Informing Neuromodulation Therapies with a Control-Theory Approach to Brain Network Plasticity

Andrew Murphy, University of Pennsylvania

Abstract

Unlike the pharmacologic treatment of neuropsychiatric disorders, non-invasive neuromodulation is spatially targeted and therefore results in far fewer side effects. However, the clinical effect of neuromodulatory treatments is often unpredictable: some patients improve greatly while others receive little or no benefits. This unpredictability may in part be due to differences in both functional and structural brain connections between individuals, which may alter how neuromodulatory stimuli diffuse through the brain. Critically, clinical effects are a product of this diffusion. To understand the behavioral effects of neuromodulatory stimuli, one must ask (1) how stimulation of a particular region alters the patient’s brain network, and (2) how the patient’s white matter tracts constrain how stimuli diffuse through and alter the brain network. To address these questions, we used functional MRI imaging in conjunction with transcranial magnetic stimulation (TMS) to probe brain network responses to neuromodulation. With regard to the first question, we focused narrowly on the network connection between the default mode system (DMS) and the frontoparietal system (FPS) during the performance of a working memory task. We found that the strength of this connection is governed by activity in two separate and novel regions within the FPS: increased activity in one region strengthens this connection, while increased activity in the other region weakens it. While the first study did not involve TMS, it demonstrates that network changes may be predictably linked to activity in particular brain regions. Therefore, in the second study we used a frontal cortex TMS dataset to probe the effects of white matter architecture on resulting brain network changes. We build a predictive model of TMS-induced network changes and found that our model was more accurate when it incorporated white matter architecture versus when it did not. These results indicate that considering the white matter architecture of a given patient may improve the precision of TMS targeting, and therefore possibly improve clinical outcomes. Taken together, these results (1) suggest that brain interconnections are specifically related to regional activations, (2) partially illuminate this complex relationship, and (3) indicate this relationship may be an essential and understudied component of TMS effects.

Subject Area

Neurosciences|Medical imaging

Recommended Citation

Murphy, Andrew, "Informing Neuromodulation Therapies with a Control-Theory Approach to Brain Network Plasticity" (2019). Dissertations available from ProQuest. AAI22624822.
https://repository.upenn.edu/dissertations/AAI22624822

Share

COinS