In this project, new mathematical analytical and computational methods for addressing problems of information processing in the auditory system are developed. More specifically, the investigators focus on several key pieces of the pathways responsible for encoding of the interaural time differences (ITDs). ITD encoding is thought to be particularly important for localization of low-frequency sound sources or low-frequency-modulated sound sources. First, the role of active dendritic conductances in auditory coincidence detectors is explored. Second, the researchers investigate and classify the ranges of parameters, cellular types, and types of interaction that improve or decrease the precision and reliability of the timing information. Third, a systematic study of a periodically driven neuronal model with additive white or colored noise as a stochastic dynamical system is conducted. Finally, a computational model of the early auditory pathway with natural or electrical (cochlear implant) stimulation is built. This model is used to study responses of cells to the interaural time differences in the amplitude modulated signal. All four components of the project contribute to a longer-term goal to build a comprehensive model for designing and testing new protocols for binaural electrical stimulations. The research activities result in the creation of new computational tools for studying the auditory system, as well as other neuroscience applications. This research also requires the development of new analytical techniques that further the mathematical research, and can be used by computational neuroscientists, probabilists, and applied mathematicians.
Cochlear implants are prosthetic devices that allow restoration of hearing for people with severely impaired hearing. In normal hearing air vibrations produced by auditory stimuli are transformed by the auditory receptors into electrical signals, which travel along the auditory nerve to the brain. In a patient with a cochlear implant this process is replaced by the direct electrical stimulation of the auditory nerve. Hence, the success of the procedure crucially depends on the algorithms of converting sounds to electrical signals. To look at it differently, the success depends on our understanding of the neural coding of the normally functioning early auditory processing. Of special interest, recently, is the binaural processing (using information from both ears) as binaural cochlear implantation is becoming available. In this project, several key pieces of the pathways involved in a binaural task of sound source localization are the primary focus. Using a combination of mathematical modeling, computer simulations and theoretical analysis, the role of detailed geometrical structure of the cells, types of interaction, specific membrane channels, and tolerance to imperfect (noisy) inputs is investigated. These are components of a longer-term goal to build a comprehensive model for designing and testing new protocols for binaural electrical stimulations.
