Contextual Spike-Timing Codes in Sensory Processing

感觉处理中的上下文尖峰计时代码

• 批准号: 6671314
• 负责人: Gregory Arthur Clark
• 金额: $ 22.97万
• 依托单位: UNIVERSITY OF UTAH
• 依托单位国家: 美国
• 财政年份: 2003
• 资助国家: 美国
• 起止时间: 2003-07-01 至 2007-05-31
• 项目状态: 已结题
• 来源: https://reporter.nih.gov/project-details/6671314
• 关键词: Hermissenda; behavioral /social science research tag; computational neuroscience; computer simulation; learning; light intensity; neural information processing; neural transmission; sensory discrimination; sensory signal detection; visual feedback; visual perception; visual photoreceptor

DESCRIPTION (provided by applicant): 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.

描述(由申请人提供)：我们提出了一系列协调的计算和电生理学研究，以识别、分析和实现光强度编码中的上下文尖峰计时码，使用易于理解的简单视觉系统(Hermissenda Eye)，该系统允许详细的细胞机制分析和生物逼真的霍奇金-赫胥黎水平模拟。我们进一步建议研究神经“噪声”矛盾地改善而不是降低光强编码的机制，以及上下文尖峰计时代码与学习相关突触可塑性的相互作用。通过明确引入上下文尖峰-定时关系来恢复网络功能的实现实验可以提供关于这些代码的功能后果的强有力的、新颖的证据。具体目标1：网络架构的作用。试点模拟和生理数据表明，网络相互作用(前馈、反馈和侧向抑制)改变了上下文尖峰-时序关系，并可能实质上有助于改善网络中相对于单个细胞的信息处理。我们将从模拟和生物学的角度研究这些密码的存在、起源和后果。具体目标2：对光强度进行编码。初步证据表明，光强度的增加会增加A型光感受器尖峰的频率和同步性，这表明多种编码方案具有潜在的互补作用。此外，突触和离子神经“噪声”的加入改善了光强度的速率和同步编码。我们建议量化和比较不同光强的速率和上下文尖峰计时码；检查它们的性能如何受到神经噪声的存在的影响；并将相关算法实现到神经接口(或模拟接口)中，以确定它们是否改善了系统和光强编码的控制。具体目标3：语境中的尖峰时序编码和学习。初步数据表明，B-A连接的突触易化改变了这些光感受器之间的时序关系，而噪声提高了A型细胞对增强突触输入的反应的可靠性，而不是降低。我们建议研究与学习相关的突触可塑性在上下文棘波计时编码中的作用，包括它对光强度编码的影响和噪声诱导的增强机制。综上所述，这些研究将阐明信息处理的神经生物学策略，并确定有效的基于生物的解决方案，这些解决方案可能有利地整合到人工智能系统、机器人系统或临床神经假体中。