Beta/Gamma Oscillations and Event-Related Potentials Indicate Aberrant Multisensory Processing in Schizophrenia

Abstract

Recent behavioral and neuroimaging studies have suggested multisensory processing deficits in patients with schizophrenia (SCZ). Thus far, the neural mechanisms underlying these deficits are not well understood. Previous studies with unisensory stimulation have shown altered neural oscillations in SCZ. As such, altered oscillations could contribute to aberrant multisensory processing in this patient group. To test this assumption, we conducted an electroencephalography (EEG) study in 15 SCZ and 15 control participants in whom we examined neural oscillations and event-related potentials (ERPs) in the sound-induced flash illusion (SIFI). In the SIFI multiple auditory stimuli that are presented alongside a single visual stimulus can induce the illusory percept of multiple visual stimuli. In SCZ and control participants we compared ERPs and neural oscillations between trials that induced an illusion and trials that did not induce an illusion. On the behavioral level, SCZ (55.7%) and control participants (55.4%) did not significantly differ in illusion rates. The analysis of ERPs revealed diminished amplitudes and altered multisensory processing in SCZ compared to controls around 135 ms after stimulus onset. Moreover, the analysis of neural oscillations revealed altered 25-35 Hz power after 100 to 150 ms over occipital scalp for SCZ compared to controls. Our findings extend previous observations of aberrant neural oscillations in unisensory perception paradigms. They suggest that altered ERPs and altered occipital beta/gamma band power reflect aberrant multisensory processing in SCZ.

Keywords: attention; audiovisual; crossmodal; electroencephalography; neural oscillations; oscillatory activity; perception; sound-induced flash illusion.

FIGURE 1

Setup of the sound-induced flash illusion paradigm. Participants fixated a central white cross while being presented with stimuli of the SIFI paradigm (Left). In a critical SIFI trial (i.e., A₂V₁) a single flash presented alongside two rapidly repeating tones is either perceived as one or two flashes. (Right) depicts the timeline of the critical SIFI trial. The visual stimulus and the first auditory stimulus are presented simultaneously. The second auditory stimulus is presented 57 ms after the onset of the first stimulus. Six hundred milliseconds after the onset of the first stimulus, the fixation cross is replaced by a response cue, which comprised an empty circle that is presented in the center of the screen.

FIGURE 2

ERP effects of multisensory A₂V₁ trials. Outcome of the non-parametric tests with cluster-based correction for multiple comparisons between Perceptions (illusion vs. no-illusion), between Groups (SCZ vs. HC), and for the Perception differences between Groups (i.e., the Perception by Group interaction). The upper planes in panels (A–C) illustrate the ERP results for Perception (A), the main effect of Group (B), and the Perception by Group interaction (C). The lower planes depict topographic maps for the observed results with highlighted significant cluster electrodes. Time-point 0 indicates the onset of the first auditory and visual stimulus.

FIGURE 3

ERP effects of visual A₀V₁ and A₀V₂ trials. Outcome of the non-parametric tests with cluster-based correction for multiple comparisons between Conditions (A₀V₁ vs. A₀V₂) between Groups (SCZ vs. HC), and for the Condition differences between Groups (i.e., the Condition by Group interaction). The upper planes in panels (A–C) illustrate significant main effects of Condition (A), the ERP results for Group (B) and the Condition by Group interaction (C). The lower planes depict topographic maps for the observed effects with highlighted significant cluster electrodes. Time-point 0 indicates the onset of the first visual stimulus.

FIGURE 4

Total oscillatory power over occipital cortex. (A) Time-frequency representations (TFRs) at occipital electrodes in response to the critical multisensory A₂V₁ trials. In the early time-frequency window (100–150 ms, 25–35 Hz; highlighted in a box) oscillatory response patterns accompanying illusion and no-illusion trials were different between groups. In the control group, the 25–35 Hz total power was stronger in illusion compared to no-illusion trials. No such difference was found in SCZ. No effects were found in the late time-frequency window (200–240 ms, 25–35 Hz). (B) TFRs at occipital electrodes in response to unisensory visual trials. Total power in the early and late time-frequency window (each highlighted in a box) was significantly higher in the A₀V₂ condition compared to the A₀V₁ condition. No Group differences (i.e., SCZ vs. HC) or Group by Condition interactions were found. Time-point 0 indicates the onset of the first auditory and first visual stimulus.