An ASIC for Recording and Stimulation in Stacked Microchannel Neural Interfaces

Abstract

This paper presents an active microchannel neural interface (MNI) using seven stacked application specific integrated circuits (ASICs). The approach provides a solution to the present problem of interconnect density in three-dimensional (3-D) MNIs. The 4 mm 2 ASIC is implemented in 0.35 µm high-voltage CMOS technology. Each ASIC is the base for seven microchannels each with three electrodes in a pseudo-tripolar arrangement. Multiplexing allows stimulating or recording from any one of 49 channels, across seven ASICs. Connections to the ASICs are made with a fiveline parallel bus. Current controlled biphasic stimulation from 5 to 500 µA has been demonstrated with switching between channels and ASICs. The high-voltage technology gives a compliance of 40 V for stimulation, appropriate for the high impedances within microchannels. High frequency biphasic stimulation, up to 40 kHz is achieved, suitable for reversible high frequency nerve blockades. Recording has been demonstrated with mV level signals; commonmode inputs are differentially distorted and limit the CMRR to 40 dB. The ASIC has been used in vitro in conjunction with an oversize (2 mm diameter) microchannel in phosphate buffered saline, demonstrating attenuation of interference from outside the microchannel and tripolar recording of signals from within the microchannel. By using five-lines for 49 active microchannels the device overcomes limitations when connecting many electrodes in a 3-D miniaturized nerve interface.

A dual-threshold adaptive action potential detector for neural recording systems

“…In this design, the bias currents Ibl and Ib2 are adjusted simultaneously to achieve the bias current adjustment of M11 and the input tubes M7 and M8, followed by changing the span conductance of the input tubes and the resistor value of the source level negative feedbackand finally changing the overall The equivalent cross-conductance of the circuit is changed to increase the adjustable range of the filter cut-off frequency. [8] VDD…”

A Bidirectional ASIC for Active Microchannel Neural Interfaces

“…High-density microchannel neural interfaces (MNIs) can overcome the limitations of electrode interconnects via multiplexed, on-chip contacts that access the channels more easily and limit device complexity. Moreover, on-site electronic circuits capable of stimulation and recording of the neural signals would allow concurrent multi-channel operation and high-quality acquisition of action potentials, while the structure of the MNI enables scaling of the system by stacking multiple units [4].…”

“…Post-layout simulation results are shown in Section IV. Discussion and concluding remarks are presented in Section V. A. Neural Excitation Electrical stimulation of the neural fibres in miniaturised interfaces requires electrode voltages above 10 V to deliver the necessary currents, which range from 5 µA to 25 µA in microchannels [4]. As the channel impedances reach 1.24 MΩ [3], activation of the fibres requires a voltage compliance exceeding 30 V. Restrictions in the maximum drain-to-source (Vds) and gate-to-source (Vgs) voltages of standard CMOS transistors require advanced methods such as transistor stacking [7] or HV shielding [8] to operate safely and integrate the HV output stage with the LV circuits.…”

Advances in Neural Recording and Stimulation Integrated Circuits

“…In these circuits, the stimulator has been widely used in biomedical applications for decades, such as cardiac pacemaking, cochlear/retinal prosthesis, and cell activation (Chen et al, 2010;Sooksood et al, 2011;Noorsal et al, 2012;Wagner et al, 2018;Lee and Im, 2019;Lin and Ker, 2020;Yen and Ker, 2020). The neural recording circuit is also involved in these applications to sense the neural signal or assess stimulation efficacy and the tissue status to enable closed-loop control in simultaneous neural recording and stimulation (Yoshida and Horch, 1996;Blum et al, 2007;Rolston et al, 2009Rolston et al, , 2010Venkatraman et al, 2009;Xu et al, 2012;Ando et al, 2016;Ramezani et al, 2018;Lancashire et al, 2019;Carmona et al, 2020). The circuits for simultaneous neural recording and stimulation are used in neural prostheses, such as the bionic neural link for limb function restoration (Xu et al, 2012;Sadeghi Najafabadi et al, 2020;Żebrowska et al, 2020).…”