Abstract
Purpose
To establish a clinically applicable dark-adapted 15 Hz flicker electroretinogram (ERG) for sensitive detection of functional changes in the fast and the slow retinal rod pathways.
Methods
The study consisted of two parts. In the first part, the paradigm of the previously demonstrated “null-effect” of stimulus luminance on ERG-amplitudes for 15 Hz flicker stimuli with duration of 2.5 s was reexamined. In the second part an optimized faster protocol designed for routine clinical use was tested and compared with the results from the first part and data published in the past.
The 15-Hz flicker ERG protocol included 12 steps with different intensities ranging from 0.00019 scot cd.s/m2 to 0.05 scot cd.s/m2 and was obtained from 15 normally sighted subjects. Stimulus flashes were presented in blue (448 nm) and white, alternating in each luminance step after 20 min of dark adaptation.
A shorter version containing only 7 steps was selected with intensities ranging from 0.00031 scot cd.s/m2 to 0.005 scot cd.s/m2. Additionally, the stimulus paradigm included noise measurement to properly estimate the signal-to-noise ratio.
Results
The results from the first protocol showed a U-shaped amplitude vs. luminance response curve at 15 Hz, consistent with previously published results. Using the new, shortened protocol, we also obtained similar results, but with a much shorter recording time. Based on the signal-to-noise ratio from these measurements a reliable measurement can be obtained in only 2.5 min which presents a considerable time reduction necessary to record a 15 Hz response curve in a clinical setup.
Conclusion
The new protocol is suitable for integration into a daily clinical environment, offering short and well-balanced protocols to address complex retinal network interactions.
Keywords: 15 Hz flicker, Rod pathways, Null effect, Scotopic ERG
Introduction
The human full-field electroretinogram (ERG) represents a global electrical response from the retina to a diffuse flash of light, reflecting the neural activity of various retinal cell populations [1].
The International Society for Clinical Electrophysiology of Vision (ISCEV) recommends recording full-field ERGs for different adaptation conditions (light- and dark-adapted) and different stimulus parameters to target the rod and cone pathways independently [2].
One of the more elusive phenomena of the ERG can be seen in recordings of the retinal response (in dark-adapted state) to a series of 15 Hz flicker stimuli with progressively increasing luminance levels. The relationship between the measured response (amplitude and phase) and stimulus luminance can be described as a U-shaped curve. Various explanations have been proposed for this phenomenon over the years. Early literature attributed the trough in the function to interactions between two distinct rod pathways at scotopic luminance levels, a “slow” and a “fast” pathway [3].
Several studies have provided further evidence supporting the existence of at least two rod pathways in the mammalian retina [4–6]. According to the theory proposed by Sharpe & Stockman 6, the “slow” pathway serves as the primary channel for rod outputs, transmitting signals from rod photoreceptors to rod ON bipolar cells, then to AII-amacrine cells, followed by cone ON and OFF bipolar cells and finally to ganglion cells. Conversely, the “fast” secondary pathway facilitates direct signal transmission from rods to cone photoreceptors via gap junctions, then to cone ON and OFF bipolar cells and their ganglion cells. As the stimulus luminance increases, the phase lag between these two rod-driven signals becomes larger, reaching a maximum of 180° at 15 Hz. According to this theory, this phase opposition leads to destructive interference, resulting in the attenuation of ERG and the perceptual disappearance of flicker in psychophysical tests at a specific flash luninance where both signals are of equal amplitude but opposite phase [7, 8].
Furthermore, as the b-wave is the predominant component of the flicker response, it is believed to be at least partially modulated by Müller cells [9]. However, the role of the temporal properties of the Müller cell response in the context of the 15 Hz cancellation effect has not been addressed yet to our knowledge.
