The History and Horizons of Microscale Neural Interfaces

Abstract

Microscale neural technologies interface with the nervous system to record and stimulate brain tissue with high spatial and temporal resolution. These devices are being developed to understand the mechanisms that govern brain function, plasticity and cognitive learning, treat neurological diseases, or monitor and restore functions over the lifetime of the patient. Despite decades of use in basic research over days to months, and the growing prevalence of neuromodulation therapies, in many cases the lack of knowledge regarding the fundamental mechanisms driving activation has dramatically limited our ability to interpret data or fine-tune design parameters to improve long-term performance. While advances in materials, microfabrication techniques, packaging, and understanding of the nervous system has enabled tremendous innovation in the field of neural engineering, many challenges and opportunities remain at the frontiers of the neural interface in terms of both neurobiology and engineering. In this short-communication, we explore critical needs in the neural engineering field to overcome these challenges. Disentangling the complexities involved in the chronic neural interface problem requires simultaneous proficiency in multiple scientific and engineering disciplines. The critical component of advancing neural interface knowledge is to prepare the next wave of investigators who have simultaneous multi-disciplinary proficiencies with a diverse set of perspectives necessary to solve the chronic neural interface challenge.

Keywords: BRAIN Initiative; bias; biocompatibility; diversity; education; micro-electromechanical systems (MEMS); micromachine; multi-disciplinary; neuroscience; training.

Figure 1

Classes of microscale implantable neural technologies: (a) 50 μm polyimide-insulated tungsten microwire with chiseled tips (Tucker–Davis Technologies, Alachua, FL, USA); (b) microfabricated silicon Michigan array with iridium electrode sites (NeuroNexus Technologies, Ann Arbor, MI, USA), scale = 100 μm; (c) macromachined boron-doped silicon array (Blackrock Microsystems, Salt Lake City, UT, USA), each needle is electrically separated at the base with glass. Scale = 400 μm.

Figure 2

Advances in microscale neural interfaces: (a) 64-channel Buszaki Array (Neuronexus); (b) 128-channel Matrix Array (Neuronexus); (c) 24-channel ultra-small carbon fiber array on silicon stacks (courtesy of Paras Patel/Cynthia Chestek), scale = 100 μm; (d) high-density ultra-small microwire array (Paradromics Inc., San Jose, CA, USA), scale = 500 μm; (e) μLED silicon optoelectrode (courtesy of NeuroNex MINT Hub at University of Michigan, Ann Arbor, MI, USA (http://mint.engin.umich.edu)), scale = 100 μm; (f) a standard-sized 1.27 mm diameter Lawrence Livermore National Laboratories (LLNL) DBS-style penetrating probe constructed using microfabrication techniques, allowing for a higher-density of electrodes and avoiding typical hand-assembly techniques; and (g) A LNLL 128-channel microelectrocorticography (µECoG) array used for language mapping on awake patients. This 20-µm-thick flexible electrode array is constructed using thin-film polymers and metals and features 1.2 mm diameter electrodes.