Accessibility and manipulation of brain signals for neuroprosthetic applications.

摘要

The field of Neural Engineering spawned in response to the perpetual problem of neurology: injured central nervous system neurons do not regenerate or repair, and stem cell/molecular genetic solutions, while the ideal intervention, are far away from clinical utilization.;A current solution is to substitute computers and electrodes for neurons, as either transducers (cochlear implants, retinal implants, and visual cortex ECoG implants for sensory replacement), regulators (deep brain stimulations for Parkinson's disease), and output signal readers (motor cortex neuroprosthetics).;I have focused on improving the technology of motor neuroprosthetics, and in this dissertation I investigated three sub-systems of this relatively new technology in a rat model.;In my first experiment, I demonstrated that the cingulate cortex, part of the prefrontal cortex, can be used as an additional control signal for a motor neuroprosthetic device in the event that upper motor neurons of the motor cortex are degenerated by neurodegenerative diseases.;In my second experiment, I examined whether electrocorticograms (ECoGs) and local field potentials (LFPs) are independent from the spiking activity of motor neurons and could be thus used as additional control channels. I showed these signals are not necessarily independent, specifically, the spikes phase-lock to the field potentials at defined frequencies, and careful algorithms will have to be developed to combine spikes, LFPs, and ECoGs as different control channels for a neuroprosthetic device.;In my third experiment, I investigated the use of feedback in a neuroprosthetic model. I combined intracortical microstimulation (ICMS) of the visual cortex with simultaneous motor cortex ensemble recordings in real time to demonstrate the feasibility of a closed-loop neuroprosthesis. I showed that though sensory cortex ICMS can be combined with motor cortex recording in real-time in a viable preparation, increased technological development in simultaneous decoding with brain stimulation needs to occur before feasible clinical implementation can become a reality.;By reading and manipulating brain signals via microelectrodes, a basic level of neural control and neural replacement can be achieved. Until the day that physicians have access to technology that allows spinal cords to regrow and limbs to regenerate, current technology allows us to achieve some measure of success by replacing neurons with electrodes.