Examples of the use of 15 Hz flicker in various rare inherited retinal diseases are also found in the literature. Independent research groups, have investigated and compared the responses of control participants with responses with those of patients with Stargardt disease [10] congenital stationary night blindness [11–13], and achromatopsia [14]. In a study comparing twenty-seven patients with STGD1, carrying mutations in both alleles of the ABCA4 gene, with normal controls, scotopic ERG response amplitudes and phases to flicker intensities ranging from 0.00031 scot cd.s/m2 to 0.005 scot cd.s/m2 were applied at a flicker frequency of 15 Hz. The amplitudes of both, the slow and fast rod ERG signals, were significantly reduced in the STGD1 group. Additionally, the phases of the slow rod responses lagged significantly, whereas the fast rod response phase remained unaffected. Notably, no significant alterations were revealed by the standard scotopic ERG. This study showed that dysfunction of the ABCA4 transporter can functionally affect both the rod bipolar–AII cell pathway and the rod–cone coupling pathway [10].
Bijveld and collaborators also applied the extended 15 Hz protocol in one patient with achromatopsia, four with CSNB1 (congenital stationary night blindness type 1) patients, and three with CSNB2, comparing the results with those of control subjects. The 15 Hz responses from the patient with achromatopsia showed a response pattern comparable to that of a normal subject, including the characteristic minimum in the U-shaped curve [11].
These studies highlight the potential of 15 Hz ERG recordings in describing retinal signal transmission deficits in inherited retinal diseases. Nonetheless, existing flicker protocols present multiple limitations and pose challenges for implementation in a clinical setup. To address these issues, the present study aimed to refine the paradigm based on previous 15 Hz publications [3] and optimize it for the Espion E3 system. The primary goal was to replicate the U-shaped curves using the 15 Hz flicker paradigm. Subsequently, our aim was also to improve the protocol to facilitate integration into standard clinical scotopic ERG examinations.
Key improvements included preserving dark adaptation by focusing only on the central part of the U-shaped curve, where stimulus intensities would not interfere with dark adaptation. Additionally, recording duration was reduced to better align with standard clinical routines. Furthermore, we conducted a more precise assessment of the U-shaped curve by including recordings of intrinsic noise and subsequently normalizing the results. Consequently, the protocol has been adapted to include systematic noise calculation and normalization, improving the reliability of the measured responses.
Methods
In the initial phase of the study, 15 healthy individuals were recruited, comprising 7 females and 8 males, with ages ranging from 28 to 65 years. All participants had full visual acuity, were emmetropic or myopic with a maximum refractive error of 5 diopters and had no prior history of eye diseases or other relevant disorders.
The ERG recordings for the first experiment were conducted in a single session. For each luminance step, two results were obtained, each consisting of three sweeps (continuous stimulation intervals) lasting 5 s. An interval of 7 s was set between each result. Sweeps with low recording quality (e.g., due to eye blinks or eye movement artifacts) were automatically rejected by the software.
Stimuli were produced by Diagnosys Espion E3 system and presented in a ColorDome Ganzfeld (Diagnosys LLC, Lowell, MA) and consisted of 15 Hz flicker with time integrated (flash) luminance ranging from −3.7 Log scot cd.s/m2 (0.00019 scot) cd.s/m2 to −1.3 scot cd.s/m2 (0.05 scot cd.s/m2) in steps divided by 0.2 Log scot cd.s/m2. The stimulation was presented in two colors, blue (448 nm) and white, following 20 min of dark adaptation. The whole examination, consisting of 24 steps and including the dark adaptation period, was completed in approximately 1 h. To extract responses at 15 Hz fast Fourier transform was calculated and the absolute value at 15 Hz was extracted for future analyses.
In a second phase, a revised protocol was conducted in a separate session. For that purpose, 9 individuals, including 4 men and 5 women with ages ranging from 21 to 47 years, have been recruited.
One subject was excluded from the following analysis due to strong artifacts from the eye movements. The second abbreviated protocol comprised only 7 steps with flash luminance ranging from 3.5 Log scot cd.s/m2 (0.00031 scot cd.s/m2) to −2.3 scot cd.s/m2 (0.005 scot cd.s/m2) in steps separated by 0.2 Log scot cd.s/m2. The stimulation was performed with only the narrow band blue stimuli. For each luminance, 5 results were obtained comprising of 10 sweeps (50 repetitions altogether). Each sweep had a duration of 2.5 s, with the stimulus presented for the initial 2 s followed by 500 ms with no stimulation.
