Inferring light responses of primate retinal ganglion cells using intrinsic electrical signatures.

Objective. Retinal implants are designed to stimulate retinal ganglion cells (RGCs) in a way that restores sight to individuals blinded by photoreceptor degeneration. Reproducing high-acuity vision with these devices will likely require inferring the natural light responses of diverse RGC types in the implanted retina, without being able to measure them directly. Here we demonstrate an inference approach that exploits intrinsic electrophysiological features of primate RGCs.Approach.First, ON-parasol and OFF-parasol RGC types were identified using their intrinsic electrical features in large-scale multi-electrode recordings from macaque retina. Then, the electrically inferred somatic location, inferred cell type, and average linear-nonlinear-Poisson model parameters of each cell type were used to infer a light response model for each cell. The accuracy of the cell type classification and of reproducing measured light responses with the model were evaluated.Main results.A cell-type classifier trained on 246 large-scale multi-electrode recordings from 148 retinas achieved 95% mean accuracy on 29 test retinas. In five retinas tested, the inferred models achieved an average correlation with measured firing rates of 0.49 for white noise visual stimuli and 0.50 for natural scenes stimuli, compared to 0.65 and 0.58 respectively for models fitted to recorded light responses (an upper bound). Linear decoding of natural images from predicted RGC activity in one retina showed a mean correlation of 0.55 between decoded and true images, compared to an upper bound of 0.81 using models fitted to light response data.Significance.These results suggest that inference of RGC light response properties from intrinsic features of their electrical activity may be a useful approach for high-fidelity sight restoration. The overall strategy of first inferring cell type from electrical features and then exploiting cell type to help infer natural cell function may also prove broadly useful to neural interfaces.

Inference of nonlinear receptive field subunits with spike-triggered clustering.

Responses of sensory neurons are often modeled using a weighted combination of rectified linear subunits. Since these subunits often cannot be measured directly, a flexible method is needed to infer their properties from the responses of downstream neurons. We present a method for maximum likelihood estimation of subunits by soft-clustering spike-triggered stimuli, and demonstrate its effectiveness in visual neurons. For parasol retinal ganglion cells in macaque retina, estimated subunits partitioned the receptive field into compact regions, likely representing aggregated bipolar cell inputs. Joint clustering revealed shared subunits between neighboring cells, producing a parsimonious population model. Closed-loop validation, using stimuli lying in the null space of the linear receptive field, revealed stronger nonlinearities in OFF cells than ON cells. Responses to natural images, jittered to emulate fixational eye movements, were accurately predicted by the subunit model. Finally, the generality of the approach was demonstrated in macaque V1 neurons.

An update on retinal prostheses.

Retinal prostheses are designed to restore a basic sense of sight to people with profound vision loss. They require a relatively intact posterior visual pathway (optic nerve, lateral geniculate nucleus and visual cortex). Retinal implants are options for people with severe stages of retinal degenerative disease such as retinitis pigmentosa and age-related macular degeneration. There have now been three regulatory-approved retinal prostheses. Over five hundred patients have been implanted globally over the past 15 years. Devices generally provide an improved ability to localize high-contrast objects, navigate, and perform basic orientation tasks. Adverse events have included conjunctival erosion, retinal detachment, loss of light perception, and the need for revision surgery, but are rare. There are also specific device risks, including overstimulation (which could cause damage to the retina) or delamination of implanted components, but these are very unlikely. Current challenges include how to improve visual acuity, enlarge the field-of-view, and reduce a complex visual scene to its most salient components through image processing. This review encompasses the work of over 40 individual research groups who have built devices, developed stimulation strategies, or investigated the basic physiology underpinning retinal prostheses. Current technologies are summarized, along with future challenges that face the field.

Unusual Physiological Properties of Smooth Monostratified Ganglion Cell Types in Primate Retina.

