Journal: PLOS Biology

Article Title: Attention to audiovisual speech shapes neural processing through feedback-feedforward loops between different nodes of the speech network

doi: 10.1371/journal.pbio.3002534

Figure Lengend Snippet: (A) TRFs were estimated using the same speech amplitude envelopes as in our SER analysis, separately for the dialogue and background streams. (B) Average TRFs over frontocentral electrodes, with points indicating significant differences between the 2 TRFs (paired permutation t test df = 18, note the 2 streams have separate y-scales). (C) Left: RDMs were constructed using TRFs for all 16 conditions (first the 8 attend speech conditions and thereafter the 8 ignore speech conditions). This involved pairwise correlations for each condition combination at each time point across EEG channels. The upper left corner shows the average TRF RDMs for both dialogue and background streams. The plot in the left corner displays the correlation between an attentional task model (attend speech vs. ignore speech, att. vs. ign.) and the 2 TRF RDM time series, with significant points displayed below the plot (FDR corrected, one-sample t test, df = 19). Right: Similar to TRFs, fMRI RDMs were constructed using searchlight SVM decoding across the 16 conditions, resulting in voxel-specific RDMs. Regions with above-average correlations between the attentional task model and fMRI RDMs are displayed (HPC parcellation). Shading indicates ± SEM. Code and processed EEG and fMRI data used to generate this figure are archived on the Open Science Framework; HTTPS://DOI.ORG/10.17605/OSF.IO/AGXTH . EEG, electroencephalography; RDM, representational dissimilarity matrix; SER, speech envelope reconstruction; SVM, support vector machine; TRF, temporal response function.

Article Snippet: The EEG data were recorded with a BrainVision actiCHamp amplifier (128 channels) and a BrainVision actiCAP snap electrode cap with an actiCAP slim electrode set of 128 active electrodes (Brain Products GmbH, Gilching, Germany).

Techniques: Construct, Plasmid Preparation

Journal: PLOS Biology

Article Title: Attention to audiovisual speech shapes neural processing through feedback-feedforward loops between different nodes of the speech network

doi: 10.1371/journal.pbio.3002534

Figure Lengend Snippet: TRFs were separately estimated for the dialogue (upper part) and background streams (lower part) for each combination of semantic coherence and audiovisual quality and EEG channel. Average TRFs are displayed for frontocentral electrodes in the middle column (attend speech: red, ignore speech: blue). We constructed TRF RDMs for each time point by correlating each EEG channel TRF pairwise across conditions and participants. Similar fMRI RDMs were constructed based on SVM decoding between the 16 condition pairs from fMRI data. Thus, we constructed similar RDMs for the EEG and the fMRI, allowing us to fuse information from both datasets by correlating vectorised TRF RDMs with fMRI RDMs, controlling for task and opposite speech stream TRF RDMs (see ). To identify fMRI activations which corresponded to TRF RDMs at different time points, we conducted one-sample t tests ( df = 18, FDR corrected) averaged across the HCP parcellation ROIs. Six time points of this TRF-fMRI RSA analysis are displayed for both the dialogue (upper part) and background streams (lower part) on the right side of the figure. For the full-time series, refer to and Videos. Code and processed EEG and fMRI data used to generate this figure are archived on the Open Science Framework; HTTPS://DOI.ORG/10.17605/OSF.IO/AGXTH . EEG, electroencephalography; fMRI, functional magnetic resonance imaging; RDM, representational dissimilarity matrix; ROI, region-of-interest; RSA, representational similarity analysis; SVM, support vector machine; TRF, temporal response function.

Article Snippet: The EEG data were recorded with a BrainVision actiCHamp amplifier (128 channels) and a BrainVision actiCAP snap electrode cap with an actiCAP slim electrode set of 128 active electrodes (Brain Products GmbH, Gilching, Germany).

Techniques: Construct, Functional Assay, Magnetic Resonance Imaging, Plasmid Preparation