A novel simulation paradigm utilising MRI-derived phosphene maps for cortical prosthetic vision

Area of Science:

• Neuroscience
• Biomedical Engineering
• Ophthalmology

Background:

• Cortical visual prostheses aim to restore sight via visual cortex stimulation.
• Previous simulations used uniform phosphene maps, potentially misrepresenting real-world phosphene distributions.
• Realistic simulation is crucial for accurate assessment of prosthetic vision capabilities.

Purpose of the Study:

• To develop a realistic simulation paradigm for cortical prosthetic vision.
• To investigate the efficacy of a novel clustering algorithm in enhancing visual performance.
• To compare simulation performance against uniform phosphene maps.

Main Methods:

• Developed a simulation using Human Connectome Project retinotopy and MRI-based implant placement.
• Derived five unique retinotopic maps.
• Implemented head scanning and a density-based clustering algorithm to optimize stimulus presentation.
• Evaluated performance on visual acuity and object recognition tasks with ten subjects.

Main Results:

• Performance was significantly better with uniform control maps compared to retinotopic maps for both tasks.
• Head scanning and the clustering algorithm demonstrated potential for improving object recognition over multiple sessions.
• Retinotopic maps provide a more realistic, albeit challenging, simulation of prosthetic vision.

Conclusions:

• The developed simulation paradigm offers a more realistic representation of cortical prosthetic vision by incorporating individual brain anatomy and implant placement.
• Uniform phosphene maps may lead to overestimation of visual prosthetic performance.
• This simulation tool could aid in surgical planning for optimal implant placement in clinical settings.