Predicting Speech in Noise Perception using Measures of Neural Entrainment

The ability to comprehend speech in a noisy environment, such as a cocktail party, varies greatly even among individuals with healthy hearing. The underlying reasons for these discrepancies in speech perception capabilities remain largely unknown. One potential explanation could be related to the neural processes involved in auditory attention.

This study seeks to determine the extent to which an individual can neurally synchronize with a continuous narrative using a temporal response function (TRF). It also aims to assess whether TRF measurements can provide insight into a person's performance on tasks involving selective attention and narrative comprehension.

The researchers recruited a cohort of twenty-five young and healthy individuals who participated in speech-in-noise (SIN) perception tasks. During these tasks, their neural activity was recorded using EEG. The EEG data, along with regressors (speech envelope and gammatone spectrogram) of the auditory stimulus, was then utilized to train TRF models. The cross-validation correlation coefficients derived from these models were subsequently employed in multiple regression models to predict each participant's SIN performance.

The findings of the study revealed a significant positive relationship between neural entrainment to attended stimuli and task performance, as well as a negative correlation between neural entrainment to unattended stimuli and task performance. This pattern was observed for both selective attention and narrative comprehension measures.

Furthermore, the differences between the correlation coefficients of attended and unattended stimuli were also significantly associated with improved task performance. The authors suggest that these measures could potentially serve as biomarkers for predicting an individual's ability to perceive speech in noisy environments.