Abstract

The past two decades have witnessed the rapid growth of therapeutic brain-computer interfaces (BCI) targeting a diversity of brain dysfunctions. Among many neurosurgical procedures, deep brain stimulation (DBS) with neuromodulation technique has emerged as a fruitful treatment for neurodegenerative disorders such as epilepsy, Parkinson's disease, post-traumatic amnesia, and Alzheimer's disease, as well as neuropsychiatric disorders such as depression, obsessive-compulsive disorder, and schizophrenia. In parallel to the open-loop neuromodulation strategies for neuromotor disorders, recent investigations have demonstrated the superior performance of closed-loop neuromodulation systems for memory-relevant disorders due to the more sophisticated underlying brain circuitry during cognitive processes. Our efforts are focused on discovering unique neurophysiological patterns associated with episodic memories then applying control theoretical principles to achieve closed-loop neuromodulation of such memory-relevant oscillatory activity, especially, theta and gamma oscillations.

First, we use a unique dataset with intracranial electrodes inserted simultaneously into the hippocampus and seven cortical regions across 40 human subjects to test for the presence of a pattern that the phase of hippocampal theta oscillation modulates gamma oscillations in the cortex, termed cross-regional phase-amplitude coupling (xPAC), representing a key neurophysiological mechanism that promotes the temporal organization of interregional oscillatory activities, which has not previously been observed in human subjects. We then establish that the magnitude of xPAC predicts memory encoding success along with other properties of xPAC. We find that strong functional xPAC occurs principally between the hippocampus and other mesial temporal structures, namely entorhinal and parahippocampal cortices, and that xPAC is overall stronger for posterior hippocampal connections.

Next, we focus on hippocampal gamma power as a `biomarker' and use a novel dataset in which open-loop DBS was applied to the posterior cingulate cortex (PCC) during the encoding of episodic memories. We evaluate the feasibility of modulating hippocampal power by a precise control of stimulation via a linear quadratic integral (LQI) controller based on autoregressive with exogenous input (ARX) modeling for in-vivo use. In the simulation framework, we demonstrate proposed BCI system achieves effective control of hippocampal gamma power in 15 out of 17 human subjects and we show our DBS pattern is physiologically safe with realistic time scales.

Last, we further develop the PCC-applied binary-noise (BN) DBS paradigm targeting the neuromodulation of both hippocampal theta and gamma oscillatory power in 12 human subjects. We utilize a novel nonlinear autoregressive with exogenous input neural network (NARXNN) as the plant paired with a proportional–integral–derivative (PID) controller (NARXNN-PID) for delivering a precise stimulation pattern to achieve desired oscillatory power level. Compared to a benchmark consisted of a linear state-space model (LSSM) with a PID controller, we not only demonstrate that the superior performance of our NARXNN plant model but also show the greater capacity of NARXNN-PID architecture in controlling both hippocampal theta and gamma power. We outline further experimentation to test our BCI system and compare our findings to emerging closed-loop neuromodulation strategies.

Degree Date

Spring 5-14-2022

Document Type

Dissertation

Degree Name

Ph.D.

Department

Electrical and Computer Engineering

Advisor

Dr. Carlos Davila

Second Advisor

Dr. Bradley Lega

Third Advisor

Dr. Kevin Brenner

Fourth Advisor

Dr. Jung-Chih Chiao

Fifth Advisor

Dr. Dinesh Rajan

Subject Area

Bioengineering and Biomedical Engineering, Neuroscience

Number of Pages

165

Format

.pdf

Creative Commons License

Creative Commons Attribution-Noncommercial 4.0 License
This work is licensed under a Creative Commons Attribution-Noncommercial 4.0 License

Download

Share

COinS