Populationskodierung natürlicher Stimuli

感觉神经科学的基本目标之一是准确地了解关于外部世界的信息是如何在大脑活动中表现出来的。破解这种“神经密码”对于理解大脑是如何做出决策和决定行为，以及制定临床策略以取代因创伤或疾病而丧失的大脑功能至关重要。尽管在过去的几十年里，我们对感觉功能的理解取得了巨大的进步，但自然环境中典型的复杂刺激的神经表征仍然是一个谜。我们理解中的这种差距可以归因于感觉系统的两个复杂性质，这两个性质在自然环境中很重要，但在涉及单个神经元对简单刺激反应的记录的典型实验室实验中没有得到解决：(1)感觉神经元的反应性质不是静态的，而是不断地适应以匹配刺激的当前统计性质；(2)复杂刺激的表示不限于单个神经元的反应，而是分布在高度相互关联的群体的反应中。我和我的研究小组计划在哺乳动物的听觉系统中研究这些复杂的特性，特别是在听觉中脑的下丘(IC)。我们将同时记录小群神经元在复杂自然刺激呈现时的反应，并将使用各种系统识别、信号处理和信息论技术来表征这些神经元的反应特性。我们还将利用这些实验结果来开发一个模型，该模型在预测对自然刺激的听觉反应时结合了适应和种群互连的影响。这项研究的结果将提供有关复杂声音处理的基本信息，并将为阐明真正的神经编码提供重要的一步。

One of the fundamental goals of sensory neuroscience is to understand exactly how information about the outside world is represented in brain activity. Cracking this ‘neural code’ is essential for understanding how the brain makes decisions and determines behavior, as well as for the development of clinical strategies for replacing brain function lost to trauma or disease. Despite tremendous advances in our understanding of sensory function over the past several decades, the neural representation of complex stimuli typical of the natural environment remains a mystery. This gap in our understanding can be attributed to two complex properties of sensory systems that are important in the natural environment, but are not addressed in typical laboratory experiments involving recordings of the responses of a single neuron to simple stimuli: (1) the response properties of sensory neurons are not static, but are constantly adapted to match the current statistical properties of the stimulus and (2) the representation of complex stimuli is not confined to the response of a single neuron, but is distributed across the responses of a highly interconnected population. With my research group, I plan to study these complex properties in the mammalian auditory system, specifically in the inferior colliculus (IC) of the auditory midbrain. We will record the responses of small populations of neurons simultaneously during the presentation of complex natural stimuli and we will characterize the response properties of these neurons using a variety of system identification, signal processing, and information theoretic techniques. We will also use these experimental results to develop a model that incorporates the effects of adaptation and population interconnectivity in predicting auditory responses to natural stimuli. The results of this investigation will provide fundamental information about the processing of complex sounds and will provide an important step toward elucidating the true neural code.