The functions of the diverse retinal ganglion cell types in primates and the parallel visual pathways they initiate remain poorly understood. Here, unusual physiological and computational properties of the ON and OFF smooth monostratified ganglion cells are explored. Large-scale multi-electrode recordings from 48 macaque retinas revealed that these cells exhibit irregular receptive field structure composed of spatially segregated hotspots, quite different from the classic center-surround model of retinal receptive fields. Surprisingly, visual stimulation of different hotspots in the same cell produced spikes with subtly different spatiotemporal voltage signatures, consistent with a dendritic contribution to hotspot structure. Targeted visual stimulation and computational inference demonstrated strong nonlinear subunit properties associated with each hotspot, supporting a model in which the hotspots apply nonlinearities at a larger spatial scale than bipolar cells. These findings reveal a previously unreported nonlinear mechanism in the output of the primate retina that contributes to signaling spatial information.

Full-field interferometric imaging of propagating action potentials.

Currently, cellular action potentials are detected using either electrical recordings or exogenous fluorescent probes that sense the calcium concentration or transmembrane voltage. Ca imaging has a low temporal resolution, while voltage indicators are vulnerable to phototoxicity, photobleaching, and heating. Here, we report full-field interferometric imaging of individual action potentials by detecting movement across the entire cell membrane. Using spike-triggered averaging of movies synchronized with electrical recordings, we demonstrate deformations up to 3 nm (0.9 mrad) during the action potential in spiking HEK-293 cells, with a rise time of 4 ms. The time course of the optically recorded spikes matches the electrical waveforms. Since the shot noise limit of the camera (~2 mrad/pix) precludes detection of the action potential in a single frame, for all-optical spike detection, images are acquired at 50 kHz, and 50 frames are binned into 1 ms steps to achieve a sensitivity of 0.3 mrad in a single pixel. Using a self-reinforcing sensitivity enhancement algorithm based on iteratively expanding the region of interest for spatial averaging, individual spikes can be detected by matching the previously extracted template of the action potential with the optical recording. This allows all-optical full-field imaging of the propagating action potentials without exogeneous labels or electrodes.

Temporal structure in spiking patterns of ganglion cells defines perceptual thresholds in rodents with subretinal prosthesis.

Subretinal prostheses are designed to restore sight in patients blinded by retinal degeneration using electrical stimulation of the inner retinal neurons. To relate retinal output to perception, we studied behavioral thresholds in blind rats with photovoltaic subretinal prostheses stimulated by full-field pulsed illumination at 20 Hz, and measured retinal ganglion cell (RGC) responses to similar stimuli ex-vivo. Behaviorally, rats exhibited startling response to changes in brightness, with an average contrast threshold of 12%, which could not be explained by changes in the average RGC spiking rate. However, RGCs exhibited millisecond-scale variations in spike timing, even when the average rate did not change significantly. At 12% temporal contrast, changes in firing patterns of prosthetic response were as significant as with 2.3% contrast steps in visible light stimulation of healthy retinas. This suggests that millisecond-scale changes in spiking patterns define perceptual thresholds of prosthetic vision. Response to the last pulse in the stimulation burst lasted longer than the steady-state response during the burst. This may be interpreted as an excitatory OFF response to prosthetic stimulation, and can explain behavioral response to decrease in illumination. Contrast enhancement of images prior to delivery to subretinal prosthesis can partially compensate for reduced contrast sensitivity of prosthetic vision.

Interferometric mapping of material properties using thermal perturbation.

Optical phase changes induced by transient perturbations provide a sensitive measure of material properties. We demonstrate the high sensitivity and speed of such methods, using two interferometric techniques: quantitative phase imaging (QPI) in transmission and phase-resolved optical coherence tomography (OCT) in reflection. Shot-noise-limited QPI can resolve energy deposition of about 3.4 mJ/cm2 in a single pulse, which corresponds to 0.8 °C temperature rise in a single cell. OCT can detect deposition of 24 mJ/cm2 energy between two scattering interfaces producing signals with about 30-dB signal-to-noise ratio (SNR), and 4.7 mJ/cm2 when SNR is 45 dB. Both techniques can image thermal changes within the thermal confinement time, which enables accurate single-shot mapping of absorption coefficients even in highly scattering samples, as well as electrical conductivity and many other material properties in biological samples at cellular scale. Integration of the phase changes along the beam path helps increase sensitivity, and the signal relaxation time reveals the size of hidden objects. These methods may enable multiple applications, ranging from temperature-controlled retinal laser therapy or gene expression to mapping electric current density and characterization of semiconductor devices with rapid pump-probe measurements.

