Figure 1.
While presenting participants with an attended visual n-back task we investigated cross-modal effects of ignored auditory events consisting of frequent standard tones of 1000 Hz and infrequent deviant tones of 1300 Hz on brain responses (ISI = 1.4 s).
The experiment consisted of 5 blocks per condition (27 s) and 15 fixation (baseline) blocks (15 s).
Figure 2.
A) Behavioral results of WM-load manipulation on subjects' hit rates (correct %) and reaction times (both M±SEM).
In EEGfMRI subjects (n = 15) GLMs confirmed a significant increase in RTs and decrease in hit rates when WM-load increased. This was replicated in EEG-only subjects (n = 5), but only significant for hit rates. B: Neural activation of visual WM-load, displayed by a contrasts from a random-effects GLM testing for general unsigned differences between visual WM load conditions (F-contrast, F>14.46, p<.05, FWE-corrected, k>20). Within the left dorsolateral prefrontal cortex (DLPFC) and the left inferior parietal cortex (IPC), cluster mean voxel activation (±SEM) were displayed via bar charts. MNI coordinates indicate the location of the maximum within the respective cluster.
Table 1.
Activation patterns corresponding to the effects of visual WM load.
Table 2.
Mean values of subjects’ behavioral performance (mean percent correct hit rates and reaction times in ms, SD in brackets) in three visual WM conditions.
Figure 3.
Cross-modal effects of visual WM load on auditory processing.
Grand average waveforms represent the evoked responses to unattended standard sounds, measured inside the scanner (EEG-fMRI), and outside (EEG-only) under different crossmodal visual WM-load manipulations. For each WM condition, topographic maps are shown at the latency of the N1 peak at Fz. Line plots below the figures show the condition-effect on AEPs (absolute peak-to-peak N1-P2 amplitude).
Table 3.
Auditory evoked potential (absolute N1-P2 amplitude, M µV, SD in brackets) in different cross-modal visual WM conditions in two different populations.
Figure 4.
Sagittal and axial slices displaying overlap of inter-subject variations in N1 peak values explaining BOLD variance with frontal core WM regions.
Activation patterns resulted from a random-effects GLM including condition-wise N1 peak values for each participant as covariates, inclusively masked (p<.05 uncorrected) with the effects of interest of the HRF regressors. Green indicates the average effect of the AEP amplitudes (F>6.48, p<.05, MC-corrected), yellow indicates the main effect of AEP amplitudes, accounting for the cross-modal visual WM-condition (T>3.29, p<.05, MC-corrected). Both contrasts are overlaid on background blue coloring indicating the main n-back effect shown in Figure 3. DLPFC = dorsolateral prefrontal cortex, MPG = medial prefrontal gyrus, IPC = inferior parietal cortex.
Table 4.
N1-peak amplitudes covarying with BOLD signal (ANCOVA).
Table 5.
Region-of-interest analyses of primary auditory cortices (AC), averages of mean activation (SD in brackets), subject to a generalized linear model (GLM) testing for significant differences between conditions.
Figure 5.
Functional connectivity of left auditory cortex (PPI seed region) in different visual WM-load conditions.
The top display reflects the parametrically increasing functional connectivity with core regions from the visual WM-load (p<.05, MC-corrected, k>125). The red circled region survives inclusive masking with the initial GLM testing for a parametric WM.-load increase. Lower displays show task-dependent functional connectivity of left AC in different conditions. All contrasts resulted from a group-level random-effects GLM analyzing effects for PPI interaction parameters (task by seed region) (p<.05, FWE-corrected, k>20).
Table 6.
Functional connectivity of left auditory cortex in different visual WM conditions.
Figure 6.
Outline of the reciprocal analysis strategy carried out with simultaneous EEG-fMRI, investigating differential effects of primary task visual working memory load on secondary task fundamental auditory processing.
fMRI = functional magnetic resonance imaging, WM = working memory, ERP = event-related potential, ANCOVA = analysis of covariance, BOLD = blood oxygenation level dependent, ROI = region-of-interest, PPI = psychophysiological interaction.