Page 186 - Visions of the Future Chemistry and Life Science

Exploring human organs with computers 167

development of structurally complex biological systems. Rather than
attempting to replicate the existing organ of Corti, we could use finite-
element models to predict the degree of mechanical amplification that
could occur in regenerated hair-cell-based sensory epithelia, whose struc-
ture and properties are quite different from those of the normal organ of
Corti. Biological, carbon-based implementations of the simplified organ
could be constructed using genetic techniques, both by manipulating the
function of individual cells and controlling the way in which the develop-
ing cells form the structure. The development process itself is amenable to
finite-element analysis since it is driven mainly by local effects. The
replacement organ could be constructed from cells obtained from the even-
tual organ recipient, bypassing the problems associated with tissue rejec-
tion during transplantation. Conversely, silicon-based implementations of
the simplified model could be used in signal processing applications. For
example, a silicon cochlea could form the front-end of a speech recognition
system with a performance superior to any designed by an electrical
engineer.
It is highly likely that by the second decade of the new millennium
silicon-based computing will have reached fundamental technological or
physical limits. Computers will therefore be based on substrates that
exhibit superior performance characteristics. One possibility is the
photon. Optoelectronic devices, which use substrates such as gallium arse-
nide, permit the interconversion of electrons and photons. Hybrid comput-
ers, which may already be available commercially by 2010, would use
silicon for computation and photons for data transfer. The coherent mod-
ulation of very-high-frequency light beams enables many high-capacity

amount of force generation. But in the model, increased force generation leads to
less basilar membrane motion. This paradoxical observation is the first that is
consistent with the experimental observations that an increased amount of brain
stimulation causes a decrease in cochlear amplification. The model behaviour is
the direct result of the inflexion point at the outer edge of the outer pillar cell
becoming much more pronounced. (d) Simulating the response of the cochlea
when individual outer hair cells are stimulated in the absence of sound. There are
two motion peaks at the position of stimulation, one beneath the outer hair cells
and the other at the outer edge of the outer pillar cell. This model response is
consistent with experiments in which the cochlea is electrically stimulated, and
comparison with Figure 9.7(a) shows that the response to normal sound cannot be
predicted from these sorts of experiments.