Functional MRI investigations of cortical mechanisms of auditory spatial attention

In everyday settings, spatial attention helps listeners isolate and understand individual sound sources. However, the neural mechanisms of auditory spatial attention (ASpA) are only partially understood. This thesis uses within-subject analysis of functional magnetic resonance imaging (fMRI) data to address fundamental questions regarding cortical mechanisms supporting ASpA by applying novel multi-voxel pattern analysis (MVPA) and resting-state functional connectivity (rsFC) approaches. A series of fMRI studies of ASpA were conducted in which subjects performed a one-back task in which they attended to one of two spatially separated streams. Attention modulated blood oxygenation level-dependent (BOLD) activity in multiple areas in the prefrontal, temporal, and parietal cortex, including non-visuotopic intraparietal sulcus (IPS), but not the visuotopic maps in IPS. No spatial bias was detected in any cortical area using standard univariate analysis; however, MVPA revealed that activation patterns in a number of areas, including the auditory cortex, predicted the attended direction. Furthermore, we explored how cognitive task demands and the sensory modality of the inputs influenced activity with a visual one-back task and a visual multiple object tracking (MOT) task. Activity from the visual and auditory one-back tasks overlapped along the fundus of IPS and lateral prefrontal cortex (lPFC). However, there was minimal overlap of activity in the lPFC between the visual MOT task and the two one-back tasks. Finally, we endeavored to identify visual and auditory networks using rsFC. We identified a dorsal visual attention network reliably within individual subjects using visuotopic seeds. Using auditory seeds, we found a prefrontal area nested between segments of the dorsal visual attention network. These findings mark fundamental progress towards elucidating the cortical network controlling ASpA. Our results suggest that similar lPFC structures support both ASpA and its visual counterpart during a spatial one-back task, but that ASpA does not drive visuotopic IPS in the parietal cortex. Furthermore, rsFC reveals that visual and auditory seed regions are functionally connected with non-overlapping lPFC regions, possibly reflecting spatial and temporal cognitive processing biases, respectively. While we find no evidence for a spatiotopic map, the auditory cortex is sensitive to direction of attention in its patterns of activation