To calculate the response amplitude, the continuous wavelet transforms (CWT) (Matlab 2020b, The MathWorks, Natick, MA, USA). The analytic wavelet used to compute the CWT was the classical Morlet (Gabor), wavelet. The magnitude of responses was calculated separately for 15 Hz responses with L1 normalization. The L1 normalization is an automatic normalization in the MATLAB implementation of CWT to provide a more accurate representation of the signal. for the whole 2.5 s stimuli interval. According to the stimulation paradigm and amplitude response curve the first transient period of 500 ms with a strong change in power responses and last period of 500 ms were not considered for the analysis. Therefore, the amplitude of responses is calculated as the mean value of CWT transformations between 500 and 2000 ms. The intrinsic noise levels, calculated from the period 2–2.5 s without stimuli, were used for noise estimate and eventual normalization. The process of normalization to noise was performed to allow a better estimate of the signal amplitude at the turning point of the u-curve, but additionally, such type of normalization allows a better group comparison.
Signal-to-noise ratio (SNR) of the second revised protocol was unknown. Therefore, an additional analytical test was performed to determine the required number of sweeps for a stable SNR profile. From 50 sweeps collected for each subject, we averaged a specific number of randomly selected sweeps to calculate the dependency between the number of sweeps and the SNR. The results from this analysis established a minimum recording duration necessary to achieve stable frequency responses and /or signal-to-noise.
In both measurements, the pupils of the tested eyes were dilated with 1% tropicamide. The ERG traces were recorded with DTL electrodes placed along the lower lid margin and made contact with the cornea. Gold cup electrodes positioned on the temple region and forehead served as reference and ground, respectively.
The experiments adhered to the tenets of the Declaration of Helsinki and were approved by the institutional ethics committee (University of Tübingen). Informed consent was obtained from each subject before the experiments.
Results
The standard 15 Hz protocol resulted in the U-shaped curve responses for both stimuli colors. The lowest amplitude value was achieved at the step number 4 with 0.00125 scot. cd.s/m2 for both stimulus colors (Table 1; Fig. 1). The responses following the blue stimuli were slightly larger than those with the white stimuli in all steps.
Table 1.
Stimulus charachteristics and response amplitudes for the standard long protocol. The table summarizes the wavelength and luminous intensity of each stimulus used. Mean response amplitude and standard deviations across participants are also reported
| Step number | stimuli | scot cd.s/m2 | Amplitude (µV) mean ± std |
|---|---|---|---|
| 1 | BLUE | 0.00019 | 2.4 ± 1.4 |
| 1 | WHITE | 0.00019 | 2.5 ± 1.1 |
| 2 | BLUE | 0.00031 | 2.2 ± 1.0 |
| 2 | WHITE | 0.00031 | 2.3 ± 1.2 |
| 3 | BLUE | 0.0005 | 1.7 ± 0.9 |
| 3 | WHITE | 0.0005 | 1.7 ± 0.9 |
| 4 | BLUE | 0.00125 | 1.2 ± 0.7 |
| 4 | WHITE | 0.00125 | 1.2 ± 0.7 |
| 5 | BLUE | 0.002 | 1.4 ± 0.7 |
| 5 | WHITE | 0.002 | 1.2 ± 0.7 |
| 6 | BLUE | 0.003 | 2.0 ± 0.9 |
| 6 | WHITE | 0.003 | 1.8 ± 0.9 |
| 7 | BLUE | 0.005 | 3.1 ± 1.3 |
| 7 | WHITE | 0.005 | 2.6 ± 1.2 |
| 8 | BLUE | 0.008 | 4.6 ± 2.1 |
| 8 | WHITE | 0.008 | 4.1 ± 1.8 |
| 9 | BLUE | 0.013 | 6.5 ± 3.0 |
| 9 | WHITE | 0.013 | 6.1 ± 2.8 |
| 10 | BLUE | 0.02 | 8.9 ± 3.7 |
| 10 | WHITE | 0.02 | 8.0 ± 3.6 |
| 11 | BLUE | 0.03 | 10.0 ± 4.3 |
| 11 | WHITE | 0.03 | 10.2 ± 4.1 |
| 12 | BLUE | 0.05 | 11.0 ± 3.8 |
| 12 | WHITE | 0.05 | 10.7 ± 4.3 |
Fig. 1.
