Modulation of brain activity by selective attention to audiovisual dialogues

Abstract

<jats:title>Abstract</jats:title><jats:p>In real-life noisy situations, we can selectively attend to conversations in the presence of irrelevant voices, but neurocognitive mechanisms in such natural listening situations remaiin largely unexplored. Previous research has shown distributed activity in the mid superior temporal gyrus (STG) and sulcus (STS) while listening to speech and human voices, in the posterior STS and fusiform gyrus when combining auditory, visual and linguistic information, as well as in lefthemisphere temporal and frontal cortical areas during comprehension. In the present functional magnetic resonance imaging (fMRI) study, we investigated how selective attention modulates neural responses to naturalistic audiovisual dialogues. Our healthy adult participants (N = 15) selectively attended to video-taped dialogues between a man and woman in the presence of irrelevant continuous speech in the background. We modulated the auditory quality of dialogues with noise vocoding and their visual quality by masking speech-related facial movements. Both increased auditory quality and increased visual quality were associated with bilateral activity enhancements in the STG/STS. In addition, decreased audiovisual stimulus quality elicited enhanced fronto-parietal activity, presumably reflecting increased attentional demands. Finally, attention to the dialogues, in relation to a control task where a fixation cross was attended and the dialogue ignored, yielded enhanced activity in the left planum polare, angular gyrus, the right temporal pole, as well as in the orbitofrontal/ventromedial prefrontal cortex and posterior cingulate gyrus. Our findings suggest that naturalistic conversations effectively engage participants and reveal brain networks related to social perception in addition to speech and semantic processing networks.</jats:p>

Selective attention to audiovisual speech routes activity through recurrent feedback-feedforward loops between different nodes of the speech network

“…We find this unlikely, however, because such lapses would likely cause more variance in the background stream TRFs, rather than the change in amplitude seen in the present results. Furthermore, previous studies using the same paradigm [42] have found that in general, participants do not remember topics of the background stream. ignore speech task, for the exact order of the conditions see "Multivariate analysis of TRFs") and channel (128 channels) separately for the dialogue and the background streams.…”

“…The stimuli comprised dialogues between two (female and male) native Finnish speakers. Written informed consent has been obtained from the individual(s) for the publication of any potentially identifiable images or data included in this manuscript (see also, [38][39][40]42]). The dialogue topics were about neutral everyday subjects such as the weather.…”

“…The original dialogues [42] were recorded so that the talkers sat next to one another with their faces slightly tilted towards each other (see Fig. 1A).…”

“…That is, ECoG studies are limited by the extent of the implanted electrodes, while MEG/EEG source localization is relatively inaccurate especially in the case of simultaneously firing neuronal populations. [36] In contrast, functional magnetic resonance imaging (fMRI) provides better spatial resolution, revealing that selective attention to cocktail-party speech modulates information processing in not only lowlevel auditory regions but also in extensive superior temporal, inferior parietal, and inferior frontal brain regions (e.g., [37][38][39][40][41][42]). Furthermore, multivariate pattern analyses on cocktail party fMRI data have indicated that neuronal populations that show differential responses during selective attention to speech are distributed globally in disparate cortical regions.…”

Breathlessness in COPD: linking symptom clusters with brain activity

“…In the case of words, activity within lingual-fusiform gyrus has been shown to be influenced by emotional content [84], increasing with negative valence [85]. While there is some evidence to suggest that lingual gyrus may be connected to fear and anxiety related processes [86,87], other work has highlighted that greater cognitive demands and focused attention can also increase brain activity within a "motivated attentional framework" [84,88,89].…”