Delay-period activity in frontal, parietal, and occipital cortex tracks different attractor dynamics in visual working memory
Yu Q, Panichello MF, Cai Y, Postle BR, and Buschman TJ
Under Review. Available on bioRxiv.
One important neural hallmark of working memory is persistent elevated delay-period activity in frontal and parietal cortex. In human fMRI, delay-period BOLD activity in frontal and parietal cortex increases monotonically with memory load and asymptotes at an individual’s capacity. Previous work has demonstrated that frontal and parietal delay-period activity correlates with the decline in behavioral memory precision observed with increasing memory load. However, because memory precision can be influenced by a variety of factors, it remains unclear what cognitive processes underlie persistent activity in frontal and parietal cortex. Recent psychophysical work has shown that attractor dynamics bias memory representations toward a few stable representations and reduce the effects of internal noise. From this perspective, imprecision in memory results from both drift towards stable attractor states and random diffusion. Here we asked whether delay-period BOLD activity in frontal and parietal cortex might be explained, in part, by these attractor dynamics. We analyzed data from an existing experiment in which subjects performed delayed recall for line orientation, at different loads, during fMRI scanning. We modeled subjects’ behavior using a discrete attractor model, and calculated within-subject correlation between frontal and parietal delay-period activity and estimated sources of memory error (drift and diffusion). We found that although increases in frontal and parietal activity were associated with increases in both diffusion and drift, diffusion explained the most variance in frontal and parietal delay-period activity. In comparison, a subsequent whole-brain regression analysis showed that drift rather than diffusion explained the most variance in delay-period activity in lateral occipital cortex. These results provide a new interpretation for the function of frontal, parietal, and occipital delay-period activity in working memory.