Neural Coding and Perception of Learned Vocalizations

描述(由申请者提供)：语音交流对人类学习和社会互动非常重要。言语知觉障碍是听觉加工障碍和自闭症等神经系统疾病的特征。在apd患者和一些学习障碍儿童中，可以观察到诸如发声等频谱复杂声音的神经处理异常。这项研究的目的是了解听觉学习(知觉)和发声学习(产生)如何影响发声和其他声音的神经处理。学习识别和产生复杂交流发声的动物是研究学习如何改变发声神经编码的合适模型系统。鸣禽非常适合这项研究，因为：1)它们在听觉识别学习方面非常熟练；2)它们像人类一样，但不同于大多数其他动物，通过模仿声音来学习产生自己的发声；3)处理发声的听觉大脑区域是已知的。鸣禽也会对不同的发声产生不同的行为反应，这些发声的行为重要性或突显程度不同。在我们的模式物种斑马雀中，听觉中脑神经元对包括复杂发声在内的各种声音都能产生强大而可靠的反应。研究将检验三个主要假设。首先，听觉识别学习改变了中脑对歌曲和其他声音的编码(目标1)。我们将训练成年鸟类识别并对特定的歌曲做出反应，然后比较中脑神经元对可识别的歌曲和未经训练的歌曲的反应。对可识别和未经训练的歌曲的神经反应的预测差异是尖峰频率、神经辨别、频谱调谐和频谱时间调谐。其次，中脑神经元对发声的反应取决于发声的行为突显(目标2)。我们将训练成年鸟将积极或消极的行为突显与特定的歌声联系起来，然后比较中脑神经元对具有正向和负向价态的显著歌声以及没有经过训练的歌声的反应。第三，歌曲学习(通过模仿来学习制作歌曲)改变了中脑对歌曲的编码(目标3)。我们将在有和没有学习歌曲的条件下饲养幼鸟，并比较已经学会和没有学会产生歌曲的鸟类中脑神经元的听觉反应。拟议的实验将使用行为训练/测试、神经生理学、计算数据分析、解剖分析和发声学习中的操作来确定学习如何塑造神经元对交流发声的反应。了解听觉和发声学习对发声学习者声音编码的影响可能为听觉训练如何改变人脑中的语音编码提供候选的神经机制，并为语音习得如何塑造语音处理提供思路。

DESCRIPTION (provided by applicant): Vocal communication is important for human learning and social interaction. Speech perception impairments are features of neurological disorders such as auditory processing disorder (APD) and autism. In APD patients and some learning-disabled children, abnormalities in the neural processing of spectrotemporally complex sounds such as those in vocalizations are observed. The goal of this research is to understand how auditory learning (perception) and vocal learning (production) affect the neural processing of vocalizations and other sounds. Animals that learn to recognize and produce complex communication vocalizations are appropriate model systems in which to study how learning alters the neural coding of vocalizations. Songbirds are well suited for this research because: 1) they are highly skilled in auditory recognition learning; 2) they learn to produce their own vocalizations by vocal imitation, like humans but unlike most other animals; and 3) the auditory brain regions that process vocalizations are known. Songbirds also produce different behavioral responses to vocalizations that differ in behavioral importance, or salience. In our model species, the zebra finch, auditory midbrain neurons produce robust and reliable responses to a wide range of sounds, including complex vocalizations. Studies will test three main hypotheses. First, auditory recognition learning alters the midbrain encoding of songs and other sounds (Aim 1). We will train adult birds to recognize and respond to specific songs and then compare the responses of midbrain neurons to recognizable songs and untrained songs. Predicted differences in the neural responses to recognizable and untrained songs are spike rate, neural discrimination, spectral tuning and spectrotemporal tuning. Second, the responses of midbrain neurons to vocalizations depend on the behavioral salience of vocalizations (Aim 2). We will train adult birds to associate positive or negative behavioral salience with specific songs, and then compare the responses of midbrain neurons to salient songs with positive and negative valence, and songs without trained salience. Third, song learning (learning to produce songs by imitation) alters midbrain encoding of songs (Aim 3). We will raise juvenile birds under conditions in which song learning does and does not occur, and compare the auditory responses of midbrain neurons from birds that have and have not learned to produce song. The proposed experiments will use behavioral training/testing, neurophysiology, computational data analysis, anatomical analysis, and manipulations in vocal learning to identify how learning shapes neuronal responses to communication vocalizations. Understanding the effects of auditory and vocal learning on sound coding in vocal learners may provide candidate neural mechanisms for how auditory training alters speech coding in the human brain and ideas about how speech acquisition shapes speech processing.