Hyperacuity in the Auditory System

听觉系统的超敏锐度

• 批准号: 0109872
• 负责人: Michael Reed
• 金额: $ 38.25万
• 依托单位: Duke University
• 依托单位国家: 美国
• 项目类别: Continuing Grant
• 财政年份: 2001
• 资助国家: 美国
• 起止时间: 2001-08-15 至 2007-07-31
• 项目状态: 已结题
• 来源: https://www.nsf.gov/awardsearch/showAward?AWD_ID=0109872&HistoricalAwards=false
• 关键词: Hyperacuity; Auditory; System

Reed0109872 Certain cells in the mammalian auditory brainstem (threeneurons from the periphery) show exceptionally small standarddeviations in the microsecond range of the timing of the firstaction potential response under repeated trials of the samesound. Moreover, the latency and its standard deviation arequite independent of frequency and decibel level over largeranges. This behavior is exceptional since the inputs to thesecells come from octopus and other cells in the contralateralcochlear nucleus, a nucleus which is ennervated by auditory nervefibers that have much larger standard deviations and highdependence on frequency and decibel level. Thus, this system isan example of how the brain can use relatively sloppy andvariable devices (neurons) to perform surprisingly accuratecalculations. Professor Michael Reed, with colleague physiologistJoseph Blum, investigates this system with three researchprojects: (i) They create and investigate by machine computationa large scale mathematical model of the whole system using knownphysiological properties of auditory nerve fibers and recentlydiscovered properties of octopus cells; (ii) They usemathematical analysis investigate of the improvement of thestandard deviation of latency in converging networks where thetarget cells require coincident inputs within time windows inorder to fire; (iii) They use partial differential equations tounderstand the special properties of dendritic informationprocessing in octopus cells. Using mathematics and machine computation, Professor MichaelReed continues to study how groups of cells (neurons) in theauditory brainstem perform calculations that individual cells cannot do by themselves. Neurons are inherently sloppy and variabledevices that perform differently even under (apparently) the sameconditions with the same inputs. Nevertheless, large groups ofthese neurons are able to perform together, reliably, highlyaccurate calculations. This project tries to understand thisapparent paradox. The project sheds light on how and why humanbrains are organized the way they are and it suggests newmechanisms for the development of man-made thinking devices.

Reed0109872 哺乳动物听觉脑干中的某些细胞（来自外围的三个神经元）在重复试验相同声音的情况下，第一动作电位反应的时间在微秒范围内表现出异常小的标准偏差。 此外，延迟及其标准偏差在大范围内与频率和分贝水平完全无关。 这种行为是特殊的，因为这些细胞的输入来自章鱼和对侧耳蜗核中的其他细胞，耳蜗核受听觉神经纤维的支配，听觉神经纤维具有更大的标准偏差，并且对频率和分贝水平具有高度依赖性。 因此，这个系统是大脑如何使用相对马虎和可变的设备（神经元）来执行令人惊讶的精确计算的一个例子。 迈克尔·里德教授和同事生理学家约瑟夫·布鲁姆通过三个研究项目研究了这个系统：（i）他们利用已知的听觉神经纤维的生理特性和最近发现的章鱼细胞的特性，通过机器计算创建和研究整个系统的大规模数学模型； (ii) 他们使用数学分析来研究收敛网络中延迟标准差的改进，其中目标细胞需要在时间窗口内一致输入才能触发； (iii) 他们使用偏微分方程来理解章鱼细胞树突信息处理的特殊性质。 迈克尔·里德教授利用数学和机器计算继续研究听觉脑干中的细胞群（神经元）如何执行单个细胞本身无法完成的计算。 神经元本质上是草率且可变的设备，即使在（显然）具有相同输入的相同条件下，其表现也不同。 然而，大量的这些神经元能够一起执行可靠、高度准确的计算。 这个项目试图理解这个明显的悖论。 该项目揭示了人脑的组织方式和原因，并提出了开发人造思维设备的新机制。