BMI Systems - Brain-Machine Interface Engineering - page 203

sites typically consist of exposed metal pads located on rigid shanks that are connected via intercon-
nect traces to output leads or to signal processing circuitry on a monolithic substrate. Some place
multiple recording sites along the length of the shank to allow for interrogating cells at varying
depths in the neural tissue [ 36 , 37 ]. An electrode design known as the Utah probe [ 38-40 ] uses
micromachined square arrays of 100 single contact electrodes bonded to a demultiplexer connec-
tor with local amplification. Despite advances in micromachined neural probes, the fixed nature of
the probes limit the ability to optimize the recording of the neural signal. If the probe is spatially
far from the neuron of interest, the measured action potential is low. The multisite micromachined
neural probe provides one approach to resolve this issue. By using a multiplexer to sample the neural
signal at different sites at different depths into the neural tissue, a discrete number of locations may
be monitored without additional surgery to reposition the probe. An alternative approach is to use
moveable probes instead of fixed probes. It is well known that the ability to reposition electrodes
after implantation surgery can improve the signal-to-noise ratio (SNR) and yield of neuronal re-
cordings. A variety of microdrives have been proposed that either mechanically or electrically move
arrays of electrodes in unison or independently with manual or automated mechanisms. Fabrication
of microdrives often requires sophisticated miniature mechanical components and significant skill,
time, and cost. For precise targeting of individual units, the probe excursion should traverse 1-2 m
in intervals of 5-30 µm. Recently, motorized screw microdrives have been reported to provide a
10-fold increase in yield compared to manual drives [ 41 ]. In addition, miniature piezoelectric linear
actuators require no gears or screws for electrode movement [ 31 ]. Physically, the design challenges
also involve specifying small form factors and lightweight components that can fit on the skull of
the patient.
Rigid substrate micromachined probes described e may introduce chronic tissue damage
due to mechanical forces encountered from strain transferred from the mount to the probes float-
ing in neural tissue [ 42 ]. As one way to mitigate this challenge, Rousche et al. [ 43 ] designed a
flexible substrate microelectrode array, where thin metal electrode sites and subsequent wiring are
enclosed between polymers. This flexible electrode array adds needed strain relief, yet cannot be
inserted into the brain matter directly. An incision must be created first in order for the electrode
to be implanted. Ref. [ 44 ] incorporates rigid probes with planar electrode sites on the probe shank
hybrid-packaged to a flexible cable that connects electrodes to output ports. This design may be
inserted directly into neural tissue, keeping damage to a minimum, and provides greater reliability
of the probe assembly.
Here, we report a neural microelectrode array design that leverages the recording properties
of conventional microwire electrode arrays with the additional features of precise control of the
electrode geometries and flexible micromachined ribbon cable integrated with the rigid probes. The