Capacitive coupling within high-density microelectrode arrays can degrade neural recording signal or disperse neural stimulation current. Material deterioration in a chronically implanted neural stimulation/recording system can cause such an undesired effect. We present a simple method with an iterative algorithm to quantify the cross-coupling capacitance, in-situ.

The NeuroTalk TM interface, described in a companion paper allows for a standardized method of communication with implanted modules that contain custom application specific integrated circuits (ASIC). Here, we describe an example of one such ASIC that has been designed for use in a visual prosthesis. The ASIC is small enough to be incorporated within a 16-channel multielectrode stimulation array implanted in the cortex. In this paper we describe a version designed to operate over a 4-wire buss called NeuroTalk 2. Other versions of the ASIC operate using a single coil input for power and data, over a transcutaneous magnetic link.

With the availability of modern application specific integrated circuit (ASIC) design tools, simulation packages, and low-cost commercial silicon foundry processes, it is becoming increasingly easy for any laboratory, or small company, to develop a custom ASIC. For stimulation, as well as recording, chips that perform specialized functions can be designed, fabricated, and tested within a time period of 2-3 months. In many cases, the desired functionality can only be obtained by using VLSI design methods. Despite this increase in ASIC functionality, as related to neural engineering applications, there exists no common interface protocol for communicating with, and controlling, neural engineering ASICs. This would be analogous to each company that manufactures PC-based systems to have no common method of communication, e.g. USB, GBIB, RS-232, etc. While it might seem elusive, we propose the specification and development of a universal interface protocol for neural engineering ASICs. We have named this interface, NeuroTalktrade.

Over 200 iridium microelectrodes with activated iridium oxide (AIROF) charge-injection sites were evaluated in vitro by cyclic voltammetry and voltage transient measurements during current pulsing. Variability in the electrochemical behavior of these nominally identical microelectrodes was correlated with variations in the amount of AIROF present and the physical condition of the charge-injection tip. Electrolyte leakage under polymer insulation and delamination of AIROF both gave rise to characteristic shapes in cyclic voltammograms which could be used to identify their presence.