Yeajin Cho

Neural circuits have explained motor learning observed in animal populations, and plastic synaptic reorganization, but simulation of network modeling at larger scales lacks circuit organization. As a result, while large scale models capture general population behavior, there remains no canonical circuit, or simplest circuit-based functional subunit, that can be studied to determine the most discrete level of neuromodulation effect. The purpose of this study was to integrate a canonical neural circuit of the motor cortex with mathematical representation of membrane potentials across multiple neurons. The circuit model was evaluated by comparing changes in modeled resting outputter cell spike rate to changes in empirically recorded motor evoked potential amplitudes following intermittent theta burst stimulation (iTBS) delivered via transcranial magnetic stimulation. Effect sizes measured the magnitude of effect in both modeled and experimental data. The model successfully showed an increase in resting spike rate with relation to increased AMPA receptor permeability, scaled by simulated effects of iTBS to induce long-term potentiation (LTP), in an excitatory change similar to that seen in empirically recorded motor evoked potentials from human subjects. Thus, the model captured the mechanism of LTP in its simplest form. The representation is incomplete, however, as the model overestimated the responsivity based on the increased effect size relative to empirical data and noise. The overestimation is likely due to the lack of the disinhibition mechanism being fully represented in this iteration of the circuit, the simplicity of the cellular compartments modeled relative to the physiological complexity of neurons, and model parameter approximations of values controlling AMPA receptors. The circuit model is a building block for wider and more complex cortical networks that can be used to test mechanisms and applications of neuromodulation modalities.

Transcranial magnetic stimulation (TMS) is a treatment procedure for some neuropsychiatric disorders, and it has been used for brain mapping, as well as diagnosis and treatment of neuromuscular dysfunctions. There is a disconnect between TMS modeling and clinical data: several groups have reported the simulated induced electric field and measured resting motor threshold (RMT) with inconsistent results in the relationship between RMT and brain scalp distance. This necessitates the use of simulation parameters that further account for individual differences in neuroanatomy. We recruited ten healthy subjects and obtained empirical RMT, magnetic resonance imaging (MRI), and diffusion tensor imaging (DTI). We developed anatomically accurate brain models from MRI and simulated TMS to determine the percent depolarized volume of gray matter (DVG) from TMS-induced electric fields. Corticospinal fiber tracts were extracted from the primary motor cortex from DTI to obtain fiber tract surface areas (FTSAs) for each participant. Linear mixed-effects models were used to evaluate the effect of DVG and FTSA on RMT. We report that DVG correlates with RMT when accounting for corticospinal FTSA.