganglion cell activity.R2repr and r2ML values increased for cones with increasing ML whereas they decreased for rods. For rhodopsin-driven MTFs at low MLs and L*-opsin-driven MTFs at high MLs amplitudes decreased with increasing frequency, with much faster decreasing amplitudes for rhodopsin. A delay was calculated from MTF phases showing larger delays for rhodopsin- vs. low delays for L*-opsin-driven responses.

Conclusion

Opsin-isolating wnERGs in mice show characteristics of different retinal cell types and their connected pathways.

Stimulator compensation and generation of Gaussian noise stimuli with defined amplitude spectra for studying input–output relations of sensory systems

Gaussian noise is an important stimulus for the study of biological systems, especially sensory and neural systems. Since these systems are inherently nonlinear, the properties of the noise strongly influence the outcome of the analysis. Therefore, it is crucial to use a well-defined and controlled noise stimulus. In this paper, we first use the example of an insect filiform sensillum, a simple mechanoreceptor with a single sensory cell, to show that changes in the amplitude and spectral properties of the noise stimulus indeed affect the linear transfer function of the sensillum. We then explain step-by-step how to use the inverse fast Fourier transform to generate a Gaussian noise that has an arbitrary user-defined amplitude spectrum, including a band-limited white noise with a perfectly sharp cutoff edge. Finally, we demonstrate how such a perfect band-limited Gaussian white noise stimulus can also be generated with a non-perfect stimulator using a simple procedure that compensates for the filtering properties of the stimulator. With this approach, one can generate well-defined Gaussian noise stimuli that can be adapted to any application. For example, one can generate visual, sound, or vibrational stimuli for experimental research in visual physiology, auditory physiology, and biotremology, as well as inputs for testing various models in theoretical research.

Luminance white noise electroretinograms (wnERGs) in mice

Purpose

To record and analyse electroretinograms (ERGs) to luminance stimuli with white noise temporal profiles in mice. White noise stimuli are expected to keep the retina in a physiologically more natural state than, e.g., flashes. The influence of mean luminance (ML) was studied.

Methods

Electroretinograms to luminance temporal white noise (TWN) modulation (wnERGs) were measured. The white noise stimuli contained all frequencies up to 20 Hz with equal amplitudes and random phases. Responses were recorded at 7 MLs between -0.7 and 1.2 log cd/m². Impulse response functions (IRFs) were calculated by cross correlating the averaged white noise electroretinogram (wnERG) responses with the stimulus. Amplitudes and latencies of the initial trough and subsequent peak in the IRFs were measured at each ML. Fourier transforms of the IRFs resulted in modulation transfer functions (MTFs). wnERGs were averaged across different animals. They were measured twice and the responses at identical instances in the 1st and 2nd recordings were plotted against each other. The correlation coefficient (r ² _repr) of the linear regression quantified the reproducibility. The results of the first and second measurement were further averaged. To study the underlying ERG mechanisms, the ERG potentials at the different MLs were plotted against those at the lowest and highest ML. The correlation coefficients (r ² _ML) were used to quantify their similarities.

Results

The amplitudes of the initial (a-wave-like) trough of the IRFs increased with increasing ML. The following positive (b-wave-like) peak showed a minimum at -0.4 log cd/m² above which there was a positive correlation between amplitude and ML. Their latencies decreased monotonously with increasing ML. In none of the IRFs, oscillatory potential (OP)-like components were observed. r ² _repr values were minimal at a ML of -0.1 log cd/m², where the MTFs changed from low-pass to band-pass. r ² _ML values increased and decreased with increasing ML when correlated with responses obtained at the highest or the lowest ML, respectively.

Conclusion

White noise electroretinograms can be reliably recorded in mice with luminance stimuli. IRFs resemble flash ERGs superficially, but they offer a novel procedure to study retinal physiology. New components can be described in the IRFs. The wnERGs are either rod- or cone-driven with little overlap.

Electroretinographic responses to luminance and cone-isolating white noise stimuli in macaques

Electroretinograms (ERGs) are mass potentials with a retinal origin that can be measured non-invasively. They can provide information about the physiology of the retina. Often, ERGs are measured to flashes that are highly unnatural stimuli. To obtain more information about the physiology of the retina, we measured ERGs with temporal white noise (TWN) stimuli that are more natural and keep the retina in a normal range of operation. The stimuli can be combined with the silent substitution stimulation technique with which the responses of single photoreceptor types can be isolated. We characterized electroretinogram (ERG) responses driven by luminance activity or by the L- or the M-cones. The ERGs were measured from five anesthetized macaques (two females) to luminance, to L-cone isolating and to M-cone isolating stimuli in which luminance or cone excitation were modulated with a TWN profile. The responses from different recordings were correlated with each other to study reproducibility and inter-individual variability. Impulse response functions (IRFs) were derived by cross-correlating the response with the stimulus. Modulation transfer functions (MTFs) were the IRFs in the frequency domain. The responses to luminance and L-cone isolating stimuli showed the largest reproducibility. The M-cone driven responses showed the smallest inter-individual variability. The IRFs and MTFs showed early (high frequency) components that were dominated by L-cone driven signals. A late component was equally driven by L- and M-cone activity. The IRFs showed characteristic similarities and differences relative to flash ERGs. The responses to TWN stimuli can be used to characterize the involvement of retinal cells and pathways to the ERG response. It can also be used to identify linear and non-linear processes.

