EEG-Based Intersubject Correlations Reflect Selective Attention in a Competing Speaker Scenario

Rosenkranz, Marc; Holtze, Björn; Jaeger, Manuela; Debener, Stefan

doi:10.3389/fnins.2021.685774

Cited by 13 publications

(28 citation statements)

References 45 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…We observed that ISC was stronger during story listening compared to rest for all three spatial components. This is consistent with previous studies showing that ISC is stronger when attending to a coherent narrative compared to when attention is directed elsewhere (S. S. Cohen et al, 2018; Ki et al, 2016; Kuhlen et al, 2012; Rosenkranz et al, 2021) or compared to rest (Hasson et al, 2004; Wilson et al, 2008). Inter-subject correlation is often thought of as a neural signature of engagement, because ISC is stronger for captivating or exciting stimuli (Hasson et al, 2010; Schmälzle et al, 2015) and predicts behavioral engagement measures (S. S. Cohen et al, 2017; Dikker et al, 2017; Dmochowski et al, 2014; Poulsen et al, 2017; Song et al, 2021).…”

Section: Discussionsupporting

confidence: 92%

“…We observed that ISC was stronger during story listening compared to rest for all three spatial components. This is consistent with previous studies showing that ISC is stronger when attending to a coherent narrative compared to when attention is directed elsewhere (Cohen et al, 2018;Ki et al, 2016;Kuhlen et al, 2012;Rosenkranz et al, 2021) or compared to rest (Hasson et al, 2004;Wilson et al, 2008). Observing significant ISC despite the presence of background noise suggests that participants were able to remain engaged despite presumably missing bits of the story during periods when the SNR was low.…”

Section: Speech Intelligibility Declines With Decreasing Signal To Noise Ratiosupporting

confidence: 91%

“…The strength of synchrony across individuals, quantified as inter-subject correlation (ISC; Hasson et al, 2004, 2008; Dmochowski et al, 2012, 2014), is stronger when stimuli are captivating or exciting (Hasson et al, 2010; Schmälzle et al, 2015), and is predictive of behavioral measures reflecting engagement (S. S. Cohen et al, 2017; Dikker et al, 2017; Dmochowski et al, 2014; Poulsen et al, 2017; Song et al, 2021) and recall of the materials (Chan et al, 2019; S. S. Cohen et al, 2018; S. S. Cohen & Parra, 2016; Davidesco et al, 2019; Hasson, Furman, et al, 2008; Piazza et al, 2021; Song et al, 2021; Stephens et al, 2010). Conversely, ISC is reduced when individuals do not attend to naturalistic materials (S. S. Cohen et al, 2018; Ki et al, 2016; Kuhlen et al, 2012; Rosenkranz et al, 2021) or when stimuli are unstructured or temporally scrambled (Dmochowski et al, 2012; Hasson, Yang, et al, 2008).…”

Section: Introductionmentioning

confidence: 99%

See 2 more Smart Citations

Neural activity during story listening is synchronized across individuals despite acoustic masking

Irsik

Johnsrude

Herrmann

2021

Preprint

View full text Add to dashboard Cite

Older people with hearing problems often experience difficulties understanding speech in the presence of background sound. As a result, they may disengage in social situations, which has been associated with negative psychosocial health outcomes. Measuring listening (dis-)engagement during challenging listening has received little attention thus far. We recruit normal-hearing human adults (both sexes) and investigate how speech intelligibility and engagement during naturalistic story listening is affected by the level of acoustic masking (12-talker babble). In Experiment 1, we observed that word-report scores were above 80% for all but the lowest SNR (-3 dB SNR) we tested, at which performance dropped to 54%. In Experiment 2, we calculated inter-subject correlation (ISC) in electroencephalography (EEG) data to identify dynamic spatial patterns of shared neural activity evoked by the stories. ISC followed a similar overall quadratic pattern as intelligibility data, however comparing ISC and intelligibility directly demonstrated that word-report performance declined more strongly with decreasing SNR compared to ISC. Observing significant ISC despite the presence of background noise suggests that participants were able to remain engaged despite missing segments of the story during especially difficult SNRs. Our work provides a novel approach to observe speech intelligibility and listener engagement using ecologically valid spoken materials which can be used to investigate (dis)engagement in older adults with hearing impairment.

