(32) High-Input-Range Recording for Investigating Direct, Stimulation-Evoked Neural Responses

Introduction: Electrical neuromodulation has enabled innovative treatments for various neurological disorders, yet the underlying biophysical mechanisms driving therapeutic outcomes are not fully understood. Recording neural activity during stimulation can be useful for investigating these mechanisms. However, neural recording hardware is sensitive to large voltage transients and often saturates during stimulation, preventing the system from capturing evoked neural activity during and immediately after stimulation.

Methods: We developed a neuromodulation system with high-input-range recording amplifiers and programmable, current-controlled stimulation sources to investigate the mechanisms of neuromodulation typically masked by artifacts. The device design enables concurrent recording and stimulation through the same electrode and is less susceptible to saturation during stimulation, enabling uninterrupted recordings of neural activity at the stimulation site. We designed a MATLAB user interface to configure the hardware, set stimulation parameters, visualize live recording, and save data. We performed in vitro testing of the system in phosphate-buffered saline with both platinum-iridium microprobes and cortical surface microelectrodes. We delivered symmetric, biphasic stimulation pulses with incrementally increasing amplitudes to determine current ranges that can be delivered without causing saturation. We also recorded a simulated neural signal while delivering stimulation and applied artifact subtraction techniques to recover neural activity directly after stimulation. We performed in vivo testing of the system in a non-human primate using a hybrid system that allows implanting the intracortical microprobes and a cortical surface microelectrode array in the same brain location. We recorded while delivering symmetric, biphasic stimulation pulses with each electrode type to investigate the capabilities of the system in vivo.

Results: We found that currents above 200 µA could be delivered without saturation for both microprobes and surface microelectrodes in saline. With current that nearly causes saturation on the stimulating electrode, we found the voltage transients on neighboring electrodes to be one to two orders of magnitude smaller. By applying artifact removal techniques, simulated neural data could be recovered ~50 µs after stimulation, which we estimated to be 1.5 – 9x faster than commercial devices. In vivo stimulation with microprobes saturated with 50 µA, while surface microelectrodes could deliver over 200 µA without saturation.

Conclusion: Future work will involve modifying the hardware to improve artifact mitigation and testing the effect of various stimulation parameters on neural responses in vivo. This work aims to advance our understanding of the therapeutic mechanisms of neuromodulation to support the development of more effective treatments for neurological disorders.