Back

Attentional disengagement during external and internal distractions reduces neural speech tracking in background noise

Ren, Y.; Cui, M. E.; Herrmann, B.

2025-10-17 neuroscience
10.1101/2025.10.17.683146 bioRxiv
Show abstract

Within-situation disengagement - the mental withdrawal during conversations in acoustically challenging environments - is a common experience of older people with hearing difficulties. Yet, most research on the neural mechanisms of attentional disengagement from speech listening has focused on the distraction by one competing speaker, whereas within-situation disengagement is often characterized by distraction towards external visual stimuli or internal thoughts and occurs in situations with ambient, multi-talker background masking. Across three electroencephalography (EEG) experiments in human participants of either sex, the current study examined how disengagement due to external and internal distractions affect the neural tracking of speech masked by different levels of multi-talker babble (speech in quiet, +6 dB, and -3 dB SNR). We observed enhanced early neural responses (<0.2 s) to the speech envelope for speech masked by background babble compared to speech in quiet (Experiments 1-3), suggesting stochastic facilitation. Importantly, neural tracking of the speech envelope was reduced when individuals were distracted by a visual-stimulus stream (Experiment 2) and by internal thought and imagination (Experiment 3). There were some indices suggesting the greatest disengagement-related decline in neural speech tracking occurs for the most difficult speech-masking condition, but this was not consistent across all measures. The current data show that disengagement due to external and internal distractions yield decreases in neural speech tracking, potentially suggesting converging neural pathways through which gain is downregulated in auditory cortex. These results indicate that disengagement from listening can be identified through non-invasive neural measures.

Published in The Journal of Neuroscience (predicted rank #1) · training set

Matching journals

The top 7 journals account for 50% of the predicted probability mass.

50% of probability mass above

"Similar papers" are the closest papers from that journal in the model's embedding space. They show what the match is built on, but the ranking comes mostly from a classifier over the whole training set, not from these examples alone.