Auditory spatial attention modulates the unmasking effect of perceptual separation in a "cocktail party" environment

Spatial release from informational masking enhances the early cortical representation of speech sounds

INTRODUCTION

Spatial separation between competing speech streams reduces their confusion (informational masking), improving speech processing under challenging listening conditions. The precise stages of auditory processing involved in this benefit are not fully understood. This study used event-related potentials to examine the processing of target speech under conditions of informational masking and its spatial release.

METHODS

Participants detected noise-vocoded target speech presented with two-talker noise-vocoded masking speech. In separate conditions, the same set of targets were spatially co-located with maskers to produce informational masking and spatially separated from maskers using a perceptual manipulation to release the informational masking.

RESULTS

An increase in N1 and P2 amplitude, consistent with cortical auditory evoked potentials, and a later sustained positivity (P300) were observed in response to target onsets only under conditions supporting release from informational masking. At target intensities above masking threshold in both spatial conditions, N1 and P2 latencies were shorter when targets and maskers were perceptually separated.

DISCUSSION

These results indicate that spatial release from informational masking benefits speech representation beginning in the early stages of auditory perception. Additionally, these results suggest that the auditory evoked potential itself may be heavily dependent upon how information is perceptually organized rather than physically organized.

The Effects of Attention on the Syllable-Induced Prepulse Inhibition of the Startle Reflex and Cortical EEG Responses against Energetic or Informational Masking in Humans

Prepulse inhibition (PPI) is the reduction in the acoustic startle reflex (ASR) when the startling stimulus (pulse) is preceded by a weaker, non-starting stimulus. This can be enhanced by facilitating selective attention to the prepulse against a noise-masking background. On the other hand, the facilitation of selective attention to a target speech can release the target speech from masking, particularly from speech informational masking. It is not clear whether attentional regulation also affects PPI in this kind of auditory masking. This study used a speech syllable as the prepulse to examine whether the masker type and perceptual spatial attention can affect the PPI or the scalp EEG responses to the prepulse in healthy younger-adult humans, and whether the ERPs evoked by the prepulse can predict the PPI intensity of the ASR. The results showed that the speech masker produced a larger masking effect than the noise masker, and the perceptual spatial separation facilitated selective attention to the prepulse, enhancing both the N1 component of the prepulse syllable and the PPI of the ASR, particularly when the masker was speech. In addition, there was no significant correlation between the PPI and ERPs under any of the conditions, but the perceptual separation-induced PPI enhancement and ERP N1P2 peak-to-peak amplitude enhancement were correlated under the speech-masking condition. Thus, the attention-mediated PPI is useful for differentiating noise energetic masking and speech informational masking, and the perceptual separation-induced release of the prepulse from informational masking is more associated with attention-mediated early cortical unmasking processing than with energetic masking. However, the processes for the PPI of the ASR and the cortical responses to the prepulse are mediated by different neural mechanisms.

Neural Correlates of Attentional Modulation of Prepulse Inhibition

Prepulse inhibition (PPI) refers to the suppression of the startle reflex when the intense startling stimulus is shortly (20–500 ms) preceded by a weak non-startling stimulus (prepulse). Although the main neural correlates of PPI lie in the brainstem, previous research has revealed that PPI can be top-down modulated by attention. However, in the previous attend-to-prepulse PPI paradigm, only continuous prepulse but not discrete prepulse (20 ms) could elicit attentional modulation of PPI. Also, the relationship between the attentional enhancement of PPI and the changes in early cortical representations of prepulse signals is unclear. This study develops a novel attend-to-prepulse PPI task, when the discrete prepulse is set at 150 ms at a lead interval of 270 ms, and reveals that the PPI with attended prepulse is larger than the PPI with ignored prepulse. In addition, the early cortical representations (N1/P2 complex) of the prepulse show dissociation between the attended and ignored prepulse. N1 component is enhanced by directed attention, and the attentional increase of the N1 component is positively correlated with the attentional enhancement of PPI, whereas the P2 component is not affected by attentional modulation. Thus, directed attention to the prepulse can enhance both PPI and the early cortical representation of the prepulse signal (N1).