Multichannel thin-film electrode for intramuscular electromyographic recordings

Dario Farina, Ken Yoshida, Thomas Stieglitz, Klaus Peter KochJournal of Applied Physiology 2007 · DOI

TL;DR

Researchers have developed a novel multichannel thin-film electrode system for intramuscular electromyography (EMG). This flexible, microfabricated electrode allows for unprecedented spatial sampling of motor unit activity within muscles, paving the way for more detailed insights into neural control and potentially advanced prosthetic interfaces.

What Problem Does This Paper Address?

Understanding how the nervous system controls movement relies heavily on analyzing the electrical activity of individual motor units within muscles. Traditional methods, like single-wire or needle electrodes, are limited in their ability to capture signals from many locations simultaneously inside a muscle. This "spatial sampling" limitation means researchers can only investigate a small fraction of the active motor units in a localized area, hindering a comprehensive understanding of muscle control strategies. Existing multichannel systems often use closely spaced needle electrodes, which are cumbersome, impractical for dynamic movements, and uncomfortable. The paper addresses this by proposing a new, flexible thin-film electrode technology that enables high-density spatial sampling of intramuscular EMG signals.

Methods

The core innovation lies in the fabrication of a thin-film electrode system using microfabrication techniques on a silicon wafer. This process creates a flexible polyimide-based structure measuring 220 µm wide and 10 µm thick, featuring eight circular platinum-platinum chloride recording sites (40 µm diameter) distributed on both its front and back surfaces. The inter-site spacing is 1,500 µm.

For implantation into rabbit gastrocnemius muscles, the thin-film electrode was attached to a fine tungsten needle via polyaramid filaments. This needle served solely as an introducer, allowing the flexible electrode to be precisely positioned perpendicular to muscle fiber orientation. After insertion, the needle and filaments were removed, leaving the thin-film structure embedded in the muscle.

EMG signals were amplified and filtered using ultra-low-noise preamplifiers. Signal analysis involved a decomposition algorithm to identify individual motor unit action potentials from the multiunit recordings. This algorithm allowed for the merging of signals from different channels to identify the same motor unit detected at multiple sites.

Datasets

Not discussed in the paper. The study utilized data collected from six experiments on anesthetized rabbits.

Key Insights and Ideas

  • High-Density Spatial Sampling: The thin-film technology enables a significantly higher density of recording sites within a small muscle volume compared to traditional methods. This is crucial for capturing the complex spatial distribution of motor unit activity.
  • Flexibility and Minimally Invasive Insertion: The thin, flexible nature of the electrode, combined with the needle-insertion technique, offers a less invasive and more adaptable approach for intramuscular recordings, especially compared to rigid needle electrodes.
  • Potential for Advanced Prosthetics: The ability to precisely record motor unit signals from multiple locations opens doors for more sophisticated myoelectric control of prosthetics, allowing for finer discrimination of neural commands.
  • Robustness to Sterilization: The thin-film system demonstrated resilience to various sterilization procedures, including autoclaving, which is essential for reusable medical devices.

Key Results

  • Motor Unit Identification: The system successfully identified 67 motor units across the experiments.
  • Signal Quality: The peak-to-peak amplitude of detected motor unit action potentials was 75 ± 12 µV, with a root mean square (RMS) noise level of 1.6 ± 0.4 µV after offline band-pass filtering (200 Hz to 5 kHz).
  • Stability: The noise level and action potential amplitude remained consistent for measurements separated by up to 40 minutes, even with passive limb movements, indicating good system stability.
  • Spatial Filtering Effectiveness: Linear combinations of signals from adjacent detection sites (single, double, and triple differentials) were shown to attenuate or amplify motor unit action potentials, demonstrating the utility of spatial filtering for signal enhancement and selectivity.

Limitations and Future Work

  • Insertion Modality: The current insertion method, involving a needle passing through the muscle, is suitable for animal studies but would require adaptation for in-vivo human implantation (e.g., bending the electrode tip).
  • Limited Number of Channels (Prototype): The prototype was limited to eight channels as a proof of concept. The technology, however, allows for scaling to a larger number of channels and intersite distances, which would increase the sampled muscle area.
  • Experimental Conditions: The tests were performed on animals with open skin and exposed muscle. Future work should focus on subcutaneous insertion in humans.
  • Voluntary Contraction Analysis: The study used induced reflexes and nerve crushes. Future research could explore the system's performance during voluntary contractions, which involve more complex motor unit recruitment strategies.
  • Optimization of Design: While the current design is based on successful nerve electrode technology, further optimization of electrode geometry and intersite spacing could be beneficial for muscle recordings.

Summary updated

Abstract

It is currently not possible to record electromyographic (EMG) signals from many locations concurrently inside the muscle in a single wire electrode system. We developed a thin-film wire electrode system for multichannel intramuscular EMG recordings. The system was fabricated using a micromachining process, with a silicon wafer as production platform for polyimide-based electrodes. In the current prototype, the flexible polymer structure is 220 microm wide, 10 microm thick, and 1.5 cm long, and it has eight circular platinum-platinum chloride recording sites of 40-microm diameter distributed along the front and back surfaces with 1,500-microm intersite spacing. The system prototype was tested in six experiments where the electrode was implanted into the medial head of the gastrocnemius muscle of rabbits, perpendicular to the pennation angle of the muscle fibers. Asynchronous motor unit activity was induced by eliciting the withdrawal reflex or sequential crushes of the sciatic nerve using a pair of forceps. Sixty-seven motor units were identified from these recordings. In the bandwidth 200 Hz to 5 kHz, the peak-to-peak amplitude of the action potentials of the detected motor units was 75 +/- 12 muV and the root mean square of the noise was 1.6 +/- 0.4 muV. The noise level and amplitude of the action potentials were similar for measures separated by up to 40 min. The experimental tests demonstrated that thin film is a promising technology for a new type of flexible-wire intramuscular EMG recording system with multiple detection sites.
BibTeX
@article{farina2007multichannel,
  author = {Dario Farina and Ken Yoshida and Thomas Stieglitz and Klaus Peter Koch},
  title = {{Multichannel thin-film electrode for intramuscular electromyographic recordings}},
  journal = {Journal of Applied Physiology},
  year = {2007},
  doi = {10.1152/japplphysiol.00788.2007},
}
Figure