Fig 1.
(A) The timing of one trial. (B) On any given trial, the subject was relying on one of six possible states of covert attention. These corresponded to a visual (upper panel) or auditory focus of attention (lower panel). In addition, subjects focused attention to the left, right or upward perceptual space. (C) A structural diagram showing that three stimuli each are presented in the visual and auditory modalities while the subjects also maintain spatial attention to any of three locations. Examples of targets indicating a switch or hold between sensory modalities or moon cues indicating a shift in spatial attention are also shown. (D) An example of a run and the independent blocking structure for object and spatial attention. The individual blocks of the two attended sensory modalities (visual and auditory conditions) as a function of spatial location were analyzed in both EEG and fMRI data analysis.
Fig 2.
Behavioral performance and the correlation between behavioral reaction time.
(A) The performance accuracy during visual and auditory target detection. (B) The false alarms for visual and auditory blocks. (C) The mean RT from combing visual and auditory on three target locations, i.e. left (L) up (U) right (R). Error bars represent, for each condition, ±SE across subjects.
Fig 3.
Group-level ITC and ERSP scalp topographies based on theta band oscillation.
Up-left: Group-level topography of ITC for visual attention. Up-right: Group-level topography of ERSP for visual attention. Down-left: Group-level topography of ITC for auditory attention. Down-right: Group-level topography of ERSP for auditory attention. All results averaged across the three spatial (L, U and R) conditions.
Fig 4.
Crossmodal effect of sensory modality on differences of BOLD signal.
A direct comparison between BOLD activity from all visual attention blocks was compared with all auditory attention blocks. The auditory cortex (STG) shows significantly greater activity (blue) during auditory attention, while the visual cortex, the regions relevant with DMN (ACC, AG, PCC and Precuneus) and left dorsal frontal parietal network shows significantly greater activity (yellow and red) during the visual attention compared to the auditory task. This was an exploratory analysis, with p<0.05 corrected. Results are collapsed across three spatial locations (L, U and R).
Table 1.
Average MNI coordinates (peak), size and t-value of clusters showing significant Crossmodal effect of sensory modality on differences of BOLD signal (P < 0.05, corrected).
Fig 5.
Supramodal correlates of theta amplitude based on conjunction analysis.
(p < 0.05, corrected). Results are collapsed across 3 spatial (L, U and R) conditions. There is an overall negative relationship of theta and BOLD in bilateral PCC and CN, left mSFG and AG, right SFG (blue). There is an overall positive relationship of theta and BOLD in right ventral PreG (red).
Table 2.
Average MNI coordinates (peak), size and t-value of clusters showing significantly correlation between EEG oscillation and BOLD activity (P < 0.05, corrected).
Fig 6.
Modality specific interactions between theta and BOLD activity.
(p < 0.05, corrected, Red and yellow: positive relationship; blue: negative relationship). (A) Correlates of theta amplitude and BOLD for all visual conditions. Positive relationships are located in bilateral SMA and PostG, left SFG and SPL, right PreG, RO and IL. Negative relationships are localized to mSFG, ACC, SFG, PCC and CN both hemispheres, and right AG and IPL. (B). Correlates of theta amplitude and BOLD for all auditory conditions. Positive relationships are located in right PreG, RO and STG. Negative relationships are localized to mSFG, ACC, Cuneus and PCC both hemispheres, and left AG and MOG.