Cortical and thalamic mechanisms of selective auditory attention

摘要 近年来，在有创和无创脑调控技术方面取得了许多进展， 运营然而，这些技术通常不能在“概念证明”之外有效地使用。 由于大脑中最基本功能的细胞网络起源尚不清楚，因此没有实验。的一部分 原因是，虽然认知神经科学家已经了解了很多关于支配 大脑操作和计算建模者在创建几乎每一个模型方面都取得了飞跃， 大脑回路，这两个领域仍然只有稀疏连接。我们提出的项目将弥补这一差距 认知神经科学、电生理学和计算建模之间的联系， 行为实验中多个空间尺度上的活动，并将这些数据连接到详细的 听觉丘脑皮质系统的计算模型。这一过程将提供具体的预测， 听觉系统功能的神经调节，并形成新的治疗方法的坚实基础。 我们的项目重点是定义三种不同的听觉机制的细胞网络基础， 感知过程，用于语音处理。首先是神经元的灵活性 三角洲-θ波段的振荡，赋予它们动态调整周期的能力， 自然听觉刺激序列的准节奏结构，包括物种特异性 发声和说话。第二种支持有效听觉处理的机制是振荡 相位重置，这使得能够通过神经元振荡支持精确跟踪刺激序列， 尤其是关注的听觉流的图形-背景分离。第三基本 处理连续听觉刺激流的机制是解析，这使得大脑能够 分割和分组声学元素，使它们形成可由大脑解释的单元。这三 这些机制形成了理解听觉环境所需的复杂计算的基础。 我们将在猕猴中同时进行丘脑-皮层电生理记录，以确定 神经元活动模式的时空组织支持上述基本听觉 处理机制。在行为任务中收集的数据将告知我们详细的丘脑皮层 计算模型，这反过来将提供准确的预测有效的神经调节方法， 诱导或暂时抑制不同听觉过程（例如流）背后的神经元活动模式 分离或解析。除了先进的时间分辨单个单元和神经元整体活动分析， 我们将能够验证基于行为偏差的神经调节的有效性。 我们提出的项目开发的基于模型的有针对性的神经调节技术将为 用于治疗神经精神和发育障碍的新治疗方法， 其特点是神经元振荡系统动力学特性的缺陷。

SUMMARY ABSTRACT There have been many recent developments in invasive and non-invasive techniques for modulating brain operations. However, these techniques typically cannot be efficiently used beyond “proof of concept” experiments since the cellular-network origins of the most basic functions in the brain are not known. Part of the reason for this is that while cognitive neuroscientists have learned a lot about the principles that govern brain operations, and computational modelers have made leaps and bounds in creating models of nearly every brain circuit, these two fields remain only sparsely connected. Our proposed project will bridge the gap between cognitive neuroscience, electrophysiology, and computational modeling by measuring neuronal activity on multiple spatial scales in behavioral experiments, and connecting these data to detailed computational models of the auditory thalamocortical system. This process will provide specific predictions for the neuromodulation of auditory system function and form a solid base for novel therapeutic approaches. Our project focuses on defining the cellular-network underpinnings of three distinct mechanisms of auditory perceptual processes, which are utilized for speech processing. The first is the flexibility of neuronal oscillations in the delta-theta bands that endows them with the capability to dynamically adapt their cycles to the quasi-rhythmic structure of naturalistic auditory stimulus sequences, including species specific vocalizations and speech. The second mechanism that supports efficient auditory processing is oscillatory phase reset, which enables the precise tracking of stimulus sequences by neuronal oscillations supporting, amongst other things the figure-ground segregation of attended auditory streams. The third fundamental mechanism for processing continuous auditory stimulus streams is parsing, which enables the brain to segment and group acoustic elements so that they form units that are interpretable by the brain. These three mechanisms form the basis of the complex computations needed to make sense of the auditory environment. We will perform concurrent thalamus-cortex electrophysiological recordings in macaques to determine the spatiotemporal organization of neuronal activity patterns supporting the above described fundamental auditory processing mechanisms. The data collected during behavioral tasks will inform our detailed thalamocortical computational model, which will in turn provide precise predictions on efficient neuromodulation approaches to induce, or temporarily inhibit the neuronal activity patterns underlying distinct auditory processes like stream segregation or parsing. Besides advanced time-resolved single unit and neuronal ensemble activity analyses, we will be able to verify the effectiveness of neuromodulation based on behavioral biases. The model based, targeted neuromodulation techniques developed by our proposed projects will pave the way for novel therapeutic approaches in the treatment of neuropsychiatric and developmental disorders that are hallmarked by deficits in the dynamical properties of neuronal oscillatory systems.