Dynamic Signal Processing Analyses of Neural Plasticity

描述（由申请人提供）：为了响应par02 -010，我们建议在统计学家（麻省总医院的Emery N. Brown博士，合作伙伴）、神经科学实验家（麻省理工学院的Matthew a . Wilson博士和纽约大学的Wendy Suzuki博士）和控制工程师（新南威尔士大学的Victor Solo博士）之间建立一个生物工程研究伙伴关系，以开发一种系统工程方法来理解神经可塑性。生物工程的研究领域将是神经信号处理算法的发展，通过结合点过程理论和自适应估计来研究学习和记忆形成过程中的神经可塑性。实验研究将研究大鼠、转基因小鼠和灵长类动物海马和邻近内侧颞叶结构（嗅内皮层、嗅周皮层和海马旁皮层）内的神经活动动态。这些实验研究将为有针对性的研究提供基础，设计新的神经信号处理方法，适用于同时记录的多个神经尖峰序列的动态分析。我们设计的算法将用于分析本调查中提出的实验研究中收集的数据。实验学家和定量科学家之间的密切合作将确保所设计的方法适合所收集的数据。该合作伙伴关系的长期目标是：建立一种结合实验-信号处理的方法来表征神经可塑性及其与学习、记忆形成和行为的关系；开发广泛适用的信号处理工具来分析神经系统的动态行为；建立一个跨学科的研究环境，使本科生、研究生和博士后能够在前沿实验方法和信号处理技术方面得到良好的训练，这些都是研究大脑中神经信息编码动力学等复杂问题所共同需要的。这项工作对健康的影响是对神经可塑性的更基本的理解，因此，它如何影响正常的生理过程，如生长、发育和学习，以及病理状况，如吸毒成瘾、阿尔茨海默病和慢性疼痛。更精确的神经信息动力学定量表征与改进的信号处理算法相结合，也可能导致创造机器-脑接口和设计神经假体装置的创新方法。

DESCRIPTION (provided by applicant): In response to PAR-02-010 we propose to form a Bioengineering Research Partnership between a statistician (Dr. Emery N. Brown of Massachusetts General Hospital, Partnership Director), neuroscience experimentalists (Dr. Matthew A. Wilson of the Massachusetts Institute of Technology and Dr. Wendy Suzuki of New York University) and a control engineer (Dr. Victor Solo of the University of New South Wales) to develop a systems engineering approach to understanding neural plasticity. The area of bioengineering research will be the development of neural signal processing algorithms by combining the theory of point processes and adaptive estimation to study neural plasticity during learning and memory formation. The experimental investigations will study the dynamics of neural activity within the hippocampus and adjacent medial temporal lobe structures (entorhinal, perirhinal and parahippocampal cortices) in rats, genetically altered mice, and primates. These experimental studies will provide the basis for a focused investigation that designs new methods for neural signal processing appropriate for dynamic analysis of multiple simultaneously recorded neural spike trains. The algorithms we design will be used to analyze the data collected in the experimental studies proposed in this investigation. The close collaboration between the experimentalists and the quantitative scientists will ensure that the methods designed are appropriate for the data collected. The long-term objectives of this partnership are: to establish a combined experimental-signal processing approach to characterizing neural plasticity and how it relates to learning, memory formation and behavior; to develop broadly applicable signal processing tools for analyzing the dynamic behavior of neural ensembles; and to establish an interdisciplinary research environment so that undergraduates, graduate students and post-doctoral fellows can be well trained in both the cutting edge experimental methods and signal processing techniques that are jointly required to study a complex question such as the dynamics of neural information encoding in the brain. The health implications of this work are a more fundamental understanding of neural plasticity and, as a consequence, how it affects normal physiological processes such as growth, development and learning as well as pathologic conditions such as drug addiction, Alzheimer's disease and chronic pain. More accurate quantitative characterizations of neural information dynamics coupled with improved signal processing algorithms may also lead to innovative approaches to creating machine-brain interfaces and designing neural prosthetic devices.