Electrical neuroimaging of voluntary audiospatial attention: evidence for a supramodal attention control network

Abstract

Previous attempts to investigate the supramodal nature of attentional control have focused primarily on identifying neuroanatomical overlap in the frontoparietal systems activated during voluntary shifts of spatial attention in different sensory modalities. However, the activation of the same neural structures is insufficient evidence for a supramodal system, as the same brain regions could interact with one another in very different ways during shifts of attention in different modalities. Thus, to explore the similarity of the functional networks, it is necessary to identify the neural structures involved and to examine the timing and sequence of activities within the network. To this end, we used an electrical neuroimaging technique to localize the neural sources of electroencephalographic signals recorded from human subjects during audiospatial shifts of attention and to examine the timing and sequence of activities within several regions of interest. We then compared the results to an analogous study of visuospatial attention shifts. Similar frontal and parietal regions were activated during visual and auditory shifts of attention, and the timing of activities within these regions was nearly identical. Following this modality-independent sequence of attention-control activity, activity in the relevant sensory cortex was enhanced in anticipation of the response-relevant target. These results are consistent with the hypothesis that a single supramodal network of frontal and parietal regions mediates voluntary shifts of spatial attention and controls the flow of sensory information in modality-specific sensory pathways.

Figure 1.

Illustration of events on a single trial.

Figure 2.

Spatiotemporal pattern of theta-band (4–7 Hz) activity associated with top-down attention control. Shown are surface-rendered maps of statistically significant increases in activity for shift cues relative to the no-shift cue during four 50 ms time windows. Activity for shift-left and shift-right cues was collapsed such that the left hemisphere displays activity ipsilateral to the cued location and the right hemisphere displays activity contralateral to the cued location. MFG, middle frontal gyrus.

Figure 3.

Time courses of beamformer power estimates in auditory, parietal, and frontal ROIs. Beamformer time courses are provided in q values, which represent the percentage change in shift-cue activity relative to the neutral-cue activity in the same time interval.

Figure 4.

Attention control activities in auditory and visual space. a, Overview of the neural sources of attention-control-related theta-band EEG activity during shifts of attention in auditory space (present study) and visual space (Green and McDonald, 2008). b, Proposed model of supramodal attention control. MFG, Middle frontal gyrus; IOG, inferior occipital gyrus; ITG, inferior temporal gyrus; SFG, superior frontal gyrus.