Multi-scale simulations predict responses to non-invasive nerve root stimulation

OBJECTIVE Established biophysical neurone models have achieved limited success in reproducing electrophysiological responses to non-invasive stimulation of the human nervous system. This is related to our insufficient knowledge of the induced electric currents inside the human body. Despite the numerous research and clinical applications of non-invasive stimulation, it is still unclear which internal sites are actually affected by it. APPROACH We performed multi-scale computer simulations that, by making use of advances in computing power and numerical algorithms, combine a microscopic model of electrical excitation of neurones with a macroscopic electromagnetic model of the realistic whole-body anatomy. MAIN RESULTS The simulations yield responses consistent with those experimentally recorded following magnetic and electrical motor root stimulation in human subjects, and reproduce the observed amplitudes and latencies for a wide variety of stimulation parameters. SIGNIFICANCE Our findings demonstrate that modern computational techniques can produce detailed predictions about which and where neurones are activated, leading to improved understanding of the physics and basic mechanisms of non-invasive stimulation and enabling potential new applications that make use of improved targeting of stimulation.