Beta Oscillations Distinguish Between Two Forms of Mental Imagery While Gamma and Theta Activity Reflects Auditory Attention

Abstract

Visual sensory processing of external events decreases when attention is internally oriented toward self-generated thoughts and also differences in attenuation have been shown depending on the thought's modality (visual or auditory thought). The present study aims to assess whether such modulations occurs also in auditory modality. In order to investigate auditory sensory modulations, we compared a passive listening condition with two conditions in which attention was internally oriented as a part of a task; a visual imagery condition and an inner speech condition. EEG signal was recorded from 20 participants while they were exposed to auditory probes during these three conditions. ERP results showed no differences in N1 auditory response comparing the three conditions reflecting maintenance of evoked electrophysiological reactivity for auditory modality. Nonetheless, time-frequency analyses showed that gamma and theta power in frontal regions was higher for passive listening than for internal attentional conditions. Specifically, the reduced amplitude in early gamma and theta band during both inward attention conditions may reflect reduced conscious attention of the current auditory stimulation. Finally, different pattern of beta band activity was observed only during visual imagery which can reflect cross-modal integration between visual and auditory modalities and it can distinguish this form of mental imagery from the inner speech. Taken together, these results showed that attentional suppression mechanisms in auditory modality are different from visual modality during mental imagery processes. Our results about oscillatory activity also confirm the important role of gamma oscillations in auditory processing and the differential neural dynamics underlying the visual and auditory/verbal imagery.

Keywords: auditory attention; beta band activity; gamma band activity; mental imagery; self-generated thoughts; theta band activity.

FIGURE 1

Schematic of experimental paradigm. Each trial started by pressing a button triggering a fixation cross presented between 1 and 1.5 s. When the word “imagine” appears followed by the green circle, participants were instructed to think about any mental image, avoiding auditory elements. If the word “speech” was followed by a green circle participants were instructed to think using only their inner speech. Finally, if any of the two words were followed by a red circle, participants were instructed to passively hearing the auditory stimuli. Color cues were counterbalanced between blocks. Auditory probes are presented in each condition (1200 Hz, 75 dB SPL) with duration of 100 ms. The inter stimulus interval was randomly jittered between 1000 and 1500 ms.

FIGURE 2

ERPs to auditory probe. ERP waveforms for three different midline electrodes; Fz, FCz, and Cz. There are not differences between conditions for any of the early sensory components of ERP.

FIGURE 3

Time-frequency plots in the frontal area. Time-frequency plot of spectral power for each condition; passive listening (left), inner speech (middle) and visual imagery (right).

FIGURE 4

Time-frequency plots of significant differences between conditions in frontal areas. (A) Time-frequency power subtraction between passive listening and visual imagery. (B) Time-frequency power subtraction between passive listening and inner speech. (C) Time-frequency power subtraction between inner speech and visual imagery. The areas delimited by lines show the frequency and time range in which significant differences are found after multiple comparison correction described in Section “Permutation Test and Multiple Comparison Correction” for non-parametric permutation test (p < 0.05).

FIGURE 5

Time-frequency plots in the parietal area. Time-frequency plot of spectral power for each condition; passive listening (left), inner speech (middle), and visual imagery (right).

FIGURE 6

Time-frequency plots of significant differences between conditions in parietal areas. (A) Time-frequency power subtraction between visual imagery and passive listening. (B) Time-frequency power subtraction between visual imagery and inner speech. (C) Time-frequency power subtraction between passive listening and inner speech. (D) Time-frequency power subtraction between inner speech and visual imagery. The areas delimited by lines show the frequency and time range in which significant differences are found after multiple comparison correction described in Section “Permutation Test and Multiple Comparison Correction” for non-parametric permutation test (p < 0.05).