The old protocol for all subjects. Bar plots of mean amplitude values of responses over all 15 subjects for each step. Blue bars represent the blue stimuli and red bars represent the white stimuli. Error bars are calculated as standard deviation over all subjects and particular steps. The standard 15 Hz protocol resulted in the U-shaped curve responses for both stimulus colors. The lowest amplitude value was achieved at the step number 4 with 0.00125 scot. cd.s/m2 flash luminance for both stimulus colors
In the second short protocol, only 7 steps were recorded per subject. The convolution obtained from the continuous wavelet transform in Fig. 2 shows transiently increased responses in the first 500 ms with stabilization of amplitude in the period between 500 and 2000 ms. The period of 2000–2500 ms is without stimuli and should correspond to intrinsic noise of the recording. The example of time frequency analysis of a subject is presented in Fig. 2.
Fig. 2.
Time frequency and response amplitudes analysis of a single subject. An example of response amplitudes for all steps in the short protocol calculated as continuous wavelet transform at 15 Hz for a single subject. In this measurement only 7 steps were recorded per subject. The convolution obtained from the continuous wavelet transform shows transiently increased responses in the first 500 ms with stabilization of amplitude in the period between 500 and 2000 ms. The period of 2000–2500 ms is without stimuli and should correspond to intrinsic noise of the recording
The responses over all the subjects can be described as U-shaped curve with responses at minimum in step number 4 with flash luminance of 0.00125 scot ocd.s/m2. (Table 2; Fig. 3).
Table 2.
New short protocol: time-integrated luminance and amplitude of responses
| Steps | Flash Luminance scot. cd.s/m2 | Amplitude (µV) mean ± std |
|---|---|---|
| 1 | 0.00031 | 2.9 ± 1.2 |
| 2 | 0.0005 | 2.6 ± 0.9 |
| 3 | 0.00079 | 2.2 ± 0.9 |
| 4 | 0.00125 | 2.2 ± 0.7 |
| 5 | 0.002 | 2.4 ± 0.7 |
| 6 | 0.003 | 2.8 ± 0.7 |
| 7 | 0.005 | 3.9 ± 1.3 |
Fig. 3.
The new short protocol for all subjects – not normalized for noise. Bar plots of the steps in the short protocol (mean value of the period 500–2000 ms) averaged over all the subjects, not normalized for noise. The responses over all the subjects can be described as U-shaped curve with a minimum response in step number 4 with stimuli of 0.00125 scot cd.s/m
The normalized signal of the short protocol responses exhibits the same shape as the not normalized ones. At step 4, the value approaches 1, suggesting an absence of measurable 15 Hz responses at that step (Fig. 4). Looking at the individual level of normalized responses, notably all subjects show a clear existence of the U-shaped response (Fig. 5.) and almost identical responses at the minimum. However, the amplitudes at steps 1,2 or 6,7 show a much bigger heterogeneity.
Fig. 4.
The new short protocol for all subjects – normalized for noise. Bar plots of the steps in the short protocol (mean value of the period 500–2000 ms) averaged over all subjects and normalized for noise calculated from the 2000–2500 ms period. The normalized signal of the short protocol responses exhibits the same shape as the not normalized one. With the value at the step 4 approximating the value 1, it indicates no measurable 15 Hz responses at that step
Fig. 5.
Bar plots of the steps in the short protocol (mean value of the period 500–2000 ms) individually for each subject and normalized for noise calculated from the 2000–2500 ms period
Additionally, we conducted an analysis to establish the number of sweeps required for signal stabilization, using step number 7 with a clear signal. The relationship between the number of randomly selected sweeps to the signal-to-noise profile, averaged over all subjects, is shown in the Fig. 6. In general, a single sweep has a good signal-to noise ratio profile but adding more sweeps up to approximately 10 sweeps can lead to further enhances the signal-to-noise ratio (Fig. 5).
