Statistical Modeling of Brain to Muscle Stimulus-Response, Utilizing Transcranial Magnetic Stimulation (TMS)

Steiner, Caitlin, Statistics - Graduate School of Arts and Sciences, University of Virginia
Keenan, Daniel, Department of Statistics, University of Virginia

Understanding how sensory information from the external world results in brain-initiated action, or lack of action, in humans is the fundamental goal of neuroscience. Inducing brain stimulation, analogous to those arising from the environment, has enabled researchers to start comprehending brain functioning. In the medical context, it is generally accepted that only noninvasive stimulation, such as Transcranial Magnetic Stimulation (TMS), of the brain will have broad applicability when trying to quantify the brain to muscle interface (corticospinal system) through the use of stimulus response relationships.

The key indicator of excitability of a muscle is the stimulus intensity versus motor-evoked potential (MEP) recruitment curve. Conventional analysis of the recruitment curve assumes a sigmoidal shape with constant additive Gaussian noise. However, two central problems arise: (1) it is extremely difficult to identify the threshold at which there will be a muscle response and, (2) there is an intensity dependent variability within the muscle response due to intrinsic, visceral, and extrinsic factors. Motivated by these limitations a new mathematical model that allows for intensity dependent variability about the MEP response and calculates the cortico-motor threshold (a fundamental neurological concept) within its parameterization has been developed. In a comparison, my new model performed better than the conventional (``gold standard") approach and still maintains the asymptotic properties of maximum likelihood estimates.

A third limitation emerges with the use of TMS, (3) the standard technique by which muscle-evoked potential measurements are made, while simple to calculate has no strong physiological basis. Additionally, my research has established the initial framework by which both mathematical and biological meaning of the MEP measurement could progress from. By accurately capturing the MEP response measurement and the relationship of the recruitment curve, neuroscientists can further understand states of the corticospinal system and subject specific parameters for TMS can be tested quickly and without unnecessary exposure to magnetic stimulation.

PHD (Doctor of Philosophy)
All rights reserved (no additional license for public reuse)
Issued Date: