Virtual Retinal Implant Simulator

Overview

This web application simulates the visual experience of a patient utilizing a retinal prosthesis (bionic eye). Retinal implants bypass damaged photoreceptors in the eye by directly stimulating the surviving retinal ganglion cells with a grid of microelectrodes. The patient does not see a normal image; instead, they perceive a grid of glowing electrical dots known as phosphenes.

How to Use

The simulation uses your device's webcam to process live visual data (if declined, it will gracefully fall back to synthetic geometric patterns). Interaction supports both Keyboard/Mouse and Touchscreens:

• Resolution: Change the electrode grid size. Lower resolutions (16x16) reflect early FDA-approved devices like the Argus II. Higher resolutions simulate theoretical future advancements.
• Disease Models: Simulates how pre-existing tissue damage degrades the artificial implant's effectiveness.
  • Macular Degeneration: Destroys central vision (scotoma).
  • Glaucoma: Damages peripheral vision, creating a "tunnel vision" effect.
  • Diabetic Retinopathy: Causes patchy, irregular loss of vision across the retina due to microvascular damage.
• Sonification: Toggles a sensory substitution audio scanner. It translates the image into sound by sweeping left-to-right. Pitch represents the vertical position of bright spots, and volume represents intensity—a common technique to aid the blind.
• Demo Mode: Automatically cycles through resolutions and disease states. Starts automatically after 10 seconds of inactivity.

Keyboard Shortcuts: Press 1-4 to change disease models, S to toggle sound, and D to toggle Demo mode.

Technical Details

The core rendering engine utilizes vanilla HTML5 <canvas> without external libraries. Visual processing extracts the luminance of the live video feed (or procedural fallback) mapped down to a quantized grid. Phosphenes are rendered dynamically using varying opacities to simulate the dynamic range of microelectrode electrical stimulation.

Disease simulation is achieved through synthetic spatial masks algorithmically generated at runtime: geometric radial masking for AMD and Glaucoma, and a deterministic hash-based noise function for Diabetic Retinopathy.

Future Directions

Future iterations of this simulator could include advanced color encoding algorithms, subretinal vs epiretinal stimulation mappings (which account for axonal streak distortion), and localized temporal decay (phosphene fading over time).