Computational modeling of the effects of transcranial magnetic stimulation on cortical neurons

• Autor(a) principal: Cuziol, Vitor Valsichi
• Data de Publicação: 2020
• Tipo de documento: Dissertação
• Idioma: eng
• Título da fonte: Biblioteca Digital de Teses e Dissertações da USP
• Texto Completo: https://www.teses.usp.br/teses/disponiveis/59/59143/tde-22062020-195016/
• topic: Ativação neuronal; Computational model; Computational neuroscience; Estimulação magnética transcraniana; Modelo computacional; Neurociência computacional; Neuroestimulação; Neuronal activation; Neurostimulation; Transcranial magnetic stimulation

Transcranial magnetic stimulation (TMS) is a noninvasive technique of brain stimulation that has been widely used in both cognitive function studies and clinical applications. However, the biophysical mechanisms by which TMS activates cortical neurons and networks are still poorly understood. The present work aimed to create a computational model of the neuronal effects of single-pulse TMS combining compartmental models of neurons and a subject-specific electric field solution. The model consists of neurons of cortical layers L2/3 and L5, transformed to conform to cortical curvature and subjected to extracellular quasipotentials following a monophasic current waveform. First, excitation thresholds and sites of action potential initiation are determined through simulation of membrane dynamics with neurons being synaptically isolated, then epidural response is simulated by connecting them in a feedforward network. Excitation occured at morphological discontinuities such as axon terminals, and thresholds were mostly correlated with total electric field magnitude instead of the component normal to cortex. Coil orientations perpendicular to central sulcus presented lowest thresholds, with L5 neurons, in general, being more easily excitable than L2/3. The simulated epidural response of the network presented amplitude and duration in accord with experimental recordings, supporting the hypothesis of transsynaptic activation, with the time of propagation of action potentials in L2/3 axonal arbors suggesting a role in latency of I1-waves. By incorporating neuroanatomical factors to a neuronal network, the current model offers a computational framework for exploring TMS parameters and advancing the personalized use of neurostimulation.