Novel annular electrodes for a cortical visual prosthesis

The research presented here was conducted within the scope of the Monash Vision Group’s (MVG) development of a cortical visual prosthesis “bionic eye” to provide visual information to blind people, using electrodes that penetrate into primary visual cortex to stimulate layer 4 and evoke the perception of spots of light called phosphenes. Platinum/iridium (PtIr), the electrode material used by MVG, is estimated to have a charge injection capacity of 50 - 150 µC/cm²/phase, much lower than the charge densities required for phosphene perception using standard electrode designs with small exposed stimulation tips. To overcome this, MVG proposed a novel design with an annular active area on the electrodes as the larger active surface area could reduce the charge densities required for effective stimulation. My original contribution is the engineering and biomedical examination of this design for effective cortical stimulation within the electrochemical “safe” limits for PtIr. I first used analytical and finite element modelling for theoretical predictions. Analytical modelling showed that due to the kill zone (the region of dead nerve cells) expected around each electrode after implantation into cortex, the electrode surface area for minimum power consumption during threshold stimulation is 49,700 µm, much larger than the surface areas of stimulating intracortical electrodes currently used. Thus, increasing the electrode’s surface area from current designs will reduce the power consumption, an important consideration in the wirelessly powered device MVG proposes to develop. Finite element modelling demonstrated the annulus electrode has a more uniform primary current density distribution than the commonly used tip electrode, likely reducing any tissue or electrode damage during stimulation. Balancing the advantage of increasing electrode surface area in reducing power consumption, is that a smaller surface area is needed for specificity and resolution of stimulation. In acute studies in rats I determined the minimum surface areas of the annular electrodes for safe and effective stimulation. The effect of coating electrodes with sputtered titanium nitride (TiN) to increase charge injection capacity was also studied. Using electrodes with varying surface areas, the charge injection capacity of TiN-coated and uncoated PtIr electrodes was measured with voltage transients, and compared to charge densities to evoke a motor response in anaesthetised rats. The minimum geometric surface areas for effective stimulation within safe limits were 38,000 µm² and 16,000 µm² for uncoated and TiN-coated PtIr electrodes respectively. Finally, chronic studies were performed to test the safety and stability of stimulation over the long term when a glial sheath forms around the electrode through the brain’s inflammatory response to a foreign object. Charge densities of up to 71.4 µC/cm²/phase were measured, which is greater than the charge injection capacity of 18 µC/cm²/phase for PtIr found in the acute studies. No neuronal loss or evidence of electrode degradation occurred that could be attributed to stimulation,suggesting less conservative safety limits could be used. Overall, the results support the use of annular electrodes in stimulating cortical prostheses. It is recommended that the electrodes are coated with TiN so that smaller electrodes can be used and device resolution increased.