Fig. 6.
Signal to noise ratio dependency on the number of trials randomly selected from 50 sweeps. The signal to noise ratio shown here is the mean value calculated from all participants
Therefore, for normal-sighted subject the new protocol consisted of 7 steps with 10 repetitions per step and a duration of 2.5 s, with the total recording time of approximately 195 s including the interstimulus intervals time.
Discussion
Based on earlier studies [3, 7], we recreated a shortened version of the 15 Hz flicker ERG measurement. Initially, we established new values for the stimulus intensities across 12 steps. The results aligned with previous findings reconstructed a U-shaped response curve when stimulated with both blue and white light [3]. The minimum amplitude value of 15 Hz flicker response was 1.2 µV at a luminance of 0.00125 cd.s/m2 scotopic for blue stimuli and 1.2 µV at 0.00125 cd.s/m2 scotopic with white stimuli. The shape and amplitude of the curves showed only minimal differences between blue and white stimuli indicating that the responses are exclusively driven by rods.
The main purpose of this study was to establish a protocol suitable for routine clinical application. To that end, we used only luminance levels in direct proximity to the minimum of the U-shaped curve. By avoiding high luminance, we were able to maintain the dark adaptation status during the entire recording process. Consequently, as there were no differences between the stimulus colors, we decided to use only the blue stimuli for the shortened protocol. These changes significantly reduced the recording duration. Furthermore, one intermediate stimulus (0.00079 cd·s/m2 scotopic) was newly introduced based on insights from the full dataset, as it proved important to preserving the curve’s shape and thus the ability to detect pathological deviations.
In the short 15 Hz protocol, we presented 50 sweeps with each trial consisting of 2000 ms of flicker 15 Hz stimuli followed by a 500 ms period of “noise recording”. The results of the short recordings show that the U-response curve was respected. Moreover, the inclusion of the noise recording allowed a proper evaluation of the signal at the minimum of the U-response and normalization with noise allows a better comparison between different subjects. Our results demonstrate that at luminance level corresponding to the minimum of the U-curve, the difference between the intrinsic power at 15 Hz and stimulus is only around 10%. This evidence suggests that the effect of 15 Hz is not just a small reduction, but rather almost a complete cancellation of the ERG signal at that specific luminance. These findings question to some extent the hypothesis of two pathways or the ON versus OFF mechanisms. Moreover, the power extracted with the time–frequency analysis in 15 Hz responses (Fig. 2) is of significant interest, as it reveals multiple stages in retinal processing for this type of stimulation. The initial transition and “accommodation” period lasted about 400 ms, thereafter the power of 15 Hz normalized depending on the strength of the stimulus. The end of the stimulation is depicted in the frequency analysis as a sharp drop of power. This observation may not fully align with the explanation concerning the involvement of two rod pathways, or the interference phenomena, which would suggest an additional peak activity due to the negative phase interferences from the following stimuli. Therefore, to some extent, the absence of a peak in power at the end of the stimulus presentation is more supportive of the ON–OFF hypothesis.
Although the origin of this effect is still not proven in human retina, we can infer that it emerges from a high order of interactions among various neuronal processing pathways within the retina. Considering that this interaction encompasses multiple retinal cell types, it is not surprising that the effect is also highly sensitive to specific retinal pathologies, as shown in a few clinical studies.
For instance, in two patients with Stargardt's macular dystrophy (carrying mutations in the ABCA4 gene), Scholl and colleagues [10] observed considerable amplitude reductions of both rod ERG pathways, despite both patients showing normal scotopic b-wave amplitudes in the standard ERG. However, the timing features showed good correspondence between the scotopic b-wave implicit time and the phase data of the slow rod ERG signals. The amplitude reductions were very similar for both rod ERG pathways, but more pronounced for the patient with advanced Stargardt disease. In this context, the scotopic 15-Hz flicker ERG can reveal rod dysfunctions not detectable by means of standard ERG techniques. However, due to the small number of subjects, further investigations are needed to determine whether the stimulus parameters used so far indeed reveal rod photoreceptor dysfunctions in the ERG and thus could serve as a functional correlation to the molecular genetic finding that the primary defect in Stargardt disease resides in the rods.
