In Vivo Electrical Stimulation of a Retinal Prosthesis Containing Conductive Diamond Electrodes

Abstract

Purpose: Nitrogen incorporated ultrananocrystalline diamond (N-UNCD) has recently been shown to be a promising material as a stimulating electrode. The aim of this study was to assess if N-UNCD electrodes are capable of stimulating retinal ganglion cells (RGCs) in vivo, at stimulus intensities considered safe for this material.

Methods: Hermetic arrays containing 120x120 µm N-UNCD electrodes were fabricated on a polycrystalline diamond substrate and embedded in a silicone carrier. A pars plana vitrectomy was performed in five normally-sighted anesthetized cats. Implants were inserted through a 5mm incision and fixed epiretinally using a titanium tack. Impedances were measured before implantation and in vivo. Multiunit responses to monopolar stimulation with biphasic current pulses (500 µs per phase), were recorded from the visual cortex using 60-electrode penetrating arrays (Blackrock Microsystems, UT). Eyes were enucleated, fixed and examined histologically for trauma to tissue surrounding the implant.

Results: Impedances in vivo (26.6±2 kΩ, Mean±SEM) did not differ (p>0.05) to those measured in vitro (26.4±2.3 kΩ). Epiretinal stimulation led to robust activation of the visual cortex with reliable thresholds measured from stimulating 30 of 42 electrodes. Best thresholds ranged between 29.5-442.6 µC/cm². Cathodic first pulses (54.9±2.8 nC) resulted in significantly lower (paired t-test, p<0.001) thresholds than anodic first pulses (73±5.2 nC). Histological analysis revealed focal damage to the retina surrounding the tacked edge of the implant. The retina beneath the diamond array remained attached. The extent of damage attributed to tacking alone versus the combined pressure from the tack and silicone implant body was unclear.

Conclusions: For all but three electrodes, charge densities required to evoke cortical responses were well within the previously established electrochemical safe limit for diamond (300 µC/cm²). N-UNCD electrodes were successfully used to acutely stimulate RGCs via an epiretinal approach with some electrodes requiring very low intensities to activate the visual cortex. Variability in impedances and thresholds, along with histological analyses, suggest that further optimization of the implant shape and tack insertion procedure is required to consistently obtain low thresholds and minimize damage. In addition, performance of these electrodes needs to be evaluated in chronic studies.

Keywords: 508 electrophysiology: non-clinical • 688 retina • 755 visual cortex