Background: Epiretinal implants based on microelectro-mechanical system (MEMS) technology with a polyimide (PI) material are being proposed for application. Many kinds of non-photosensitive PIs have good biocompatibility and stability as typical MEMS materials for implantable electrodes. However, the effects of MEMS microfabrication, sterilization and implantation using a photosensitive polyimide (PSPI) microelectrode array for epiretinal electrical stimulation has not been extensively examined. Methods: A novel PSPI (Durimide 7510) microelectrode array for epiretinal electrical stimulation was designed, fabricated based on MEMS processing and microfabrication techniques. The biocompatibility of our new microelectrode was tested in vitro using an MTT assay and direct contact tests between the microelectrode surface and cells. Electrochemical impedance characteristics were tested based on a three-electrode testing method. The reliability and stability was evaluated by a chronic implantation of a non-functional array within the rabbit eye. Histological examination and SEM were performed to monitor possible damage of the retina and microelectrodes. Electrically evoked potentials (EEPs) were recorded during the acute stimulation of the retina. Results: The substrate was made of PSPI and the electrode material was platinum (Pt). The PSPI microelectrode array showed good biocompatibility and appropriate impedance characteristics for epiretinal stimulation. After a 6-month epiretinal implantation in the eyes of rabbits, we found no local retinal toxicity and no mechanical compression caused by the array. The Pt electrodes adhesion to the PSPI remained stable. A response to electrical stimuli was with recording electrodes lying on the visual cortex. Conclusion: We provide a relevant design and fundamental characteristics of a PSPI microelectrode array. Strong evidences on testing indicate that implantation is safe in terms of mechanical pressure and biocompatibility of PSPI microelectrode arrays on the retina. The dual-layer process we used proffers considerable advantages over the more traditional single-layer approach and can accommodate much many electrode sites. This lays the groundwork for a future, high-resolution retinal prosthesis with many more electrode sites based on the flexible PSPI thin film substrate.

Retinal electrical stimulation is a widely utilized method to restore visual function for patients with retinal degenerative diseases. Transcorneal electrical stimulation (TES) represents an effective way to improve the visual function due to its potential neuroprotective effect. However, TES with single electrode fails to spatially and selectively stimulate retinal neurons. Herein, a computational modeling method was proposed to explore the feasibility of spatially selective retinal stimulation via temporally interfering electric fields. An eyeball model with multiple electrodes was constructed to simulate the interferential electric fields with various electrode montages and current ratios. The results demonstrated that the temporal interference (TI) stimulation would gradually generate an increasingly localized high-intensity region on retina as the return electrodes moved towards the posterior of the eyeball and got closer. Additionally, the position of the convergent region could be modulated by regulating the current ratio of different electrode channels. The TI strategy with multisite and steerable stimulation can stimulate local retinal region with certain convergence and a relatively large stimulation range, which would be a feasible approach for the spatially selective retinal neuromodulation.

Epiretinal prostheses have been greatly successful in helping restore the vision of patients blinded by retinal degenerative diseases. The design of stimulating electrodes plays a crucial role in the performance of epiretinal prostheses. The objective of this study was to investigate, through computational modeling analysis, the effects on the excitation of retinal ganglion cells (RGCs) when different three-dimensional (3-D) electrodes were placed in the epiretinal space. 3-D finite element models of retinal electrical stimulation were created in COMSOL using a platinum microelectrode, a vitreous body, multi-layered retinal tissue, and retinal pigment epithelium (RPE). Disk and non-planar electrodes with different 3-D structures were used in the epiretinal electrical stimulation. In addition, a multi-RGC model including ionic mechanisms was constructed in NEURON to study the excitability of RGCs in response to epiretinal electrical stimulation by different types of electrodes. Threshold current, threshold charge density, and the activated RGC area were the three key factors used to evaluate the stimulating electrode’s performance. As the electrode-retina distance increased, both threshold current and threshold charge density showed an approximately linear relationship. Increasing the disk electrode’s diameter resulted in an increase in threshold current and a decrease in threshold charge density. Non-planar electrodes evoked different activation responses in RGCs than the disk electrode. Concave electrodes produced superior stimulation localization and electrode safety while convex electrodes performed relatively poorly. Investigation of epiretinal electrical stimulation using different 3-D electrodes would further the optimization of electrode design and help improve the performance of epiretinal prostheses. The combination of finite element analysis in COMSOL and NEURON software provides an efficient way to evaluate the influences of various 3-D electrodes on epiretinal electrical stimulation. Non-planar electrodes had larger threshold currents than disk electrodes. Of the five types of electrodes, concave hemispherical electrodes may be the ideal option, considering their superior stimulation localization and electrode safety.