Restoring Sight with Retinal Prostheses.

Vision begins when the eye's optical system-the cornea, iris, and crystalline lens-projects an image onto the retina, the thin and nearly transparent sheet of neural tissue that lines the back of the eye (see figure 1). Photoreceptors located at the back of the retina transduce incident photons into neural signals that are relayed to the brain. Those signals form the basis for visual perception. In humans, cone photoreceptors, which number about 6 million, dominate the central region of the visual field and are responsible for color and high-resolution day vision. Rod photoreceptors, which number about 120 million, dominate the periphery and mediate night vision.

Spatiotemporal characteristics of retinal response to network-mediated photovoltaic stimulation.

Subretinal prostheses aim at restoring sight to patients blinded by photoreceptor degeneration using electrical activation of the surviving inner retinal neurons. Today, such implants deliver visual information with low-frequency stimulation, resulting in discontinuous visual percepts. We measured retinal responses to complex visual stimuli delivered at video rate via a photovoltaic subretinal implant and by visible light. Using a multielectrode array to record from retinal ganglion cells (RGCs) in the healthy and degenerated rat retina ex vivo, we estimated their spatiotemporal properties from the spike-triggered average responses to photovoltaic binary white noise stimulus with 70-µm pixel size at 20-Hz frame rate. The average photovoltaic receptive field size was 194 ± 3 µm (mean ± SE), similar to that of visual responses (221 ± 4 µm), but response latency was significantly shorter with photovoltaic stimulation. Both visual and photovoltaic receptive fields had an opposing center-surround structure. In the healthy retina, ON RGCs had photovoltaic OFF responses, and vice versa. This reversal is consistent with depolarization of photoreceptors by electrical pulses, as opposed to their hyperpolarization under increasing light, although alternative mechanisms cannot be excluded. In degenerate retina, both ON and OFF photovoltaic responses were observed, but in the absence of visual responses, it is not clear what functional RGC types they correspond to. Degenerate retina maintained the antagonistic center-surround organization of receptive fields. These fast and spatially localized network-mediated ON and OFF responses to subretinal stimulation via photovoltaic pixels with local return electrodes raise confidence in the possibility of providing more functional prosthetic vision. NEW & NOTEWORTHY Retinal prostheses currently in clinical use have struggled to deliver visual information at naturalistic frequencies, resulting in discontinuous percepts. We demonstrate modulation of the retinal ganglion cells (RGC) activity using complex spatiotemporal stimuli delivered via subretinal photovoltaic implant at 20 Hz in healthy and in degenerate retina. RGCs exhibit fast and localized ON and OFF network-mediated responses, with antagonistic center-surround organization of their receptive fields.

Optophysiology of cardiomyocytes: characterizing cellular motion with quantitative phase imaging.

Quantitative phase imaging enables precise characterization of cellular shape and motion. Variation of cell volume in populations of cardiomyocytes can help distinguish their types, while changes in optical thickness during beating cycle identify contraction and relaxation periods and elucidate cell dynamics. Parameters such as characteristic cycle shape, beating frequency, duration and regularity can be used to classify stem-cell derived cardiomyocytes according to their health and, potentially, cell type. Unlike classical patch-clamp based electrophysiological characterization of cardiomyocytes, this interferometric approach enables rapid and non-destructive analysis of large populations of cells, with longitudinal follow-up, and applications to tissue regeneration, personalized medicine, and drug testing.

