Modeling Surface Electrical Stimulation
Dr. Katharine Polasek, Engineering
Surface electrical stimulation is a non-invasive method for interfacing with the nervous system. However, variations in electrode placement, limb position and an individual’s anatomy make the results difficult to replicate, even in the same person. This project developed and validated a model to allow systematic varying of these key parameters to assess their contribution to nerve activation. The objective of this research in the future is to use the model to develop a surface electrical stimulation system that allows for precise nerve activation. A three-dimensional finite element model of the elbow was created, using two illustrated cross sections of the arm swept together and extruded in both directions. Three versions of the median nerve were modeled with 10 fascicles in random arrangements and 250 axons per fascicle. The fascicle diameters and distribution of axon diameters were based on measurements of a human median nerve at the elbow. Parameters were adjusted to reflect the differences in channel properties between motor and sensory axons; fifteen percent of the axons had motor properties and the remaining axons had sensory properties. Simulations were run, applying a voltage across two electrodes placed on the skin. The threshold voltages predicted by the model were consistent with experimental results from human subjects. In addition, the model predicted 20% activation of sensory axons before any motor axons, in the same way that sensory perception occurred before muscle movement experimentally. Using an ANOVA analysis, it was found that voltage, axon diameter, axon type, and fascicle location were all significant predictors of percent activation. The model has been validated with experimental results from human subjects. Next, the model will be used to analyze differences in nerve activation for different electrode arrangements and an algorithm will be developed to adjust and optimize nerve activation for various individuals.
A recommended citation will become available once a downloadable file has been added to this entry.