The choice of the number of sweeps was made arbitrarily, as we aimed to perform additional analyses to determine the required number of repetitions necessary to achieve a good signal-to-noise ratio. The results obtained from estimating the required number of trials for stabilizing the signal-to-noise ratio indicate that even with a small number of trials it is feasible to achieve a satisfactory signal-to-noise ratio. Thus, adding more than ten sweeps would not contribute significantly to increase the signal over noise. Our data indicate that even a single sweep shows a good signal-to-noise ratio, and the addition of more trials yields only a marginal improvement of 20% once the ceiling effect is reached. Therefore, with a duration of 2.5 s and 10 repetitions, the recording time per step will result in 25 s. Given that only 7 steps are required, the whole recording can be completed in around 3 min. Amplitude differences were observed between the full and abbreviated 15-Hz ERG protocols, ranging from about 20% (at 0.005 cd-s/m2) to 50% (at 0.002 cd-s/m2). These differences can be attributed to multiple factors: one major factor is the smaller sample size used in the abbreviated protocol, which likely increased interindividual variability and reduced the robustness of the mean response. In addition, the difference in the number of sweeps and total recording time between the two protocols may have affected the signal-to-noise ratio, particularly at luminance levels near the minimum of the curve, where the amplitude of the response is naturally reduced. Given the nonlinear behavior of the U-shaped curve, small variations in recording conditions or participant characteristics may have a disproportionate impact on the amplitudes at these critical points. Nevertheless, the shape and relative position of the minimum remained consistent across protocols, supporting the validity of the abbreviated version for clinical use.
Furthermore, since the luminance steps are also selected within the scotopic range, it is possible to implement this protocol in the standard clinical recording of ERG during a dark-adapted state. New studies addressing the test–retest variability of the 15 Hz protocols could be easily completed even in patient populations. The estimate of the recording time is calculated on the signal-to-noise information from healthy participants, therefore the adjustment in a patient population with small b wave response due to degeneration of retina should be taken into account and tested separately in that population.
The study has several important limitations. The small sample size and focus on healthy participants limit the predictability of how sensitive this protocol would be in a clinical population. Nevertheless, integrating this protocol to the standard recording during clinical visits would enable a comprehensive evaluation of the sensitivity of 15 Hz recording as a screening tool for early signs of retinal degeneration. Hence, readily applicable protocols such as this one could provide a possible novel biomarker for assessing early signs of a disease progression and some treatment effects. Currently, there is an unmet need for such a new type of objective and fast measurement to monitor and evaluate the progression of retinal degeneration with heightened sensitivity to any retinal damage. Additionally, we acknowledge that the results reported here reflect data from healthy individuals, and the protocol is in process to be tested in patients with retinal degeneration, particularly in those with reduced rod functions. Our primary goal was to integrate this measurement into routine clinical practice, allowing for the testing of different patient populations within the time and protocol constraints of a standard ISCEV ERG session and without disrupting dark adaptation. However, since the protocol has not yet been applied to patient populations, we recognize the need for future studies to evaluate the performance of this abridged protocol in clinical populations.
Funding
The research included in this publication was carried out during a foreign internship as part of the Preludium Bis2 project (led by Prof. Katarzyna Nowomiejska and financed by the National Science Center; UMO-2020/39/0/NZ5/00570).
Declarations
Conflict of interest
All authors certify that they have no affiliations with or involvement in any organization or entity with any financial interest (such as honoraria; educational grants; participation in speakers’ bureaus; membership, employment, consultancies, stock ownership, or other equity interest; and expert testimony or patent-licensing arrangements), or non-financial interest (such as personal or professional relationships, affiliations, knowledge or beliefs) in the subject matter or materials discussed in this manuscript.
Ethical approval
All procedures performed in studies involving human participants were in accordance with the ethical standards of the institutional research committee (bioethics committee of University of Tübingen) and with the 1964 Helsinki declaration and its later amendments or comparable ethical standards.