Activation of ganglion cells and axon bundles using epiretinal electrical stimulation.

Epiretinal prostheses for treating blindness activate axon bundles, causing large, arc-shaped visual percepts that limit the quality of artificial vision. Improving the function of epiretinal prostheses therefore requires understanding and avoiding axon bundle activation. This study introduces a method to detect axon bundle activation on the basis of its electrical signature and uses the method to test whether epiretinal stimulation can directly elicit spikes in individual retinal ganglion cells without activating nearby axon bundles. Combined electrical stimulation and recording from isolated primate retina were performed using a custom multielectrode system (512 electrodes, 10-µm diameter, 60-µm pitch). Axon bundle signals were identified by their bidirectional propagation, speed, and increasing amplitude as a function of stimulation current. The threshold for bundle activation varied across electrodes and retinas, and was in the same range as the threshold for activating retinal ganglion cells near their somas. In the peripheral retina, 45% of electrodes that activated individual ganglion cells (17% of all electrodes) did so without activating bundles. This permitted selective activation of 21% of recorded ganglion cells (7% of expected ganglion cells) over the array. In one recording in the central retina, 75% of electrodes that activated individual ganglion cells (16% of all electrodes) did so without activating bundles. The ability to selectively activate a subset of retinal ganglion cells without axon bundles suggests a possible novel architecture for future epiretinal prostheses.NEW & NOTEWORTHY Large-scale multielectrode recording and stimulation were used to test how selectively retinal ganglion cells can be electrically activated without activating axon bundles. A novel method was developed to identify axon activation on the basis of its unique electrical signature and was used to find that a subset of ganglion cells can be activated at single-cell, single-spike resolution without producing bundle activity in peripheral and central retina. These findings have implications for the development of advanced retinal prostheses.

Optimization of return electrodes in neurostimulating arrays.

OBJECTIVE: High resolution visual prostheses require dense stimulating arrays with localized inputs of individual electrodes. We study the electric field produced by multielectrode arrays in electrolyte to determine an optimal configuration of return electrodes and activation sequence. APPROACH: To determine the boundary conditions for computation of the electric field in electrolyte, we assessed current dynamics using an equivalent circuit of a multielectrode array with interleaved return electrodes. The electric field modeled with two different boundary conditions derived from the equivalent circuit was then compared to measurements of electric potential in electrolyte. To assess the effect of return electrode configuration on retinal stimulation, we transformed the computed electric fields into retinal response using a model of neural network-mediated stimulation. MAIN RESULTS: Electric currents at the capacitive electrode-electrolyte interface redistribute over time, so that boundary conditions transition from equipotential surfaces at the beginning of the pulse to uniform current density in steady state. Experimental measurements confirmed that, in steady state, the boundary condition corresponds to a uniform current density on electrode surfaces. Arrays with local return electrodes exhibit improved field confinement and can elicit stronger network-mediated retinal response compared to those with a common remote return. Connecting local return electrodes enhances the field penetration depth and allows reducing the return electrode area. Sequential activation of the pixels in large monopolar arrays reduces electrical cross-talk and improves the contrast in pattern stimulation. SIGNIFICANCE: Accurate modeling of multielectrode arrays helps optimize the electrode configuration to maximize the spatial resolution, contrast and dynamic range of retinal prostheses.

Photovoltaic Pixels for Neural Stimulation: Circuit Models and Performance.

Photovoltaic conversion of pulsed light into pulsed electric current enables optically-activated neural stimulation with miniature wireless implants. In photovoltaic retinal prostheses, patterns of near-infrared light projected from video goggles onto subretinal arrays of photovoltaic pixels are converted into patterns of current to stimulate the inner retinal neurons. We describe a model of these devices and evaluate the performance of photovoltaic circuits, including the electrode-electrolyte interface. Characteristics of the electrodes measured in saline with various voltages, pulse durations, and polarities were modeled as voltage-dependent capacitances and Faradaic resistances. The resulting mathematical model of the circuit yielded dynamics of the electric current generated by the photovoltaic pixels illuminated by pulsed light. Voltages measured in saline with a pipette electrode above the pixel closely matched results of the model. Using the circuit model, our pixel design was optimized for maximum charge injection under various lighting conditions and for different stimulation thresholds. To speed discharge of the electrodes between the pulses of light, a shunt resistor was introduced and optimized for high frequency stimulation.

