Decoding Brain Responses to Names and Voices across Different Vigilance States

Area of Science:

• Neuroscience
• Sleep Science
• Auditory Processing

Background:

• The brain processes auditory stimuli during sleep, but the underlying mechanisms are not fully understood.
• Paralinguistic features of sound, such as voice familiarity, influence offline auditory processing.

Purpose of the Study:

• To investigate similarities in brain activity across different sleep stages (N1-N3, REM, wake) using multivariate pattern analysis.
• To determine if paralinguistic and subjective salience of auditory stimuli can be decoded during sleep.
• To explore the differences in brain processing modes between wakefulness and sleep.

Main Methods:

• Multivariate pattern analysis (MVPA) applied to EEG data from 16 healthy sleepers.
• Stimuli varied in subjective (own name) and paralinguistic (familiar/unfamiliar voice) salience.
• Classification and generalization analyses across sleep stages (N1-N3, REM) and wakefulness.

Main Results:

• Paralinguistic salience (voice familiarity) was reliably decoded from EEG in N2 and N3 sleep.
• Classifiers trained on N2 data generalized to N3, and vice versa, indicating similar processing modes.
• No generalization of decoding was found between wakefulness and any sleep stage.
• Similar fronto-central topographical patterns were observed in N2 and N3 stages.

Conclusions:

• A uniform NREM-specific mechanism underlies auditory salience detection during sleep.
• Brain processing modes significantly differ between wakefulness and sleep.
• The findings suggest distinct neural architectures for processing auditory information while awake versus asleep.