Neural mechanisms for auditory-motor integration and learning: prediction and reward

本研究计划将探讨听觉运动学习中预测与奖赏的神经系统。它基于音乐和言语的解剖学和生理学模型，这些模型提出，人类大脑中存在一条听觉背流，可以预测声音和动作之间的联系。目的1旨在分离背流结构在听觉-运动学习中的作用，并确定预测运动反应的产生和振荡特征。预测在音乐奖励或快乐中也很重要，这种奖励或快乐来自于期望和结果之间的紧张关系，并由多巴胺调节的纹状体网络介导。重要的是，纹状体奖励机制对学习也至关重要，但音乐愉悦和学习之间的联系尚未被探索。因此，目标2将考察可预测性和愉悦感对学习演奏旋律的影响。最后，音乐愉悦的模型提出纹状体奖赏网络与皮质系统相互作用。在目标3中，我们的目标是首次确定音乐凹槽背景下背流感觉运动预测网络和纹状体奖励网络之间的相互作用。我们在实验室中开发了一系列新的范例，并将它们与脑成像和神经刺激方法相结合，以解决以下假设：(1)将使用重复经颅磁刺激测试背侧和腹侧运动前皮层在获得听觉-运动关联中的作用。此外，我们假设预测的听觉-运动关联是在背流振荡网络中编码的。我们将通过将单脉冲经颅磁刺激与脑电图和脑磁图相结合来验证这一点。(2)我们已经证明，旋律的可预见性和愉悦性调节了听觉-运动学习。我们假设这种学习是由背流和奖励网络共同作用的结果，我们将使用功能磁共振成像进行测试。(3)我们将使用脑磁图和功能磁共振成像来验证背流和奖励网络之间的相互作用是音乐节奏的基础。这项工作在探索行动的预测控制和那些介导奖励的神经系统之间的相互作用以及检查潜在的大脑振荡机制方面具有创新性。结果将促进我们对背流对听觉-运动整合、预测和学习的贡献的理解。它们还将有助于完善连接听觉和运动信息的振荡机制模型，以预测行动控制。目的2和3最初是为了测试背流预测和学习机制以及纹状体奖励机制之间的相互作用。这将为内在奖励对运动学习的影响提供重要的见解，并有助于确定声音、动作和愉悦之间紧密联系的神经机制，这些联系是人类与音乐互动的特征。

This research program will probe the neural systems underlying prediction and reward in auditory-motor learning. It is based on anatomical and physiological models of music and speech which propose that there is an auditory dorsal stream in the human brain that predictively links sound and action. Objective 1 aims to dissociate the role of dorsal stream structures in auditory-motor learning, and to identify the generators and oscillatory signatures of predictive motor responses. Prediction is also critical in musical reward or pleasure which emerges from the tension between expectations and outcomes, and is mediated by dopamine-modulated striatal networks. Importantly, striatal reward mechanisms are also crucial for learning, but the link between musical pleasure and learning has not been explored. Thus Objective 2 will examine the impact of predictability and pleasure on learning to play a melody. Finally, models of musical pleasure propose that striatal reward networks interact with cortical systems. In Objective 3 we aim for the first time to identify interactions between dorsal-stream sensorimotor prediction networks and striatal reward networks in the context of musical groove - the pleasurable desire to move to music. We have developed a series of novel paradigms in our lab and will combine them with converging brain imaging and neuro-stimulation methods to address the following hypotheses: (1) The role dorsal- and ventral premotor cortex in acquiring auditory-motor associations will be tested using repetitive TMS. Further, we hypothesize that predictive auditory-motor associations are encoded in dorsal stream oscillatory networks. We will test this by combining single-pulse TMS with EEG and MEG. (2) We have shown that melodic predictability and pleasure modulate auditory-motor learning. We hypothesize that this learning results from the combined contributions of dorsal stream and reward networks and we will test this using fMRI. (3) We will use MEG and fMRI to test the hypothesis that interactions between the dorsal stream and reward networks underlie musical groove. This work is innovative in exploring the interaction between neural systems for predictive control of action and those mediating reward, and in examining the underlying brain oscillatory mechanisms. The results will advance our understanding of dorsal stream contributions to auditory-motor integration, prediction and learning. They will also serve to refine models of the oscillatory mechanisms that link auditory and motor information for predictive control of action. Objectives 2 and 3 are original in testing the interaction between dorsal stream prediction and learning mechanisms and striatal reward mechanisms. This will provide critical insights into the impact of intrinsic reward on motor learning, and will contribute to identifying the neural mechanisms underlying the strong link between sound, action and pleasure that characterize human interactions with music.