show abstract

Section: Discussionsupporting

confidence: 92%

“…We observed that ISC was stronger during story listening compared to rest for all three spatial components. This is consistent with previous studies showing that ISC is stronger when attending to a coherent narrative compared to when attention is directed elsewhere (Cohen et al, 2018;Ki et al, 2016;Kuhlen et al, 2012;Rosenkranz et al, 2021) or compared to rest (Hasson et al, 2004;Wilson et al, 2008). Observing significant ISC despite the presence of background noise suggests that participants were able to remain engaged despite presumably missing bits of the story during periods when the SNR was low.…”

Section: Speech Intelligibility Declines With Decreasing Signal To Noise Ratiosupporting

confidence: 91%

Section: Introductionmentioning

confidence: 99%

See 1 more Smart Citation

Neural activity during story listening is synchronized across individuals despite acoustic masking

Irsik

Johnsrude

Herrmann

2021

Preprint

View full text Add to dashboard Cite

show abstract

“…However, despite being highly significant at the group-level, this effect was not ubiquitous across all participants, and was statistically significant only in a subset of participants (∼30%). To the best of our knowledge, previous studies have not directly addressed this inter-subject variance in attentional-bias, however several pervious papers that present individual-level data seem to exhibit similar variability (Ding and Simon 2012a, 2012b; Fuglsang et al 2017; Rosenkranz et al 2021). In the current data set, inter-subject variability in the magnitude of attentional bias was not correlated with working memory capacity (OSPAN), which sometimes drives individual differences in attentional abilities (Beaman et al 2007; Sörqvist and Rönnberg 2014a; Lambez et al 2020).…”

Section: Discussionmentioning

confidence: 93%

Capacity and tradeoffs in neural encoding of concurrent speech during Selective and Distributed Attention

Kaufman

Golumbic

2022

Preprint

View full text Add to dashboard Cite

Speech comprehension is severely compromised when several people talk at once, due to limited perceptual and cognitive resources. Under some circumstances listeners can employ top-down attention to prioritize the processing of task-relevant speech. However, whether the system can effectively represent more than one speech input remains highly debated. Here we studied how task-relevance affects the neural representation of concurrent speakers under two extreme conditions: when only one speaker was task-relevant (Selective Attention), vs. when two speakers were equally relevant (Distributed Attention). Neural activity was measured using magnetoencephalography (MEG) and we analysed the speech-tracking responses to both speakers. Crucially, we explored different hypotheses as to how the brain may have represented the two speech streams, without making a-priori assumptions regarding participants' internal allocation of attention. Results indicate that neural tracking of concurrent speech did not fully mirror their instructed task-relevance. When Distributed Attention was required, we observed a tradeoff between the two speakers despite their equal task-relevance, akin to the top-down modulation observed during Selective Attention. This points to the system's inherent limitation to fully process two speech streams, and highlights the complex nature of attention, particularly for continuous speech.

show abstract

“…Previous work considered CorrCA components corresponding to three highest ISC values [32]. In this work…”

Section: Resultsmentioning

confidence: 99%

Identifying Reliable Neural Components and P300 Latency using Correlated Component Analysis

Gupta

Hazarika

2022

Preprint

View full text Add to dashboard Cite

This paper applies Correlated component analysis (CorrCA) to extract the reliable and maximally correlated P300 component. Trials collected over time during an oddball electroencephalogram (EEG) experiment are first segregated into the target and non-target trial cohorts. CorrCA was then applied to both cohorts separately. Event related potential (ERP) image plots of CorrCA components from target trials showed a coherent structure for P300 latency around 400 ms when compared to plots of single electrode (Cz) trials. A difference in peak amplitude of 0.37 micro volts was observed between averaged first CorrCA component and averaged target trials at Cz electrode for able bodied subjects. Inter trial Correlation (ITC) obtained using CorrCA algorithm is applied to detect subjects who performed poorly during the oddball experiment. We also used a forward model considering sixteen channels to identify the spatial patterns of neural activity for both able and disabled subjects. It was observed that disabled subjects illustrated lower neural activation during target trials in most brain areas when compared to able-bodied subjects.

show abstract

EEG-Based Intersubject Correlations Reflect Selective Attention in a Competing Speaker Scenario

Cited by 13 publications

References 45 publications

Neural activity during story listening is synchronized across individuals despite acoustic masking

Neural activity during story listening is synchronized across individuals despite acoustic masking

Capacity and tradeoffs in neural encoding of concurrent speech during Selective and Distributed Attention

Identifying Reliable Neural Components and P300 Latency using Correlated Component Analysis

Contact Info

Product

Resources

About