Anil Kumar RamRakhyani, Zachary B. Kagan, David J. Warren, Richard A. Normann, Gianluca Lazzi, University of Utah, USA
There has been recurring interest in using magnetic neural stimulation for implantable localized stimulation. Accurate prediction of stimulator performance in animal or human subjects is an important step in the design and optimization of coil and stimulator parameters. Currently, predictions are based on the induced electric field simulation in a homogenous tissue environment which does not account for random fascicle distributions inside the nerve and high density axon distributions inside individual fascicles. In this work, we propose the use of an anatomically driven micrometer-resolution heterogeneous model of the sciatic nerve to simulate the magnetic field interactions with the axons. To improve the simulation’s prediction accuracy, we created a detailed computational model of the rat sciatic nerve based on a histological image of the sciatic nerve cross section. To simulate the field distribution inside a heterogeneous tissue environment, we developed a multiresolution impedance method that can compute the induced electric fields resulting from an applied magnetic field pulse. We combined this method with the Frankenhaeuser–Huxley (FH) axon model to simulate active neuronal membrane mechanisms. Therefore, such a hybrid multiphysics modeling approach is capable of studying the interaction of the coil’s magnetic field and individual axons. To validate the performance of the modeling framework, we used a variety of solenoid-shaped magnetic coils (each with different geometrical parameters) and predicted the stimulation threshold within the 95% confidence interval (experiments count = 4) of measured in vivo stimulation thresholds for the rat sciatic nerve. The presented simulation toolset is capable of studying the interaction of individual neurons/axons or neuronal networks with incident magnetic fields. Therefore, it can be effectively utilized in clinical settings (i.e., transcranial magnetic stimulation, spinal cord or peripheral nerve stimulation) to generate specific neuronal responses (e.g., stimulation selectivity) in the tissue depending on the coil’s shape, orientation, and position.
Keywords: Computational model, magnetic coil, magnetic stimulation, multiresolution model, multiscale modeling, peripheral nerve, sciatic nerve