Contrast Sensitivity With a Subretinal Prosthesis and Implications for Efficient Delivery of Visual Information.

PURPOSE: To evaluate the contrast sensitivity of a degenerate retina stimulated by a photovoltaic subretinal prosthesis, and assess the impact of low contrast sensitivity on transmission of visual information. METHODS: We measure ex vivo the full-field contrast sensitivity of healthy rat retina stimulated with white light, and the contrast sensitivity of degenerate rat retina stimulated with a subretinal prosthesis at frequencies exceeding flicker fusion (>20 Hz). Effects of eye movements on retinal ganglion cell (RGC) activity are simulated using a linear-nonlinear model of the retina. RESULTS: Retinal ganglion cells adapt to high frequency stimulation of constant intensity, and respond transiently to changes in illumination of the implant, exhibiting responses to ON-sets, OFF-sets, and both ON- and OFF-sets of light. The percentage of cells with an OFF response decreases with progression of the degeneration, indicating that OFF responses are likely mediated by photoreceptors. Prosthetic vision exhibits reduced contrast sensitivity and dynamic range, with 65% contrast changes required to elicit responses, as compared to the 3% (OFF) to 7% (ON) changes with visible light. The maximum number of action potentials elicited with prosthetic stimulation is at most half of its natural counterpart for the ON pathway. Our model predicts that for most visual scenes, contrast sensitivity of prosthetic vision is insufficient for triggering RGC activity by fixational eye movements. CONCLUSIONS: Contrast sensitivity of prosthetic vision is 10 times lower than normal, and dynamic range is two times below natural. Low contrast sensitivity and lack of OFF responses hamper delivery of visual information via a subretinal prosthesis.

Photovoltaic restoration of sight with high visual acuity.

Patients with retinal degeneration lose sight due to the gradual demise of photoreceptors. Electrical stimulation of surviving retinal neurons provides an alternative route for the delivery of visual information. We demonstrate that subretinal implants with 70-µm-wide photovoltaic pixels provide highly localized stimulation of retinal neurons in rats. The electrical receptive fields recorded in retinal ganglion cells were similar in size to the natural visual receptive fields. Similarly to normal vision, the retinal response to prosthetic stimulation exhibited flicker fusion at high frequencies, adaptation to static images and nonlinear spatial summation. In rats with retinal degeneration, these photovoltaic arrays elicited retinal responses with a spatial resolution of 64 ± 11 µm, corresponding to half of the normal visual acuity in healthy rats. The ease of implantation of these wireless and modular arrays, combined with their high resolution, opens the door to the functional restoration of sight in patients blinded by retinal degeneration.

Performance of photovoltaic arrays in-vivo and characteristics of prosthetic vision in animals with retinal degeneration.

Loss of photoreceptors during retinal degeneration leads to blindness, but information can be reintroduced into the visual system using electrical stimulation of the remaining retinal neurons. Subretinal photovoltaic arrays convert pulsed illumination into pulsed electric current to stimulate the inner retinal neurons. Since required irradiance exceeds the natural luminance levels, an invisible near-infrared (915 nm) light is used to avoid photophobic effects. We characterized the thresholds and dynamic range of cortical responses to prosthetic stimulation with arrays of various pixel sizes and with different number of photodiodes. Stimulation thresholds for devices with 140 µm pixels were approximately half those of 70 µm pixels, and with both pixel sizes, thresholds were lower with 2 diodes than with 3 diodes per pixel. In all cases these thresholds were more than two orders of magnitude below the ocular safety limit. At high stimulation frequencies (>20 Hz), the cortical response exhibited flicker fusion. Over one order of magnitude of dynamic range could be achieved by varying either pulse duration or irradiance. However, contrast sensitivity was very limited. Cortical responses could be detected even with only a few illuminated pixels. Finally, we demonstrate that recording of the corneal electric potential in response to patterned illumination of the subretinal arrays allows monitoring the current produced by each pixel, and thereby assessing the changes in the implant performance over time.

