We propose a series of coordinated computational and electrophysiological studies to identify, analyze, and implement contextual spike-timing codes in the encoding of light intensity, using a well-understood, simple visual system (the Hermissenda eye) that permits detailed cellular mechanistic analyses and biologically realistic, Hodgkin- Huxley level simulations. We further propose to examine mechanisms by which neural """"""""noise"""""""" paradoxically improves, rather than degrades, light intensity encoding; and the interactions of contextual spike-timing codes with learning-related synaptic plasticity. Implementation experiments to restore network function by explicitly introducing contextual spike-timing relationships may provide powerful, novel evidence regarding the functional consequences of these codes.
Specific Aim 1 : Role of Network Architecture. Pilot simulation and physiological data indicate that network interactions (feed-forward, feedback, and lateral inhibition) alter contextual spike-timing relationships and may contribute substantially to improved information processing in the network, relative to single cells. We will examine the existence, origins, and consequences of these codes in both the simulated and biological eye.
Specific Aim 2 : Encoding of Light Intensity. Preliminary evidence indicates that an increase in light intensity produces an increase in both the rate and synchrony of type A photoreceptor spiking, suggesting multiple encoding schemes with potentially complementary roles. Further, the addition of synaptic and ionic neural """"""""noise"""""""" improves both rate and synchrony encoding of light intensity. We propose to quantify and compare rate and contextual spike-timing codes across light intensities; examine how their performance is affected by the presence of neural noise; and implement relevant algorithms into a neural interface (or simulated interface) to determine whether they improve control of the system and light intensity encoding.
Specific Aim 3 : Contextual Spike-Timing Codes and Learning. Preliminary data indicate that synaptic facilitation at B-to-A connections alters the timing relationships between these photoreceptors, and that noise improves rather than degrades the reliability of type A cell responses to enhanced synaptic input. We propose to examine the role of learning-related synaptic plasticity in contextual spike timing encoding, including its effects on light intensity coding and mechanisms of noise-induced enhancements. Taken together, these studies will elucidate neurobiological strategies for information processing, and identify efficient bio-based solutions that may be advantageously incorporated into artificial intelligence systems, robotic systems, or clinical neuroprostheses.

Agency
National Institute of Health (NIH)
Institute
National Institute of Mental Health (NIMH)
Type
Research Project (R01)
Project #
1R01MH068392-01
Application #
6671314
Study Section
Integrative, Functional and Cognitive Neuroscience 8 (IFCN)
Program Officer
Glanzman, Dennis L
Project Start
2003-07-01
Project End
2007-05-31
Budget Start
2003-07-01
Budget End
2004-05-31
Support Year
1
Fiscal Year
2003
Total Cost
$229,705
Indirect Cost
Name
University of Utah
Department
Biomedical Engineering
Type
Schools of Engineering
DUNS #
009095365
City
Salt Lake City
State
UT
Country
United States
Zip Code
84112
Butson, Christopher R; Clark, Gregory A (2008) Random noise paradoxically improves light-intensity encoding in Hermissenda photoreceptor network. J Neurophysiol 99:146-54
Butson, Christopher R; Clark, Gregory A (2008) Mechanisms of noise-induced improvement in light-intensity encoding in Hermissenda photoreceptor network. J Neurophysiol 99:155-65
Nesse, William H; Clark, Gregory A; Bressloff, Paul C (2007) Spike patterning of a stochastic phase model neuron given periodic inhibition. Phys Rev E Stat Nonlin Soft Matter Phys 75:031912