Project 3: Mechanics of Long-Shank 5 mm Neural Probe Insertion into the Rat Brain: Effects of Geometry and Vibration-Assisted Insertion
Motivation
Ataxia is a neurological condition characterized by impaired movement coordination, affecting balance, walking, speech, and fine motor control. It commonly results from damage or degeneration of the cerebellum, a brain region responsible for coordinating movement, or its connections to other parts of the nervous system. This dysfunction can arise from inherited genetic disorders, neurodegenerative diseases, or acquired injuries. Research into movement coordination and disorders such as ataxia requires reliable recordings from neural circuits that extend beyond the cortical surface. Accessing deep brain structures, including the dorsal striatum, hippocampus, and thalamus, requires probes capable of reaching depths of 3–5 mm in rodents. However, longer probes are more susceptible to buckling during insertion, particularly when their cross-sectional area is reduced to limit tissue damage. We investigated how probe geometry, insertion speed, and vibration-assisted insertion affect the mechanical reliability of 5 mm neural probes to support deep-brain recording studies.
Our Solution
We designed and evaluated two 5 mm amorphous silicon carbide (a-SiC) probes with identical thickness (~12 µm) and tip angle (11°): a uniform shank measuring 175 µm wide and a tapered shank narrowing from 175 µm proximally to 75 µm at the tip. The tapered design reduces the area of the penetrating tip while retaining a wider proximal section for stiffness. We used finite element modeling (FEM) in COMSOL to predict the critical buckling force of both geometries, then compared these predictions with experimental measurements. Insertion forces were characterized in 1.2% agarose gel at four speeds (20, 100, 500, and 1000 µm/s), followed by acute insertions in rat cortex. We also tested vibration-assisted insertion using a NeuralGlider inserter at 0.5 W actuation.
Figure 1: Probe design: (a) comparison of uniform and tapered probe geometries; (b) scanning electron microscopy (SEM) of tip profiles of tapered and uniform designs defining the 11° tip angle geometry.
My Contributions
I measured the critical buckling force of both probe designs by driving each probe against a rigid silicon wafer using a NeuralGlider inserter and a 20 g S-beam load cell. I prepared the 1.2% agarose gel phantoms and conducted insertion tests at 20, 100, 500, and 1000 µm/s, recording force-displacement curves to quantify maximum force (Fmax), penetration force (FP), and dimpling depth (Dd). Based on these measurements, 100 µm/s provided the best balance of insertion force, success rate, and practical insertion speed. I then performed in vivo insertions of both probe designs into rat cortex to a depth of 4 mm under stereotaxic guidance, including additional trials with vibration assistance enabled for the tapered probe.
Multi-shank microelectrode arrays (MEAs) can record neural activity across multiple locations, but inserting several shanks simultaneously can cause substantial cortical deformation before tissue penetration. This deformation, known as cortical dimpling, can strain surrounding tissue and blood vessels. Reducing these insertion forces while maintaining sufficient stiffness to prevent shank buckling is an important consideration in electrode design. We investigated whether staggering the lengths of individual shanks could reduce insertion-related tissue deformation without requiring additional surgical equipment.
Our Solution
In collaboration with NeuroNexus, we developed a four-shank MEA with staggered lengths of 2.2, 2.4, 2.75, and 3 mm. Each shank had a thickness of 6 µm at the recording tip and 16 µm proximally. The staggered lengths allow the shanks to penetrate the cortex sequentially rather than simultaneously, while the thicker proximal sections maintain structural stiffness. Finite element modeling showed that the predicted critical buckling force exceeded cortical penetration force by at least sixfold across all shanks. We compared the staggered design against a conventional NeuroNexus array with four equal-length 3 mm shanks through in vivo insertions in rat cortex, measuring penetration force, cortical dimpling, and total insertion energy.
My Contributions
I performed the in vivo insertions and collected force-displacement measurements for both the staggered (STGRD_NNx) and conventional (CONV_NNx) arrays. Using a NeuralGlider inserter equipped with a 20 g S-beam load cell, I inserted each array into rat somatosensory cortex at 100 µm/s to a depth of 2 mm, completing five trials per group.
Figure 2: Insertion mechanics of STGRD_NNx and CONV_NNx MEAs. a) Schema illustrating the sequential penetration of STGRD_NNx MEA shanks in brain tissue, b) Representative force-displacement curves for both STGRD_NNx and CONV_NNx MEAs. Different insertion onsets for each staggered shank are labeled to match the schematic in (a), c) Example force-displacement curve showing no distinct penetration peaks for the STGRD NNx MEA.
Project 1: Chronic Stability of Ruthenium Oxide Electrodes in Rat Motor Cortex
Motivation
Implanted neural electrodes must maintain stable electrical properties and recording performance over time. One important factor is the electrode coating, which affects how electrical signals are transferred between neural tissue and the recording system. Iridium oxide coatings, including sputtered iridium oxide film (SIROF), are widely studied but rely on an expensive and relatively scarce material. Ruthenium oxide (RuOx) has shown promising electrochemical properties as a lower-cost alternative, although its long-term performance in brain tissue remains less characterized. We investigated whether RuOx-coated electrodes could maintain stable electrochemical properties and neural recordings over 16 weeks of implantation.
Our Solution
We implanted RuOx-coated amorphous silicon carbide (a-SiC) microelectrode arrays into rat motor cortex and monitored their performance weekly for 16 weeks. Electrochemical impedance spectroscopy (EIS) was used to measure the resistance and frequency-dependent electrical behavior of the electrode-tissue interface, while cyclic voltammetry (CV) was used to assess charge storage capacity. Neural recordings collected during the same sessions were analyzed for active electrode yield, signal amplitude, and signal-to-noise ratio. These measurements allowed us to track both the electrochemical properties of the electrodes and their ability to record neural activity throughout the implantation period.
My Contributions
I led experimental data collection, beginning with baseline in vitro electrochemical characterization before implantation. I then conducted weekly in vivo EIS and CV measurements using a Gamry potentiostat over the 16-week study, measuring 1 kHz impedance and charge storage capacity at sweep rates of 50 and 50,000 mV/s. I also collected weekly neural recordings and analyzed active electrode yield (AEY), peak-to-peak voltage (Vpp), and signal-to-noise ratio (SNR). Electrochemical performance remained stable through week 16, with active electrode yield ranging from 40–80% and recorded signal amplitudes of approximately 70–170 µV, without a statistically significant trend in Vpp over time (p = 0.17).
