A Computational Model of Phosphene Appearance for Epiretinal Prostheses

Area of Science:

• Biomedical Engineering
• Computational Neuroscience
• Ophthalmology

Background:

• Retinal neuroprostheses are FDA-approved for degenerative blindness.
• Predicting visual percepts (phosphenes) from electrical stimuli remains a challenge.

Purpose of the Study:

• Develop a computational model for predicting phosphene appearance.
• Account for stimulus parameters like amplitude, frequency, and pulse duration.

Main Methods:

• Phenomenological modeling of phosphene appearance.
• Utilizing a simulated retinal nerve fiber bundle map.
• Validation against psychophysical data from two independent studies.

Main Results:

• The model accurately predicts phosphene brightness, size, orientation, and elongation.
• Demonstrated generalization across different stimuli and electrodes.
• A comprehensive approach addressing both spatial and temporal phosphene aspects.

Conclusions:

• The developed model offers a significant advancement in predicting visual outcomes for retinal prosthesis users.
• The model is flexible, extensible, and adaptable to individual users.
• This work is a crucial step towards optimizing visual restoration in degenerative retinal diseases.