Polymer Cuff Electrodes Prove Viable for Selective Peripheral Nerve Stimulation

Peripheral nerve stimulation has long held promise as a treatment for chronic pain, neuromuscular rehabilitation, and nerve injury. The clinical bottleneck, however, is precision. Most implantable devices apply bulk electrical stimulation to an entire nerve trunk, activating multiple fibre types simultaneously and producing unwanted side effects — from discomfort to off-target muscle contraction.

Achieving what researchers term “fascicular selectivity” — the ability to target one bundle of nerve fibres within a mixed nerve, without recruiting its neighbours — has become a defining challenge in neural interface engineering.

A recent 2026 study from Imperial College London, published in the Journal of Neural Engineering, takes a significant step toward meeting that challenge. The research group, led by Dr Rylie A Green in the Department of Bioengineering, developed a fully polymeric transverse multipolar cuff electrode, fabricated from a conductive elastomer (CE), and demonstrated that it can achieve reliable selective fascicular activation without a single metallic contact.

A Softer Interface for Peripheral Nerves

Conventional nerve cuff electrodes are built around metallic conductors — most commonly platinum or platinum-iridium alloys. While these materials offer excellent charge injection capacity and long-term stability, their rigidity limits conformal contact with soft nerve tissue, and they introduce artefacts in computed tomography imaging that obscure the precise anatomical relationship between electrode sites and underlying fascicles.

Conductive elastomers address both constraints. The CE used in this study — a PDMS-based composite developed by Cuttaz et al and characterised in prior work published in Biomaterials Science and Advanced Science — retains an electrical conductivity of 660 S m⁻¹ while matching the mechanical compliance of soft biological tissue. An eight-contact transverse electrode array was laser-patterned from CE sheet, encapsulated in polydimethylsiloxane, and sized for the 0.9–1.1 mm diameter of a rat sciatic nerve.

Electrical interconnects within the device were formed using 125 µm diameter polytetrafluoroethylene-insulated silver wire supplied by Advent Research Materials, bonded to the CE electrode contacts and encapsulated in medical-grade silicone. The pairing of CE electrode sites with precision silver interconnects produced a fully polymeric, mechanically flexible device assembly suitable for conformal nerve wrapping.

What the Experiments Demonstrated

The electrode arrays were evaluated ex vivo on sciatic nerves harvested from five female Sprague-Dawley rats. The sciatic nerve divides into three fascicles — tibial, peroneal, and sural — each of which was recorded independently using bipolar silver hook electrodes in a purpose-built nerve chamber. Compound nerve action potentials were captured for each fascicle across a broad parameter sweep: stimulation amplitudes from 50 to 1600 µA, five waveform configurations spanning bipolar and tripolar nearest-neighbour electrode couplings, and both symmetric and asymmetric biphasic waveforms.

Fascicular selectivity was quantified using a selectivity index (SI), in which positive values indicate that one fascicle is being preferentially recruited over the other two.

Across all five nerve trials, each fascicle achieved selective activation at SI values greater than 0.65 — a result comparable to previously reported platinum-based multipolar cuffs. The CE electrodes maintained stable electrochemical performance when placed on the nerve, with average charge storage capacity of 113.3 ± 39.12 mC cm⁻² ex vivo, approximately fifty times greater than comparable platinum electrode measurements.

A notable finding was that neuroanatomical variation between individual nerves exerted substantially greater influence on achievable selectivity than differences in waveform configuration. Tripolar electrode configurations produced marginally higher selectivity than bipolar arrangements — consistent with their more localised electric field distribution — but the effect was not statistically significant across fascicles. The neuroanatomy of any given nerve, it appears, is a more decisive determinant of stimulation outcome than the precise waveform delivered.

Pairing Experiment With Anatomically Informed Simulation

A distinguishing feature of this work was the integration of microCT imaging with the ASCENT computational modelling pipeline. Because CE electrodes absorb the iodine-based Lugol’s contrast agent used for microCT staining — rather than blocking it as metallic arrays would — the full three-dimensional geometry of each cuff and the underlying fascicular anatomy could be imaged directly after each ex vivo session, without imaging artefacts.

These anatomically precise reconstructions were imported into COMSOL-based finite element models within the ASCENT framework, and the same stimulation parameter sweeps applied experimentally were run in silico. The simulations reproduced selective activation patterns in a number of cases, but systematic discrepancies emerged across the nerve cohort. The sural fascicle — the smallest of the three — was consistently overpredicted for selectivity by the model, while the tibial fascicle — the largest — was underpredicted by approximately 15%. The research team identified inaccurate modelling of perineurial charge transfer as a probable contributor, pointing to the need for improved tissue parameter assignment, particularly around fascicle-size-dependent perineurial thickness and resistivity, in future simulation frameworks.

Implications for Neural Interface Research

This study confirms that fully polymeric electrode cuffs can match the performance of metallic designs for selective nerve stimulation — while adding a practical advantage: they allow direct, artefact-free imaging of the device in place. That closes a long-standing gap between where a device is placed and what the computational model assumes about its position.

For groups developing next-generation implantable neural interfaces, the results reinforce two practical conclusions: flexible polymer electrode materials can achieve clinical-grade fascicular selectivity, and inter-subject neuroanatomical variation is a more significant determinant of outcome than choice of stimulation waveform.

The full paper is available open access via IOP Publishing: doi.org/10.1088/1741-2552/ae44d0.

Source: Bailey Z K et al (2026) J. Neural Eng. 23 016032. https://doi.org/10.1088/1741-2552/ae44d0