Meienbrock A, Naumer M J, Doehrmann O, Singer W, Muckli L
Max Planck Institute for Brain Research, Department of Neurophysiology, Frankfurt/Main, Germany.
Neuropsychologia. 2007 Feb 1;45(3):531-9. doi: 10.1016/j.neuropsychologia.2006.05.018. Epub 2006 Jun 23.
The successful integration of visual and auditory stimuli requires information about whether visual and auditory signals originate from corresponding places in the external world. Here we report crossmodal effects of spatially congruent and incongruent audio-visual (AV) stimulation. Visual and auditory stimuli were presented from one of four horizontal locations in external space. Seven healthy human subjects had to assess the spatial fit of a visual stimulus (i.e. a gray-scaled picture of a cartoon dog) and a simultaneously presented auditory stimulus (i.e. a barking sound). Functional magnetic resonance imaging (fMRI) revealed two distinct networks of cortical regions that processed preferentially either spatially congruent or spatially incongruent AV stimuli. Whereas earlier visual areas responded preferentially to incongruent AV stimulation, higher visual areas of the temporal and parietal cortex (left inferior temporal gyrus [ITG], right posterior superior temporal gyrus/sulcus [pSTG/STS], left intra-parietal sulcus [IPS]) and frontal regions (left pre-central gyrus [PreCG], left dorsolateral pre-frontal cortex [DLPFC]) responded preferentially to congruent AV stimulation. A position-resolved analysis revealed three robust cortical representations for each of the four visual stimulus locations in retinotopic visual regions corresponding to the representation of the horizontal meridian in area V1 and at the dorsal and ventral borders between areas V2 and V3. While these regions of interest (ROIs) did not show any significant effect of spatial congruency, we found subregions within ROIs in the right hemisphere that showed an incongruency effect (i.e. an increased fMRI signal during spatially incongruent compared to congruent AV stimulation). We interpret this finding as a correlate of spatially distributed recurrent feedback during mismatch processing: whenever a spatial mismatch is detected in multisensory regions (such as the IPS), processing resources are re-directed to low-level visual areas.
视觉和听觉刺激的成功整合需要有关视觉和听觉信号是否源自外部世界相应位置的信息。在此，我们报告空间一致和不一致的视听（AV）刺激的跨模态效应。视觉和听觉刺激从外部空间的四个水平位置之一呈现。七名健康人类受试者必须评估视觉刺激（即卡通狗的灰度图片）和同时呈现的听觉刺激（即犬吠声）的空间匹配度。功能磁共振成像（fMRI）揭示了两个不同的皮质区域网络，它们优先处理空间一致或空间不一致的AV刺激。早期视觉区域优先对不一致的AV刺激做出反应，而颞叶和顶叶皮质的较高视觉区域（左侧颞下回[ITG]、右侧颞上回/颞上沟后部[pSTG/STS]、左侧顶内沟[IPS]）和额叶区域（左侧中央前回[PreCG]、左侧背外侧前额叶皮质[DLPFC]）优先对一致的AV刺激做出反应。位置分辨分析揭示了视网膜拓扑视觉区域中四个视觉刺激位置各自的三种强大皮质表征，对应于V1区域水平子午线以及V2和V3区域之间背侧和腹侧边界的表征。虽然这些感兴趣区域（ROIs）未显示出空间一致性的任何显著影响，但我们在右半球的ROIs内发现了显示不一致效应的子区域（即与一致的AV刺激相比，在空间不一致时fMRI信号增加）。我们将这一发现解释为不匹配处理期间空间分布的循环反馈的相关因素：每当在多感官区域（如IPS）检测到空间不匹配时，处理资源就会重新导向低级视觉区域。