Selectivity of direct and network-mediated stimulation of the retinal ganglion cells with epi-, sub- and intraretinal electrodes.

OBJECTIVE: Intra-retinal placement of stimulating electrodes can provide close and stable proximity to target neurons. We assessed improvement in stimulation thresholds and selectivity of the direct and network-mediated retinal stimulation with intraretinal electrodes, compared to epiretinal and subretinal placements. APPROACH: Stimulation thresholds of the retinal ganglion cells (RGCs) in wild-type rat retina were measured using the patch-clamp technique. Direct and network-mediated responses were discriminated using various synaptic blockers. MAIN RESULTS: Three types of RGC responses were identified: short latency (SL, τ < 5 ms) originating in RGCs, medium latency (ML, 3 < τ < 70 ms) originating in the inner nuclear layer and long latency (LL, τ > 40 ms) originating in photoreceptors. Cathodic epiretinal stimulation exhibited the lowest threshold for direct RGC response and the highest direct selectivity (network/direct thresholds ratio), exceeding a factor of 3 with pulse durations below 0.5 ms. For network-mediated stimulation, the lowest threshold was obtained with anodic pulses in OPL position, and its network selectivity (direct/network thresholds ratio) increased with pulse duration, exceeding a factor of 4 at 10 ms. Latency of all three types of responses decreased with increasing strength of the stimulus. SIGNIFICANCE: These results define the optimal range of pulse durations, pulse polarities and electrode placement for the retinal prostheses aiming at direct or network-mediated stimulation of RGCs.

Cortical responses elicited by photovoltaic subretinal prostheses exhibit similarities to visually evoked potentials.

We have previously developed a wireless photovoltaic retinal prosthesis, in which camera-captured images are projected onto the retina using pulsed near-IR light. Each pixel in the subretinal implant directly converts pulsed light into local electric current to stimulate the nearby inner retinal neurons. Here we report that implants having pixel sizes of 280, 140 and 70 µm implanted in the subretinal space in rats with normal and degenerate retina elicit robust cortical responses upon stimulation with pulsed near-IR light. Implant-induced eVEP has shorter latency than visible light-induced VEP, its amplitude increases with peak irradiance and pulse duration, and decreases with frequency in the range of 2-20 Hz, similar to the visible light response. Modular design of the arrays allows scalability to a large number of pixels, and combined with the ease of implantation, offers a promising approach to restoration of sight in patients blinded by retinal degenerative diseases.

Photovoltaic Retinal Prosthesis with High Pixel Density.

Retinal degenerative diseases lead to blindness due to loss of the "image capturing" photoreceptors, while neurons in the "image processing" inner retinal layers are relatively well preserved. Electronic retinal prostheses seek to restore sight by electrically stimulating surviving neurons. Most implants are powered through inductive coils, requiring complex surgical methods to implant the coil-decoder-cable-array systems, which deliver energy to stimulating electrodes via intraocular cables. We present a photovoltaic subretinal prosthesis, in which silicon photodiodes in each pixel receive power and data directly through pulsed near-infrared illumination and electrically stimulate neurons. Stimulation was produced in normal and degenerate rat retinas, with pulse durations from 0.5 to 4 ms, and threshold peak irradiances from 0.2 to 10 mW/mm(2), two orders of magnitude below the ocular safety limit. Neural responses were elicited by illuminating a single 70 µm bipolar pixel, demonstrating the possibility of a fully-integrated photovoltaic retinal prosthesis with high pixel density.