Zerebrale Kontrolle motorischer Aspekte der Sprachproduktion

认知-语言操作和言语运动输出之间的接口的神经解剖学基础仍然知之甚少。这一挑战为我们研究小组的一系列功能磁共振成像（fMRI）研究提供了基础。首先，左侧运动前皮层和对侧小脑半球的相互作用为高度过度学习词串的内部言语提供支持。其次，小脑在这个过程中的参与似乎与高于3hz的音节序列有关。考虑到最近的语言产生的心理语言学模型，以及之前我们小组对共济失调性构音障碍的语言时间的研究，我们假设这个观察到的网络反映了小脑在说话者的习惯性语言速度为4到6赫兹时，从左半球的前运动存储中检索到的心理音节的连接。第三，发现前脑岛的激活与显性言语有关。该项目的重点是对该模型的心理音节成分及其与小脑和前脑岛的相互作用进行实验评估。IMS斯图加特生成的数据语料库将为正常语音的计划语音调查、慕尼黑-博根豪森城市医院ekn -临床神经心理学研究小组的临床调查以及<s:1>宾根大学正常受试者的功能磁共振成像研究提供测试材料。最后，IMS将开发和实现一个计算模型，该模型将集中于语音的时间/韵律组织及其神经控制，并明确指定语音编码期间的检索过程，包括从基于范例的音节表中检索。

The neuroanatomical substrates of the interface between cognitive-linguistic operations and speech motor output are still poorly understood. This challenge provided the basis for a series of functional magnetic resonance imaging (fMRI) studies of our research groups. First, interactions of left-sided premotor cortex and contralateral cerebellar hemisphere were found to subserve inner speech of highly overlearned word strings. Second, participation of the cerebellum in this process seems to be bound to syllable sequences above a rate of 3 Hz. Considering recent psycholinguistic models of speech production as well as preceding investigations of our groups on speech timing in ataxic dysarthria, we assume this observed network to reflect cerebellar concatenation of retrieved mental syllables from a left-hemisphere premotor store at a speaker's habitual speech tempo of 4 to 6 Hz. Thirdly, activation of the anterior insula was found associated with overt speech. The proposed project focuses on an experimental evaluation of the mentalsyllabary component of this model, and its interactions with cerebellum and anterior insula. The data corpus generated at IMS Stuttgart will provide the test materials both for the planned phonetic investigations of speech by normals, the clinical investigations at the EKN-Clinical Neuropsychology Research Group, City Hospital Munich-Bogenhausen, as well as the fMRI studies in normal subjects at the University of Tübingen. Finally, the IMS will develop and implement a computational model which will concentrate on temporal/prosodic organization of speech and its neural control and explicitly specify the retrieval process during phonetic encoding, including retrieval from the exemplar-based syllabary.