The influence of temporal frequency and stimulus size on the relative contribution of luminance and L-/M-cone opponent mechanisms in heterochromatic flicker ERGs

PURPOSE

To study the effect of stimulus size and temporal frequency on the relative contribution of luminance and L-/M-cone opponent signals in the ERG.

METHODS

In four healthy, color normal subjects, ERG responses to heterochromatic stimuli with sinusoidal, counter-phase modulation of red and green LEDs were measured. By inverse variation of red and green contrasts, we varied luminance contrast while keeping L-/M-cone opponent chromatic contrast constant. The first harmonic components in the full field ERGs are independent of stimulus contrast at 12 Hz, while responses to 36 Hz stimuli vary, reaching a minimum close to isoluminance. It was assumed that ERG responses reflect L-/M-cone opponency at 12 Hz and luminance at 36 Hz. In this study, we modeled the influence of temporal frequency on the relative contribution of these mechanisms at intermediate frequencies, measured the influence of stimulus size on model parameters, and analyzed the second harmonic component at 12 Hz.

RESULTS

The responses at all frequencies and stimulus sizes could be described by a linear vector addition of luminance and L-/M-cone opponent reflecting ERGs. The contribution of the luminance mechanism increased with increasing temporal frequency and with increasing stimulus size, whereas the gain of the L-/M-cone opponent mechanism was independent of stimulus size and was larger at lower temporal frequencies. Thus, the luminance mechanism dominated at lower temporal frequencies with large stimuli. At 12 Hz, the second harmonic component reflected the luminance mechanism.

CONCLUSIONS

The ERGs to heterochromatic stimuli can be fully described in terms of linear combinations of responses in the (magnocellular) luminance and the (parvocellular) L-/M-opponent retino-geniculate pathways. The non-invasive study of these pathways in human subjects may have implications for basic research and for clinical research.

Developments in non-invasive visual electrophysiology

To study the physiology of the primate visual system, non-invasive electrophysiological techniques are of major importance. Two main techniques are available: the electroretinogram (ERG), a mass potential originating in the retina, and the visual evoked potential (VEP), which reflects activity in the primary visual cortex. In this overview, the history and the state of the art of these techniques are briefly presented as an introduction to the special issue "New Developments in non-invasive visual electrophysiology". The overview and the special issue can be used as the starting point for exciting new developments in the electrophysiology of primate and mammalian vision.

Melanopsin and Cone Photoreceptor Inputs to the Afferent Pupil Light Response

Background: Retinal photoreceptors provide the main stage in the mammalian eye for regulating the retinal illumination through changes in pupil diameter, with a small population of melanopsin-expressing intrinsically photosensitive retinal ganglion cells (ipRGCs) forming the primary afferent pathway for this response. The purpose of this study is to determine how melanopsin interacts with the three cone photoreceptor classes in the human eye to modulate the light-adapted pupil response.

Methods: We investigated the independent and combined contributions of the inner and outer retinal photoreceptor inputs to the afferent pupil pathway in participants with trichromatic color vision using a method to independently control the excitations of ipRGCs, cones and rods in the retina.

Results: We show that melanopsin-directed stimuli cause a transient pupil constriction generated by cones in the shadow of retinal blood vessels; desensitizing these penumbral cone signals uncovers a signature melanopsin pupil response that includes a longer latency (292 ms) and slower time (4.1x) and velocity (7.7x) to constriction than for cone-directed stimuli, and which remains sustained post-stimulus offset. Compared to melanopsin-mediated pupil responses, the cone photoreceptor-initiated pupil responses are more transient with faster constriction latencies, higher velocities and a secondary constriction at light offset. The combined pupil responses reveal that melanopsin signals are additive with the cone signals.

Conclusions: The visual system uses the L–, M–, and S–cone photoreceptor inputs to the afferent pupil pathway to accomplish the tonic modulations of pupil size to changes in image contrast. The inner retinal melanopsin-expressing ipRGCs mediate the longer-term, sustained pupil constriction to set the light-adapted pupil diameter during extended light exposures.

Correlated cone noise decreases rod signal contributions to the post-receptoral pathways

This study investigated how invisible extrinsic temporal white noise that correlates with the activity of one of the three [magnocellular (MC), parvocellular (PC), or koniocellular (KC)] post-receptoral pathways alters mesopic rod signaling. A four-primary photostimulator provided independent control of the rod and three cone photoreceptor excitations. The rod contributions to the three post-receptoral pathways were estimated by perceptually matching a 20% contrast rod pulse by independently varying the LMS (MC pathway), +L-M (PC pathway), and S-cone (KC pathway) excitations. We show that extrinsic cone noise caused a predominant decrease in the overall magnitude and ratio of the rod contributions to each pathway. Thus, the relative cone activity in the post-receptoral pathways determines the relative mesopic rod inputs to each pathway.

Melanopsin photoreception contributes to human visual detection, temporal and colour processing

The visual consequences of melanopsin photoreception in humans are not well understood. Here we studied melanopsin photoreception using a technique of photoreceptor silent substitution with five calibrated spectral lights after minimising the effects of individual differences in optical pre-receptoral filtering and desensitising penumbral cones in the shadow of retinal blood vessels. We demonstrate that putative melanopsin-mediated image-forming vision corresponds to an opponent S-OFF L + M-ON response property, with an average temporal resolution up to approximately 5 Hz, and >10x higher thresholds than red-green colour vision. With a capacity for signalling colour and integrating slowly changing lights, melanopsin-expressing intrinsically photosensitive retinal ganglion cells maybe the fifth photoreceptor type for peripheral vision.