Ekert Justyna O, Gajardo-Vidal Andrea, Lorca-Puls Diego L, Hope Thomas M H, Dick Fred, Crinion Jennifer T, Green David W, Price Cathy J
Wellcome Centre for Human Neuroimaging, UCL Queen Square Institute of Neurology, 12 Queen Square, London WC1N 3AR, United Kingdom.
Wellcome Centre for Human Neuroimaging, UCL Queen Square Institute of Neurology, 12 Queen Square, London WC1N 3AR, United Kingdom; Faculty of Health Sciences, Universidad del Desarrollo, Concepcion, Chile.
Neuroimage. 2021 Dec 15;245:118764. doi: 10.1016/j.neuroimage.2021.118764. Epub 2021 Nov 27.
Prior studies have shown that the left posterior superior temporal sulcus (pSTS) and left temporo-parietal junction (TPJ) both contribute to phonological short-term memory, speech perception and speech production. Here, by conducting a within-subjects multi-factorial fMRI study, we dissociate the response profiles of these regions and a third region - the anterior ascending terminal branch of the left superior temporal sulcus (atSTS), which lies dorsal to pSTS and ventral to TPJ. First, we show that each region was more activated by (i) 1-back matching on visually presented verbal stimuli (words or pseudowords) compared to 1-back matching on visually presented non-verbal stimuli (pictures of objects or non-objects), and (ii) overt speech production than 1-back matching, across 8 types of stimuli (visually presented words, pseudowords, objects and non-objects and aurally presented words, pseudowords, object sounds and meaningless hums). The response properties of the three regions dissociated within the auditory modality. In left TPJ, activation was higher for auditory stimuli that were non-verbal (sounds of objects or meaningless hums) compared to verbal (words and pseudowords), irrespective of task (speech production or 1-back matching). In left pSTS, activation was higher for non-semantic stimuli (pseudowords and hums) than semantic stimuli (words and object sounds) on the dorsal pSTS surface (dpSTS), irrespective of task. In left atSTS, activation was not sensitive to either semantic or verbal content. The contrasting response properties of left TPJ, dpSTS and atSTS was cross-validated in an independent sample of 59 participants, using region-by-condition interactions. We also show that each region participates in non-overlapping networks of frontal, parietal and cerebellar regions. Our results challenge previous claims about functional specialisation in the left posterior superior temporal lobe and motivate future studies to determine the timing and directionality of information flow in the brain networks involved in speech perception and production.
先前的研究表明，左后颞上沟（pSTS）和左颞顶联合区（TPJ）都对语音短期记忆、言语感知和言语产生有贡献。在此，通过进行一项受试者内多因素功能磁共振成像（fMRI）研究，我们区分了这些区域以及第三个区域——左颞上沟前升支终末分支（atSTS）的反应特征，atSTS位于pSTS的背侧和TPJ的腹侧。首先，我们发现与对视觉呈现的非言语刺激（物体或非物体的图片）进行1-back匹配相比，每个区域在对视觉呈现的言语刺激（单词或伪词）进行1-back匹配时激活更强，并且与1-back匹配相比，在进行公开言语产生时激活更强，涉及8种类型的刺激（视觉呈现的单词、伪词、物体和非物体以及听觉呈现的单词、伪词、物体声音和无意义的嗡嗡声）。这三个区域的反应特性在听觉模态内有所区分。在左TPJ中，与言语刺激（单词和伪词）相比，非言语听觉刺激（物体声音或无意义的嗡嗡声）的激活更高，与任务（言语产生或1-back匹配）无关。在左pSTS中，在背侧pSTS表面（dpSTS），与语义刺激（单词和物体声音）相比，非语义刺激（伪词和嗡嗡声）的激活更高，与任务无关。在左atSTS中，激活对语义或言语内容均不敏感。使用条件×区域交互作用，在59名参与者的独立样本中对左TPJ、dpSTS和atSTS的对比反应特性进行了交叉验证。我们还表明，每个区域都参与了额叶、顶叶和小脑区域的非重叠网络。我们的结果挑战了先前关于左后颞上叶功能特化的说法，并促使未来的研究确定参与言语感知和产生的脑网络中信息流的时间和方向性。
