In vivo Recording Quality of Mechanically Decoupled Floating Versus Skull-Fixed Silicon-Based Neural Probes.

Laetitia Chauvière,Patrick Ruther,Tobias Holzhammer,Oliver Paul,Frederick Pothof,Kai S Gansel,Wolf J Singer,Arno A A Aarts,Johanna Klon-Lipok

doi:10.3389/fnins.2019.00464

Abstract

Throughout the past decade, silicon-based neural probes have become a driving force in neural engineering. Such probes comprise sophisticated, integrated CMOS electronics which provide a large number of recording sites along slender probe shanks. Using such neural probes in a chronic setting often requires them to be mechanically anchored with respect to the skull. However, any relative motion between brain and implant causes recording instabilities and tissue responses such as glial scarring, thereby shielding recordable neurons from the recording sites integrated on the probe and thus decreasing the signal quality. In the current work, we present a comparison of results obtained using mechanically fixed and floating silicon neural probes chronically implanted into the cortex of a non-human primate. We demonstrate that the neural signal quality estimated by the quality of the spiking and local field potential (LFP) recordings over time is initially superior for the floating probe compared to the fixed device. Nonetheless, the skull-fixed probe also allowed long-term recording of multi-unit activity (MUA) and low frequency signals over several months, especially once pulsations of the brain were properly controlled.

Highlights

Stable, extracellular recording of cortical activity is mandatory for brain-computer interfaces used to accurately and reliably control a robotic arm (Velliste et al, 2008; Hochberg et al, 2012)
Over the course of 8 weeks, neuronal activity was recorded within the primary visual cortex of one awake behaving monkey
We compared the neuronal activity from both types of probes and found that local field potential (LFP) could be recorded from both probe types already 1 day after probe implantation (Figure 2B where exemplary signals are from recordings during Week 8)

Summary

Introduction

Stable, extracellular recording of cortical activity is mandatory for brain-computer interfaces used to accurately and reliably control a robotic arm (Velliste et al, 2008; Hochberg et al, 2012). Technical tools applied in this context include microwires, i.e., singles wires (Nicolelis et al, 2003) or tetrodes (Gray et al, 1995), flexible, polymer-based probes (Liu et al, 2015; Luan et al, 2017) as well as a variety of silicon-based micro-electrode arrays (Wise et al, 2008; Normann and Fernandez, 2016). Neural probes based on polymeric or silicon substrates comprise a large number of recording sites arranged along slender probe shafts. With up to 1600 electrodes along a 10-mmlong probe shaft (Herbawi et al, 2018), a pronounced increase in the number of recording sites is achieved in comparison to any other recording technology

Methods

Results

Discussion

Conclusion