Sustained Magnetic Responses in Temporal Cortex Reflect Instantaneous Significance of Approaching and Receding Sounds

Abstract

Rising sound intensity often signals an approaching sound source and can serve as a powerful warning cue, eliciting phasic attention, perception biases and emotional responses. How the evaluation of approaching sounds unfolds over time remains elusive. Here, we capitalised on the temporal resolution of magnetoencephalograpy (MEG) to investigate in humans a dynamic encoding of perceiving approaching and receding sounds. We compared magnetic responses to intensity envelopes of complex sounds to those of white noise sounds, in which intensity change is not perceived as approaching. Sustained magnetic fields over temporal sensors tracked intensity change in complex sounds in an approximately linear fashion, an effect not seen for intensity change in white noise sounds, or for overall intensity. Hence, these fields are likely to track approach/recession, but not the apparent (instantaneous) distance of the sound source, or its intensity as such. As a likely source of this activity, the bilateral inferior temporal gyrus and right temporo-parietal junction emerged. Our results indicate that discrete temporal cortical areas parametrically encode behavioural significance in moving sound sources where the signal unfolded in a manner reminiscent of evidence accumulation. This may help an understanding of how acoustic percepts are evaluated as behaviourally relevant, where our results highlight a crucial role of cortical areas.

Fig 1. Experimental procedures.

A: Illustration of the sound stimuli used in this experiment. After an initial constant sound pressure level [SPL] segment of 300 ms, SPL rises (or falls) over 1000 ms; the sound terminates with another segment of constant SPL. B: Intra-trial procedure. Our analysis focuses on responses to the sound. Participants were instructed to blink only in designated periods between sounds. A visual task was introduced between sound presentations to increase alertness and render the eye blink task plausible to the participant. C: Global sound responses to all sound stimuli, for all sensors. The intensity change segment of the sound was between 300 and 1300 ms.

Fig 2. Sustained fields across the scalp.

Fields are shown for the most significant time point of the intensity change x carrier frequency interaction (1215 ms after sound onset, 915 ms into the intensity change), averaged across the group, for the four different conditions in this interaction. Dots: average sensor positions. X: Most significant point in sensor space and its contralateral counterpart (see Fig 3). Black outline: location of significant left hemispheric cluster at 1215 ms. Grey outline: location of significant right hemispheric cluster at its most significant time point, 1390 ms after sound onset.

Fig 3. Field time courses.

All time courses are shown for the most significant point in sensor space of the intensity change x carrier frequency interaction, and its contralateral counterpart, averaged across the group, for the four different conditions in this interaction. Location of this point in space is plotted in Fig 2. Averaged across the group, the closest sensors at these scalp points were MLT26/MRT16. Grey background: significant time points at these points in space.

Fig 4. Source-level analysis.

Putative sources of the observed intensity change x carrier frequency interaction: significant clusters in a group-level t-test on source inversion results of the intensity change x carrier frequency interaction (see Table 2).