Purdue University Graduate School
Browse
Dissertation_BrandonSCoventry.pdf (19.61 MB)

Closed-loop optical neuromodulation of the auditory thalamocortical pathway

Download (19.61 MB)
thesis
posted on 2021-06-24, 05:07 authored by Brandon Coventry

Neurological and sensory neuroprostheses have proven effective in treating ailments such as Parkinson’s disease and Tourette’s syndrome using deep brain stimulation (DBS), as well as restoring auditory and visual percepts using cochlear and retinal implants. These devices rely on open-loop, continuous electrical stimulation. Deficits in such devices, such as large-scale, nonspecific activation due to current spillover and the inability to selectively target local micro-circuits, result in aberrant auditory and visual percepts in sensory prostheses and intractable side effects in central nervous system stimulation implants. It is therefore critical to develop targeted, closed-loop neuroprostheses which can adjust treatment in real time. Infrared neural stimulation (INS) is an optical technique which has shown to selectively stimulate nerves and neurons using long wavelength (> 1850 nm) infrared light. Previous studies have suggested that INS activation is spatially constrained, and thus could provide targeted stimulation without the need for genetic modification inherent to other optical techniques. However, the photo-biomodulation mechanisms underlying INS, the therapeutic parameters, and underlying neurophysiological mechanisms are unknown. In this work, we explore the use of INS as a thalamocortical neuroprosthesis by showing that INS reliably produces graded responses across a continuum of wavelengths and drives naturalistic responses across thalamocortical circuits. Furthermore, we develop spiker-net, the first closed-loop optical neurostimulator which uses deep reinforcement learning techniques to actively track and learn evolving brain state patterns. Although spiker-net has been tested in conjunction with INS stimulation, it is a general closed-loop scheme that can be used with any stimulation technique to control output and neural ensemble activities. This work takes a first step in evaluating INS as an optical deep brain stimulation modality which could provide more precise and targeted treatment improving patient outcomes.


History

Degree Type

  • Doctor of Philosophy

Department

  • Biomedical Engineering

Campus location

  • West Lafayette

Advisor/Supervisor/Committee Chair

Edward L Bartlett

Additional Committee Member 2

Pedro P. Irazoqui

Additional Committee Member 3

Tamara Kinzer-Ursem

Additional Committee Member 4

Thomas M Talavage