Automated Distributed Element Model Generation for Neural Interface Co-Design

We present a scripted distributed element modeling framework to enable process-portable co-design of neural recording and stimulation circuits and other applications that require co-design with an electrode-electrolyte interface. Using distributed elements enables designers to simulate the spatial voltage and current profiles in tissue to determine key parameters such as stimulator voltage headroom, stimulation artifact, and charge-balance. Designers specify 2D or 3D physical parameters of the electrode configuration in MATLAB, which in turn generates a netlist in Cadence Virtuoso for simulation with circuitry. Using this framework, we show that time-domain artifact cancellation techniques outperform frequency-domain techniques for concurrent neural recording and stimulation.