School of Psychological Sciences, University of Manchester, Manchester, United Kingdom.
Neuropsychologia. 2012 Jan;50(1):77-84. doi: 10.1016/j.neuropsychologia.2011.10.024. Epub 2011 Nov 4.
We investigated how two distortions of the speech signal - added background noise and speech in an unfamiliar accent - affect comprehension of speech using functional Magnetic Resonance Imaging (fMRI). Listeners performed a speeded sentence verification task for speech in quiet in Standard Dutch, in Standard Dutch with added background noise and for speech in an unfamiliar accent of Dutch. The behavioural results showed slower responses for both types of distortion compared to clear speech, and no difference between the two distortions. The neuroimaging results showed that, compared to clear speech, processing noise resulted in more activity bilaterally in Inferior Frontal Gyrus, Frontal Operculum, while processing accented speech recruited an area in left Superior Temporal Gyrus/Sulcus. It is concluded that the neural bases for processing different distortions of the speech signal dissociate. It is suggested that current models of the cortical organisation of speech are updated to specifically associate bilateral inferior frontal areas with processing external distortions (e.g., background noise) and left temporal areas with speaker-related distortions (e.g., accents).
我们利用功能磁共振成像(fMRI)研究了语音信号的两种扭曲——添加背景噪声和不熟悉的口音——如何影响语音理解。听众在安静的标准荷兰语、添加背景噪声的标准荷兰语和不熟悉的荷兰口音的语音中执行快速句子验证任务。行为结果表明,与清晰语音相比,两种失真类型的反应都较慢,而两种失真类型之间没有差异。神经影像学结果表明,与清晰语音相比,处理噪声导致双侧下额前回和额前回的活动增加,而处理口音语音则募集了左侧颞上回/回的一个区域。研究得出的结论是,处理语音信号不同失真的神经基础是分离的。有人建议更新当前的皮质言语组织模型,以专门将双侧下额区域与处理外部失真(例如背景噪声)相关联,并将左侧颞区与与说话者相关的失真(例如口音)相关联。