Neuroprosthetic systems
rely on the communication of information to and from
the nervous system. This communication requires an
understanding of how neurons code information. For
example, to extract movement information from
mechanosensory neurons, one can model the relationship
between the mechanical stimulus and the neural
response. This relationship is often a complex
nonlinear function of multiple variables. Similarly,
to transmit information into the nervous system, one
must encode the input as a pattern of electrical
stimulation pulses that evokes an appropriate neural
response. A clinically successful example is the
cochlear implant which is a neuroprosthesis that
electrically stimulates the auditory nerve to produce
sensations of hearing. This device requires the
encoding of sound information into patterns of
electrical stimulation that the auditory system in the
brain can understand.
Research in the RNE Lab is currently investigating the
neural coding of motor and sensory information related to
the control of arm and leg movement. Using advanced digial
signal processing technologies, it is possible to record
and analyze the activities of many neurons simultaneously.
This approach allows us to study large parts of the neural
network rather than isolated single neurons, which was the
traditional approach in neurophysiology. By recording and
processing the neural activity in real time, we can
construct control signals for operating a prosthetic
device.
Figure 1: Multichannel Afferent
Recording
Figure 2: Multichannel DRG Recordings
