Computational Modelling of Transcranial Magnetic Stimulation

Abstract

Transcranial magnetic stimulation (referred to as TMS) is a form of non-invasive brain stimulation that has been in widespread clinical use in recent decades. In spite of its widespread applicability to a variety of mental and neurological ailments, there are a great many unknowns in the understanding of how and why TMS has beneficial effects. In this work we model transcranial magnetic stimulation in-silico, with a focus on the propagation of action potentials and implications for synaptic plasticity. The first piece of modelling work that will be focused on in this thesis is the toolbox ”Neuron Modelling for Transcranial Magnetic Stimulation” (NeMo-TMS). NeMo-TMS is a tool developed to synergize modelling of transcranial magnetic stimulation across multiple scales of brain organization. NeMo TMS links modelling of magnetically-induced electric fields in large volumes of neural tissue (such as whole brains, segments of brain tissue, or even volumes placed in-vitro) to smaller scales. These fields are coupled to single-cell models. NeMo-TMS is capable of importing arbitrary morphologies which are reconstructed from observations of real cells. With the realistic electric field and the realistic morphology, the tool enables modelling of the propagation of action potentials through the axons and dendrites of these cells, and subsequent modelling of calcium accumulation and dispersion within the cells. This thesis details the function of this tool, and provides a worked example of the tool’s use and observations of the induction and propagation of cellular action potentials in NeMo-TMS. The second major piece of modelling in this thesis is the induction of synaptic plasticity by TMS stimulation. We implement in NeMo-TMS a well-validated model of the CA1 pyramidal cell with a reduced morphology and detailed biophysical properties. This model is then combined with a model of a combined AMPA/NMDA glutamatergic synapse, with multiple detailed phenomenological synaptic plasticity pathways representing biophysical pathways. With this model, we simulate induction of long-term potentiation by application of both low-frequency TMS stimulation as well as that induced by high-frequency theta burst TMS stimulus, and simulate the dependence of plasticity outcomes on stimulation frequency, synaptic location, and pharmacological perturbation. Finally, with the tools we developed, we investigate the degree to which axonal/dendritic morphology explains discrepancies in plasticity induction thresholds between cells sourced from different species of rodent, using detailed reconstructions of the dendrites and large, branching axons sourced from mice and rats. In summary, this thesis demonstrates the development and implementation of novel tools for modelling transcranial magnetic stimulation, and demonstrates their use with scientifically relevant applications to plasticity and modelling of action potential initiation.