Optimal placement of high-channel visual prostheses in human retinotopic visual cortex

Area of Science:

• Neurotechnology
• Biomedical Engineering
• Computational Neuroscience

Background:

• Neuroprosthetics offer potential vision restoration for visual pathway damage.
• Manual electrode placement and evaluation are impractical for widespread use.
• A scalable, automated method for simulating and optimizing electrode implantation is needed.

Purpose of the Study:

• To develop a comprehensive method for automatically optimizing visual prosthesis electrode placement.
• To match predefined phosphene distributions using individual anatomical data and retinotopic predictions.
• To create a versatile simulation pipeline for evaluating electrode designs and surgical procedures.

Main Methods:

• Utilized retinotopic predictions and individual anatomy data for optimization.
• Developed a simulation pipeline adaptable to various electrode designs (e.g., 1000-channel 3D array in V1).
• Minimized discrepancies between simulated and target phosphene patterns.

Main Results:

• Individually optimized electrode placements significantly outperformed average solutions, highlighting anatomical specificity.
• Virtual implantation revealed challenges in achieving full visual field coverage with single arrays.
• Multiple electrode arrays and consideration of vasculature could improve outcomes.

Conclusions:

• The developed open-source software streamlines surgical procedure refinement for visual prostheses.
• Facilitates simulation studies for exploring electrode configuration possibilities.
• Advances the development of personalized vision restoration neurotechnologies.