Computational and Experimental Advances in Clinically Relevant Neurostimulation of Non-Human Primates

Abstract

Neural engineering capabilities in non-human primates (NHPs) dramatically lag the capabilities in rodents, despite NHPs being the most neurophysiologically similar animal model to humans. This work puts forward advances in hardware, computation, and theoretical grounding to improve neural engineering ability in NHPs, especially with respect to clinically relevant neural stimulation. We begin by investigating how cortical stimulation affects cortical coherence, a form of functional connectivity, in the NHP cortex (Chapter 2). Our findings highlight how various factors, such as the state of the cortex during stimulation, can modulate coherence. Building on this, Chapter 3 examines how network structure impacts functional reorganization resulting from optogenetic stimulation in the NHP sensorimotor cortex. The results, obtained through graph theory-inspired computational analysis, reveal that network structure, more than the stimulation protocol, can predict changes in functional connectivity. To further our understanding of behavioral and neural disruption studies in NHPs, we introduce a novel opto-electric neural interface in Chapter 4. This tool combines the specific modulation abilities of optogenetics with the dual functionality of electrical stimulation and recording. Consequently, this allows for a more precise investigation into the link between brain activity and behavior. Transitioning to computer-based work in Chapter 5, we develop a biophysically accurate neural network simulation to derive structural connectivity from functional connectivity at the local field potentials (LFP) level. This important advancement aids in understanding the relationship between LFP functional connectivity and structural connectivity. Finally, in Chapter 6, we present and validate a computational model predicting lesion sizes in a new photothrombotic stroke model. This key tool contributes to our understanding of cortical physiology and stroke in NHPs. In summary, this thesis provides a multifaceted approach to enhance our understanding of neurostimulation and its potential therapeutic benefits for neural disorders. By improving experimental and computational tools, grounding functional connectivity in a biological framework, and advancing stroke modeling in NHPs, this work helps to bring neurotechnologies closer to their clinical application in humans.