Assessment of optogenetically-driven strategies for prosthetic restoration of cortical vision in large-scale neural simulation of V1

Ján Antolík, Quentin Sabatier, Charlie Galle, Yves Frégnac, Ryad BenosmanScientific 2021 · DOI

TL;DR

This paper presents a computational model of a virtual optogenetic vision prosthesis for the primary visual cortex (V1). They demonstrate a method to translate simple visual stimuli (gratings) into light patterns that can evoke neural activity in V1 similar to natural vision, even with relatively low spatial resolution and light intensity. This work provides a framework for designing future optogenetic vision restoration strategies.

What Problem Does This Paper Address?

Restoring vision for individuals with conditions like glaucoma or diabetic retinopathy, where the retina is damaged, requires stimulating downstream visual processing areas like the primary visual cortex (V1). Optogenetics, using light-sensitive proteins to control neuron activity, offers a promising approach for this. However, a major challenge is understanding how to design light stimulation patterns that effectively mimic natural visual input and lead to a naturalistic perception. Specifically, the functional impact of optogenetically evoked neural dynamics and the optimal protocols for inducing specific perceptual experiences are poorly understood. This paper tackles this by developing a detailed computational model to explore and design such optogenetic stimulation strategies for V1.

Methods

The core of this work is a sophisticated computational model that integrates several components:

  • Large-scale spiking neural network (SNN) of cat V1: This model, based on existing work, simulates the complex dynamics of approximately 30,000 neurons and 30 million synapses, capturing the anatomical and functional properties of V1.
  • Virtual prosthetic system: This system translates visual input into light stimulation patterns. It includes:
  • Model of light-emitting elements (MLEE): Simulates an array of light sources (like LEDs) that would be placed on the cortical surface.
  • Model of light propagation: Accounts for how light is absorbed and scattered within cortical tissue.
  • Model of channelrhodopsin (ChR) dynamics: Simulates how light activates ChR proteins expressed in neurons, leading to an inward current and neuronal firing.
  • Stimulation strategy: A novel protocol is developed to translate simple visual stimuli (sinusoidal gratings) into activation signals for the MLEE. This strategy considers the orientation preference of V1 neurons and the spatial distribution of light.

The researchers then compare neural activity evoked by natural visual input (simulated) with activity evoked by their optogenetic stimulation protocol under various conditions (e.g., stimulating only excitatory neurons vs. both excitatory and inhibitory neurons).

Datasets

Not discussed in the paper. The study relies on a simulated V1 model constrained by existing experimental data on cat V1.

Key Insights and Ideas

  • Optogenetics can mimic natural V1 activity: The study demonstrates that with a carefully designed protocol, optogenetic stimulation can evoke neural activity patterns in V1 that are remarkably similar to those generated by natural visual input, even in the absence of retinal input.
  • Spatial resolution and light intensity are surprisingly forgiving: Effective stimulation was achieved with a spatial resolution as low as 100 µm and light intensities as weak as 10¹⁶ photons/s/cm². This suggests that current or near-future technology might be sufficient for some aspects of prosthetic vision.
  • Intra-cortical dynamics are crucial: The model highlights how the inherent recurrent connectivity within V1 acts to "sharpen" the imposed light patterns, pushing the evoked activity towards patterns seen during natural vision. This suggests that rather than precisely dictating every neuron's output, mimicking the statistical properties of natural input might be more effective.
  • Targeting excitatory neurons is sufficient (and potentially better): Stimulating only excitatory neurons (OptoExc) yielded results closer to natural vision dynamics than stimulating both excitatory and inhibitory neurons (OptoExcInh), suggesting that this might be a more efficient strategy for prosthetic applications.
  • MLEE density is less critical than expected: The study found that even with a coarser MLEE pitch (up to 100 µm), orientation tuning could still be effectively induced, indicating flexibility in hardware design.

Key Results

  • Optogenetic stimulation with a spatial resolution of 100 µm and light intensity of 10¹⁶ photons/s/cm² was sufficient to evoke V1 activity patterns close to those evoked by normal vision.
  • The developed stimulation protocol successfully induced orientation tuning in V1 neurons, comparable to natural vision, across different conditions.
  • The model showed that intra-cortical processing significantly sharpens the induced activity patterns, making them more similar to natural responses.
  • Differences in membrane potential dynamics and synchronization were observed between natural and optogenetic stimulation, with the OptoExc condition showing dynamics closer to natural vision.

Limitations and Future Work

  • Model Simplifications: The V1 model is a down-sampled representation of cat cortex and uses a planar approximation, not accounting for the natural folding of the human cortex or the non-linear retino-cortical magnification factor.
  • Transient Dynamics: The proposed stimulation protocol does not perfectly reproduce the transient dynamics at the onset and offset of visual stimuli.
  • Human vs. Cat Cortex: While the cat model is a good starting point, differences in human cortex may require further adaptation.
  • Dendritic Arborization: The impact of transfected dendritic arbors on spatial light integration was not fully explored.

Future work suggested includes:

  • Developing more accurate models of human V1, incorporating cortical folding and magnification.
  • Improving stimulation protocols to better capture transient dynamics.
  • Investigating the role of dendritic arbors in spatial light integration.
  • Extending the framework to other visual stimuli and more complex perceptual tasks.
  • Using the computational framework to guide future experimental validation in animal models and eventually clinical trials.

Summary updated

Abstract

The neural encoding of visual features in primary visual cortex (V1) is well understood, with strong correlates to low-level perception, making V1 a strong candidate for vision restoration through neuroprosthetics. However, the functional relevance of neural dynamics evoked through external stimulation directly imposed at the cortical level is poorly understood. Furthermore, protocols for designing cortical stimulation patterns that would induce a naturalistic perception of the encoded stimuli have not yet been established. Here, we demonstrate a proof of concept by solving these issues through a computational model, combining (1) a large-scale spiking neural network model of cat V1 and (2) a virtual prosthetic system transcoding the visual input into tailored light-stimulation patterns which drive in situ the optogenetically modified cortical tissue. Using such virtual experiments, we design a protocol for translating simple Fourier contrasted stimuli (gratings) into activation patterns of the optogenetic matrix stimulator. We then quantify the relationship between spatial configuration of the imposed light pattern and the induced cortical activity. Our simulations in the absence of visual drive (simulated blindness) show that optogenetic stimulation with a spatial resolution as low as 100 [Formula: see text]m, and light intensity as weak as [Formula: see text] photons/s/cm[Formula: see text] is sufficient to evoke activity patterns in V1 close to those evoked by normal vision.
BibTeX
@inproceedings{antolik2021assessment,
  author = {Ján Antolík and Quentin Sabatier and Charlie Galle and Yves Frégnac and Ryad Benosman},
  title = {{Assessment of optogenetically-driven strategies for prosthetic restoration of cortical vision in large-scale neural simulation of V1}},
  booktitle = {Scientific},
  year = {2021},
  doi = {10.1038/s41598-021-88960-8},
}
Figure