Abstract
Human caudolateral frontal cortex (LFC) is often characterized as domain-general or multiple-demand, due to its recruitment in a wide range of cognitive tasks (e.g. Duncan 2010, Fedorenko 2013). However, our laboratory has recently used fMRI to demonstrate that direct contrast of either visual and auditory attention (Michalka 2015) or visual and auditory working memory identifies four discrete sensory-biased attention regions in LFC. Two visual-biased areas, superior and inferior precentral sulcus (sPCS and iPCS) are interleaved with two auditory-biased areas, the transverse gyrus bridging precentral sulcus (tgPCS) and caudal inferior frontal sulcus (cIFS). These four structures can be identified bilaterally in more than 75% of subjects (n = 17) and their locations are stable across tasks (attention or working memory) and over time (>1 year; n = 7). Using fMRI BOLD signal change in visual and auditory working memory, and MVPA classifier accuracy, we compared these four LFC regions to (i) posterior sensory-specific structures in temporal, occipital, and parietal cortex, and (ii) general attention structures including anterior insula and dorsal anterior cingulate. Good classifier performance generally co-occurred with high BOLD activity, which we call "good coding." As expected, posterior structures were modality biased, showing good coding either for auditory or for visual tasks. In contrast, anterior insula and dorsal anterior cingulate showed domain generality, with good coding for both tasks. LFC structures exhibited asymmetry between modalities: visual-biased sPCS and iPCS demonstrated good coding in both visual and auditory working memory, but auditory-biased tgPCS and cIFS only showed good coding in auditory working memory. That is, visual-biased areas demonstrated the domain-general properties that have previously been observed in LFC, while auditory-biased areas were more specialized. These results extend our previous understanding of frontal contributions to attention and working memory, characterizing two distinct, sensory-biased attention networks that play quite different roles in human cognition.
Meeting abstract presented at VSS 2016