Abstract
Purpose :
Design of retinal prosthesis relies on modeling of neural stimulation, for which electrical resistivity of the retina is one of the most important parameters. Published values for retinal resistivity vary by nearly two orders of magnitude (100 - 8,000 Ω*cm). Previous methods used penetrating probes, which alter retinal morphology by tissue compression and rupture. Moreover, since the retina is composed of many structurally different layers, representing it by one average resistivity may lead to significant distortions of the model. Here, we present a non-disruptive approach to characterize the layered resistivity profile in the retina.
Methods :
Electrochemical impedance spectroscopy (EIS) of rat retinas was performed with electrodes of 20, 40 and 80μm in diameter. A custom plug pressed the retinal explant onto the array, while EIS measurements were taken at each depth increment, monitored by optical coherence tomography (OCT). MATLAB and COMSOL were used to fit the data to a Randles circuit and determine the average resistivity of the retina as a function of its total thickness.
Impedance tomography was performed with retinal explants pressed onto an array of 512-electrodes of 10μm in diameter at 60μm spacing. Current, varied from 70 to 130nA, was injected from various electrodes and potential was recorded at neighboring electrodes. Retinal resistivity was mapped as a function of depth using Res2DInv inverse-problem software.
Results :
Average resistivity based on EIS measurements increased from 200 to 4000 Ω*cm as the retina was compressed. OCT showed irreversible damage to the retina after compression past 160μm, with a corresponding resistivity of about 300 Ω*cm. Impedance tomography showed a clear delineation between the approximately 200 Ω*cm photoreceptor layer and 1400 Ω*cm inner retina – from the inner nuclear layer to ganglion cell layer.
Conclusions :
Retinal compression and rupture by penetrating probes in previous studies may explain a wide variability of resistivity measurements. Impedance tomography does not alter retinal morphology and allows resolving several layers with different resistivities. A more precise map of retinal resistivity should improve the modeling accuracy and hence advance the field of retinal prosthetics. Further improvements in resolution and precision of the resistivity mapping may be possible with closer spacing (30μm) of the electrodes in MEA.
This abstract was presented at the 2024 ARVO Annual Meeting, held in Seattle, WA, May 5-9, 2024.