Artifact-Recovery in Neuromodulators using Tunable High-Pass Corners

Abstract

In neuromodulation systems, stimulation causes artifacts which are orders of magnitude larger than the recorded neural signal. Several recent publications have proposed recorder front-ends with high dynamic range to cope with stimulation artifacts, thereby avoiding blanking switches and the resulting recorder blind-time. However, this usually comes at the cost of limited first-stage amplification due to signal swing requirements, thus degrading the noise efficiency factor (NEF). In this paper, we propose to instead combine blanking switches with a tuning mechanism on the pseudo-resistor based high-pass corner frequency of the frontend. This allows to temporarily increase the settling speed during artifact-recovery, such that undisturbed recording can commence few tens of milli-seconds after e.g., a stimulation event. The artifact-recovery scheme is demonstrated using an integrated state-of-the-art neuromodulator, and in-vitro measurements are presented to validate its usability.