Biomedical Engineering Reference
In-Depth Information
One of these nuclei is the tracheosyringeal motor nucleus (nXIIts), which
innervates the syrinx. A second nucleus of this set is the retroambigualis
nucleus (RAm), which innervates both nXIIts and the spinal motor neu-
rons which control the expiratory muscles. In birds, both inspiration and
expiration are correlated with the activity of specific sets of muscles. The
parambigualis nucleus (PAm) has neurons with activity correlated with in-
spiration [Wild 2004]. How do these nuclei interact? On one hand, the nuclei
associated with inspiration and expiration (PAm and RAm) will inhibit each
other. On the other hand, there is feedback from the air sacs to these nuclei
via the vagus nerve (NX). Intrapulmonary CO
2
receptors are innervated by
vagal afferents, as well as sensors of air sac volume [Wild 2004]. The vagal
nerve then projects to the tractus solitarius nucleus, which then projects to
PAm. We describe this complex structure to point out that the song system
should not be thought of as a music box, where a neural script forces a passive
peripheral system. On the contrary, the brain-stem nuclei are modulated on
one hand by the forebrain nuclei, and on the other hand by peripheral inputs
that act in a feedback manner. Experience has shown that forced nonlinear
systems can display quite complex behavior. This should prepare us to expect
a rich class of phenomena and sound features which are not just the result
of a complex musical script written in forebrain nuclei, but are a result of an
interaction between the script and the instrument itself.
8.6 Computational Models and Learning
The genetic constraints on song are su
ciently loose in oscine songbirds,
hummingbirds and parrots that song acquisition requires an imitation process
[Nottebohm 2002b]. However, the processes involved in learning are not com-
pletely known. In this area, we are not referring to details: there is a wide
range of phenomena which probably play complementary roles in the process
of learning, beyond reinforcement or weakening of connections between the
neurons of the motor pathway [Nordeen and Nordeen 1997].
Maybe one of the most surprising processes involved in learning is neuro-
genesis, i.e., the birth of neurons. Biologists usually assume that this oc-
curs in the developing embryo or at a young age. One of the scientific
statements that biologists have conveyed to the general public is that neu-
rons in adults cannot be replaced if damaged. For this reason, the observa-
tion that Nottebohm and coworkers reported in the 1980s was revolution-
ary [Nottebohm et al. 1986, Nottebohm 2002a]: they described anatomical
changes in the brain which were correlated with developmental changes in
song. New neurons were being detected in the brains of adult canaries. But
what was really suggestive was that Nottebohm and Alvarez Buylla were
able to follow the geometrical paths within the brain followed by these new
neurons, and found that they migrated precisely to HVC, one of the nuclei
of the motor song system [Alvarez-Buylla and Nottebohm 1988]. A fraction
