1 National Acoustic Laboratories, Australia
2 Unit of Technical Audiology, Dept. of Ear and Skin, Karolinska
Institute, Stockholm, Sweden
The object of this presentation is to show a movie of a realistic
computer model of the ear and how the coding process and auditory channel
capacity depends upon maturation of the ear before birth and the reverse
process which takes place with aging. The model uses the one-dimensional
computationally-efficient wavefilter approach instigated by Fettweis,
which describes the motion of the basilar membrane, and the resulting
excitation of the auditory nerve. A key feature of the model is the
history dependence of biological processes receiving artificial
stimulation. Into this model is introduced normal and amplified sounds of
various types (sinusoids, clicks, speech) and scalar measures of sound
fidelity are derived from the neural activity. The behaviour of the model
is presented for the normal ear with intact crossed efferent innervation,
and then in various states of pathology modelled as reduced outer hair
cell activity and presented in terms of loss of channel capacity. These
are considered to the point of total loss of activity mimicking a moderate
loss, plus loss of inner hair cells mimicking a severe-to-profound hearing
loss. Finally the model depicts the loss of transmission due to a further
frequency-independent variable (<10 dB) attenuation which has been
associated with the swelling of primary afferent dendrites and its
dependence upon sound energy. The model thus demonstrates that the aging
ear depends increasingly upon temporal coding, and this process is
accentuated by high gain aids. Improved hearing aid performance,
particularly in regard to acclimatisation, may thus come from a tradeoff
between audiological and physiological considerations.