Multi‐Unit Recording

Chronic, Multi-Site Recordings Supported by Two Low-Cost, Stationary Probe Designs Optimized to Capture Either Single Unit or Local Field Potential Activity in Behaving Rats

“…Unlike with microdrives, there is no ability to move electrode sites vertically. The collection of spiking activity and field potentials may be affected by acute injuries of neural tissue, including inflammation, severe bleeding and chance of implanting in a “dead zone” ( 54 , 73 , 99 ). The histopathology does confirm damage around the electrode tracks after being implanted chronically for months ( Figure 4 ).…”

“…Insulated metal wires are commonly used as the voltage-collecting interface in single unit recordings. Many different sizes (5–50 μm diameter) and types of wires (including tungsten, stainless steel, and nichrome) have been used to collect single units ( 54 , 71 , 72 ). Wire size influences the number of neurons recorded simultaneously, signal-to-noise ratio, and electrode impedance levels ( 73 , 74 ).…”

“…Fixed devices can have up to 384 recording sites, targeting many brain areas simultaneously as is seen with the Neuropixels silicon probe (46)(47)(48). The other approach utilizes a microdrive to move electrodes along the vertical axis to advance after surgical implantation (52)(53)(54)(55)(56)(57)(58). Microdrives are advantageous for reaching neural dense areas and moving past the site of initial injury/ inflammatory response, however fabrication can be more challenging and expensive than fixed implants and are typically limited in the number of multi-site trajectories they can support (52,53,(55)(56)(57).…”

“…Table 3 compares our probe design and quantifiable results with other custom and commercial probes. (54,71,72). Wire size influences the number of neurons recorded simultaneously, signal-to-noise ratio, and electrode impedance levels (73,74).…”

“…Therefore, we use "polytrode" sorting to compare spikes captured by 16 wires in a fixed bundle to provide a conservative metric and avoid a bias cell count driven from units that may have been recorded on multiple channels. Our probes are not arranged like a tetrode (four recording sites 25-50 µm apart) or polytrode (8-64 channels 50-70 µm apart) with fixed distances to capture multiple views of the same neuron, but since the goal of clustering is to find similarity between spikes we can sort in multiple channels, acknowledging that we are not sorting any better than a single microwire (54,73,80,81,(84)(85)(86). Signals were processed as 16 electrode groups during referencing (two polytrode groups for 32-channel probes; four polytrode groups for 64-channel probes) selected because of their fixed arrangement into different cannula during fabrication.…”