We seek to implement an aggressive neural engineering research program that will combine both basic science and clinical components. Pursuing successful solutions to the associated research challenges offers a rich opportunity for collaborative, interdisciplinary endeavors. The research program is structured into 4 main thrusts summarized below:

Thrust 1: Objective fitting of cochlear implants

The increase in the number of young children implanted with cochlear implants (CIs) has spurred the need for an objective aid to assist with the fitting procedure. The electrical stapedius reflex (ESR) threshold has strong correlation with the upper end of the dynamic range of the CI but the traditional non-invasive technique of measuring acoustic impedance has limitations. Dr. Clement's PhD research developed and demonstrated the use of a new strategy to record the ESR using electromyogram (EMG) recordings directly from the stapedius muscle. This work successfully characterized the ESR response in a number of animal models, including some awake and unrestrained examples showing that the technique could be clinically useful. In order to move toward a clinical application, we plan to conduct additional studies aimed at refining the basic technique and obtaining additional chronic data. This work will hopefully include designing an optimized BioMEMS based chronic stapedial electrode that could be manufactured on a large scale. We are developing collaboration with persons in audiology and otolaryngology, especially at Hershey Medical Center, to perform intra-operative recordings in humans undergoing cochlear implantation as a first step toward clinical trials. Some of the critical questions to be asked with this research include: Does a chronic indwelling electrode cause significant damage to the muscle? How stable are the ESR responses over time as a result of chronic stimulation as compared to behavioral threshold? What are the best recording/processing techniques for obtaining the EMG response with the telemetry systems currently available in today's cochlear implants? What types of strategies are effective for measuring the EMG during higher stimulation rates given the presence of stimulation artifact? Improving the accuracy of fitting with objective aids could allow patients to receive greater benefit from their cochlear implant.

Thrust 2: Investigations of neural ensemble coding with multi-channel neural recordings

Until recently, much of the literature investigating the neural coding of acoustic environments has been conducted with single electrode penetrations in anesthetized animals probed with simple stimuli (clicks, tones, frequency modulated sweeps, etc). While these studies have been critical for characterizing the properties of auditory neurons, recent studies have suggested that the auditory cortex, similar to the visual cortex, may be specifically designed to detect spectrotemporal features (such as those present in speech and vocalizations). It has also been hypothesized that cortical neurons dynamically participate in groups or ensembles that multiplex, process, and bind together the key information representing the complex and changing auditory landscape. Leveraging our experience in chronically implanting animals with cortical microelectrode arrays, we are pursuing studies that will provide a better understanding of how the brain processes auditory stimuli on a spatio-temporal scale. Interesting research questions to be addressed include: How many recording electrodes are required to reconstruct or estimate a presented stimulus? Can behavioral performance in discrimination tasks be accurately predicted by examining the associated similarities in the neural representations? How and to what degree do these neural representations change with learning? Building on our experience with cochlear implants and electrical stimulation, we plan to investigate cortical processing of electrical stimuli using electrode arrays that span the auditory cortex. We are very interested in understanding the fundamental differences in cortical responses to electrical and acoustic stimuli. The knowledge gained may be applied to the development of new stimulation protocols that might better recreate natural cortical responses. This may include the development of models for investigating alternative sites of stimulation such as auditory nerve and/or brainstem penetrations that might provide an increased number of independent stimulation channels relative to the traditional cochlear implant electrode. New neural interface designs and stimulation strategies could also be developed and evaluated through this research.

Thrust 3: The development of BioMEMS based neural interfacing devices.

Central to the above thrust is the necessity of a neural interface. The neural interfacing technology developed thus far has provided an exciting view into the operation of neural circuits. However, if this technology is ever to have clinical impact, devices should bio-integrate and yield stable communication channels with ever increasing numbers of neurons for significant amounts of time. Flexible polymer-based BioMEMS have the potential of better matching the mechanical properties of brain tissue while at the same time surfaces can be engineered to enhance biocompatibility. We are collaborating with experts in thin film processing and microelectronics to create such optimal and reliable neural interfacing device structures that can be readily produced. We are also interested in collaborating with experts in biomaterials and tissue engineering to investigate potential surface modifications or neural culturing techniques that might further enhance long-term biocompatibility. Finally, while flexible structures are desirable to reduce potential micro-damage in chronically implanted devices they are more difficult to implant. The other research thrusts provide a vast "test-bed" for the development of this technology and will in turn receive great benefit from it.

Thrust 4: Bio-robotics and brain-machine interface development

Bio-robotics and direct brain-machine interface research is very new and exciting. Recently researchers have demonstrated the ability to integrate a simple living neural system with a robot to create a functioning bio-robotic system that can seek out or avoid a target stimulus. Several others have shown that animals can learn to control neuroprosthetics through neural signals recorded directly from the brain. However, before developing and testing an invasive system that can be readily used by those suffering from paralysis or other neurological deficits significantly more research must be conducted. We are interested in contributing to this research effort by developing novel brain-machine interface experiments in rats that learn to control external devices with their cortical signals. Through closed-loop, on-line experiments I am interested in learning to what extent it is possible for animals to take advantage of neural plasticity to interact with the external world through these new information channels. The research will utilize microelectrode arrays for signal input/output. We are particularly interested in combining both motor and sensory channels to evaluate the effect of sensory feedback. The research effort will address questions such as: What neural areas are the best targets for obtaining relevant command signals? What are optimal strategies for extracting and processing the relevant signals? How does the brain and associated neural classification algorithms adapt together? What is the maximum information transfer rate achievable by invasive brain-machine interfaces? These efforts may lead not only to clinically viable technologies, but may also usher in a new generation of robotics that have the processing power and adaptive properties of living systems by employing neural tissue as an adaptive micro-controller.