Listening in Naturalistic Scenes: What Can Functional Near-Infrared Spectroscopy and Intersubject Correlation Analysis Tell Us About the Underlying Brain Activity?

The participants were 30 university students aged 18 to 23 years with self-reported normal hearing, so the findings do not show what happens for people with hearing loss, older adults, or clinical populations. The acoustic scenes were recreated in a laboratory with generic binaural recordings and non-individualized earphone equalization, so each participant's sense of a real space may have been limited. The brain-imaging narratives and the ratings narratives were different, although matched, so the subjective ratings may not exactly represent the listening experience during imaging. fNIRS only reaches superficial cortex and has modest spatial resolution, so it could not measure deeper regions such as the cingulo-opercular system and could not confirm exact channel anatomy. The study had no concurrent measure of attention during the long narratives, and some participants said they struggled to maintain attention, so lapses may have weakened the brain results. The content questions were answered at about 50% correct and mainly covered the start of narratives, so they did not show whether participants stayed engaged throughout. The brain synchronization values were small, with a maximum intersubject correlation around 0.03, and the work was exploratory across 52 channels, so the absence of brain-effort relationships should be read cautiously. Left-handed participants and participants without strong hand preference were included, and the left-handed subgroup was too small to analyze separately.

The supplied text does not include a conflict-of-interest declaration. The provided publication types note research support, non-U.S. government.

Do not treat the frontal and temporal brain activity as a clinical marker of listening effort. The study found common brain activity across listeners, but it did not find a relationship between brain activity and reported effort, reported intelligibility, or scene signal-to-noise ratio.