Faculty of Engineering, Built Environment and Information Technology
School of Engineering
Department of Electrical, Electronic and Computer Engineering
Selected Highlights from Research Findings
The aims of this study were to investigate the occurrence and effect of ephaptic excitation in electrical stimulation of the auditory system, to quantify the influence of ephaptic excitation on nerve stimulation, and to determine whether it is a necessary factor in neuromodelling. It was shown with a simple model that ephaptic excitation could be important at stimulus intensities close to threshold. The results show that the contribution of ephaptic excitation is significant up to at least 6 to 7 dB above threshold. Cochlear implant patients normally have a small dynamic range (average of 7 dB), indicating that the ephaptic effect might be important in models of the implanted cochlea. There is a significant qualitative difference between the results of the coincidental location and the separation of location of the active nerve fibres. When the fibres are bundled together, a constant stimulus current is used, while the current increases with an increase in spread of the active fibres when they are spatially separated. When the fibres are bundled together, there is no spatial constraint to the number of active neurons, which leads to large numbers of active neurons. This is also the reason why excitation by the electrode stimulus is at some stage outweighed by ephaptic excitation in these results, while the electrode stimulus dictates spread of excitation at high stimulus intensities (that is, large numbers of active neurons) when the active neurons are spatially separated. Results obtained for the planar orientation of the fibres could be used to estimate the extent of ephaptic excitation for the peripheral part of the fibres along the basilar membrane. In addition, ephaptic excitations were frequently observed to initiate in the central neural processes because of the larger fibre diameters, resulting in larger membrane currents per node and easier recruitment of thick fibres. This is of interest since cochlear implant users are often assumed to have an essential loss in the peripheral processes of their cochlear neurons as a consequence of long-term deafness. In conclusion, it was shown with a very simple model that ephaptic excitation could be important at stimulus intensities close to threshold. Limitations of this model include an assumed ideal volume conductor, a point source electrode and planar fibre location. Future research work in this field should focus on the development of a model that includes a more realistic description of the volume conductor (that is, the cochlea with the implanted electrodes). The results obtained from such a model should be compared with physiological results measured in animals or psychoacoustic results measured in human subjects, for example, forward masking data that provides an indication of the spread of neural excitation
Contact person: Prof T Hanekom.
|