Informed consent
Informed consent was obtained from all individual participants included in the study.
Statement of human rights
All procedures conducted in this study involving human participants adhered to the ethical standards set by the institutional and/or national research committee and were in compliance with the 1964 Helsinki Declaration and its subsequent amendaments or comparable ethical standards.
Statement of welfare of animals
No animals were used in this study.
Footnotes
Publisher's Note
Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.
Aleksandra Czarnek-Chudzik and Giulia Righetti have contributed equally to this work.
References
- 1.Maguire J, Parry NRA, Kremers J, Murray IJ, McKeefry D (2017) The morphology of human rod ERGs obtained by silent substitution stimulation. Doc Ophthalmol 134:11–24. 10.1007/s10633-017-9571-4 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 2.Robson AG et al (2022) ISCEV Standard for full-field clinical electroretinography (2022 update). Doc Ophthalmol 144:165–177. 10.1007/s10633-022-09872-0 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 3.Sharpe LT, Stockman A (1999) Rod pathways: the importance of seeing nothing. Trends Neurosci 22:497–504. 10.1016/s0166-2236(99)01458-7 [DOI] [PubMed] [Google Scholar]
- 4.Bloomfield SA, Völgyi B (2009) The diverse functional roles and regulation of neuronal gap junctions in the retina. Nat Rev Neurosci 10:495–506. 10.1038/nrn2636 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 5.Asteriti S, Gargini C, Cangiano L (2014) Mouse rods signal through gap junctions with cones. Elife 3:e01386. 10.7554/eLife.01386 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 6.Robson JG, Maeda H, Saszik SM, Frishman LJ (2004) In vivo studies of signaling in rod pathways of the mouse using the electroretinogram. Vision Res 44:3253–3268. 10.1016/j.visres.2004.09.002 [DOI] [PubMed] [Google Scholar]
- 7.Stockman A, Sharpe LT, Zrenner E, Nordby K (1991) Slow and fast pathways in the human rod visual system: electrophysiology and psychophysics. J Opt Soc Am A 8:1657–1665. 10.1364/JOSAA.8.001657 [DOI] [PubMed] [Google Scholar]
- 8.Scholl H, Langrová H, Weber B, Zrenner E, Apfelstedt-Sylla E (2001) Clinical electrophysiology of two rod pathways: normative values and clinical application. Graefes Arch Clin Exp Ophthalmol 239:71–80. 10.1007/s004170000232 [DOI] [PubMed] [Google Scholar]
- 9.Bhatt Y, Hunt DM, Carvalho LS (2023) The origins of the full-field flash electroretinogram b-wave. Front Mol Neurosci. 10.3389/fnmol.2023.1153934 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 10.Scholl HPN, Besch D, Vonthein R, Weber BHF, Apfelstedt-Sylla E (2002) Alterations of slow and fast rod ERG signals in patients with molecularly confirmed Stargardt disease type 1. Invest Ophthalmol Vis Sci 43:1248–1256 [PubMed] [Google Scholar]
- 11.Bijveld MMC, Riemslag FCC, Kappers AML, Hoeben FP, Van Genderen MM (2011) An extended 15 Hz ERG protocol (2): data of normal subjects and patients with achromatopsia, CSNB1, and CSNB2. Doc Ophthalmol 123:161–172. 10.1007/s10633-011-9293-y [DOI] [PubMed] [Google Scholar]
- 12.Scalabrino ML et al (2015) Intravitreal delivery of a novel AAV vector targets ON bipolar cells and restores visual function in a mouse model of complete congenital stationary night blindness. Hum Mol Genet 24:6229–6239. 10.1093/hmg/ddv341 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 13.Van Genderen MM et al (2009) Mutations in TRPM1 are a common cause of complete congenital stationary night blindness. Am J Human Genet 85:730–736. 10.1016/j.ajhg.2009.10.012 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 14.Rüther K, Sharpe L, Zrenner E (1994) Dual rod pathways in complete achromatopsia. Ger J Ophthalmol 3:433–439 [PubMed] [Google Scholar]