Modeling Motor and Sensory Axons

Surface electrical stimulation of the median nerve at the elbow can be used to elicit a sensation in the hand or fingers. However, perception of the stimulus can be strongly dependent on precise electrode placement. A model was needed to predict the ideal size and placement of electrodes to control activation of the nerve. To accurately predict activation of sensory and motor axons, a separate model for each axon type was required. The existing full axon model developed by McIntyre, Richardson, and Grill (2002) was created using the properties of a motor axon, but did not focus on the differences between motor and sensory axons. Previous work by this research team noted that sensation could be achieved at a lower stimulation voltage than movement. It was hypothesized that sensory axons would have a lower threshold voltage of activation than motor axons due to the differences in the properties of the membrane. The MRG (2002) model of a motor axon was modified to reflect the membrane properties of a sensory axon using motor-sensory comparisons found in literature. Both the motor and sensory models now include fast potassium channels and hyperpolarization-activated cyclic nucleotide-gated (HCN) channels. The parameters with the largest effect on threshold were found to be the conductance values of the fast sodium channel, HCN channels, and leak. The model results compare well with experimental measurements for conduction velocity, strength duration curves, and action potential shapes found in the literature. For a given fiber diameter, the sensory axon model predicted activation at a lower stimulation voltage than the motor axon model, similar to experimental results. The validated models can be used to predict activation of sensory and motor neurons in applications such as sensory feedback from prosthetics and treatment to increase neuroplasticity in patients after spinal cord injury, stroke, or amputation