Temporal integration in human auditory cortex is predominantly yoked to absolute time, not structure duration

Area of Science:

• Neuroscience
• Auditory Perception
• Computational Linguistics

Background:

• Neural integration in the auditory cortex is debated, with models differing on whether it's time-based or structure-based.
• Understanding how the brain processes speech sounds of varying durations is crucial for auditory perception models.

Purpose of the Study:

• To investigate whether neural integration in the human auditory cortex is primarily governed by absolute time or by the duration of sound structures.
• To determine the extent to which cortical computations reflect temporal processing versus structural processing in speech perception.

Main Methods:

• Utilized intracranial recordings in the human auditory cortex for spatiotemporally precise measurements.
• Developed a novel experimental and computational method to measure neural integration windows.
• Rescaled speech durations using time stretching and compression to manipulate sound structure durations.

Main Results:

• Observed significantly longer neural integration windows for time-stretched speech compared to normally paced speech.
• The increase in integration window duration was minimal (~5%) relative to the changes in speech structure durations.
• This effect was consistent across auditory cortex regions, including those specialized for speech processing.

Conclusions:

• Neural computations in the human auditory cortex are predominantly time-yoked, rather than being strictly dependent on speech structure durations.
• Findings constrain neurocomputational models of auditory and speech processing, emphasizing the role of temporal dynamics.
• The auditory cortex prioritizes temporal integration, with limited adaptation to altered speech structure durations.