Recent data from feline auditory nerve fibers (ANFs) indicate that electrically stimulated fibers can undergo large degrees of rate adaptation to pulse-train stimuli using pulse rates within the range used by clinical auditory prostheses. However, the application of Hodgkin-Huxley-type models does not produce such adaptation, which occurs over time periods on the order of 100 ms. We describe our development of a computational ANF axon model that incorporates a time-changing external potassium concentration ([K⁺]_ext) that depends on potassium currents produced by active nodal channel activity. This relatively simple and computationally tractable approach produces poststimulus time histograms that are similar to experimental (cat) data. Furthermore, this mechanism could be easily incorporated into other models to produce much more realistic estimates of the neural coding produced by repeated electric stimulation.

The Hodgkin-Huxley (HH) model does not simulate the significant changes in auditory nerve fiber (ANF) responses to sustained stimulation that are associated with neural adaptation. Given that the electric stimuli used by cochlear prostheses can result in adapted responses, a computational model incorporating an adaptation process is warranted if such models are to remain relevant and contribute to related research efforts. In this paper, we describe the development of a modified HH single-node model that includes potassium ion (K ⁺ ) concentration changes in response to each action potential. This activity-related change results in an altered resting potential, and hence, excitability. Our implementation of K ⁺ -related changes uses a phenomenological approach based upon K ⁺ accumulation and dissipation time constants. Modeled spike times were computed using repeated presentations of modeled pulse-train stimuli. Spike-rate adaptation was characterized by rate decrements and time constants and compared against ANF data from animal experiments. Responses to relatively low (250 pulse/s) and high rate (5000 pulse/s) trains were evaluated and the novel adaptation model results were compared against model results obtained without the adaptation mechanism. In addition to spike-rate changes, jitter and spike intervals were evaluated and found to change with the addition of modeled adaptation. These results provide one means of incorporating a heretofore neglected (although important) aspect of ANF responses to electric stimuli. Future studies could include evaluation of alternative versions of the adaptation model elements and broadening the model to simulate a complete axon, and eventually, a spatially realistic model of the electrically stimulated nerve within extracochlear tissues.

It is important and recommended to detect hearing loss as soon as possible. If it is found early, proper treatment may help improve hearing and reduce the negative consequences of hearing loss. In this study, we developed smartphone-based hearing screening methods that can ubiquitously test hearing. However, environmental noise generally results in the loss of ear sensitivity, which causes a hearing threshold shift (HTS). To overcome this limitation in the hearing screening location, we developed a correction algorithm to reduce the HTS effect. A built-in microphone and headphone were calibrated to provide the standard units of measure. The HTSs in the presence of either white or babble noise were systematically investigated to determine the mean HTS as a function of noise level. When the hearing screening application runs, the smartphone automatically measures the environmental noise and provides the HTS value to correct the hearing threshold. A comparison to pure tone audiometry shows that this hearing screening method in the presence of noise could closely estimate the hearing threshold. We expect that the proposed ubiquitous hearing test method could be used as a simple hearing screening tool and could alert the user if they suffer from hearing loss.