Neural Systems Underlying the Implementation of Working Memory Removal Operations

Recent research demonstrated that analyses of neuroimaging data can verify the removal of information from working memory (WM) when it is replaced, suppressed, and cleared completely in the mind compared to when that information is maintained In addition, univariate activation across the brain while someone performed these operations revealed activation patterns that resembled those of functionally communicating brain networks. Currently, less is known about the similarities and differences of the activation patterns across these functionally related brain regions for each of the distinct WM operations. Here, we address these gaps in a sample of 55 humans (17 male; age, M = 23.52, SD = 4.93) and identify the key functional brain networks underlying the WM operations. We also quantify the uniqueness of each network’s activation pattern for each operation. Using bagging enhanced Leiden community detection, we uncovered four reliable, functional networks that showed dissociable activation patterns across the WM operations. The Frontoparietal Control Network was identified and elicited unique activation patterns for each of the four operations. The uniqueness of these patterns suggests that this network has an important role in allowing someone to differentiate and perform each operation. The Default Mode Network was also identified and differentiated suppress and clear - which require disengagement in WM - from each other and from maintain and replace - which require deliberate engagement with information in WM. The activation patterns observed across our identified networks inform us of the prototypical functional communication across the brain when someone purposely manipulates information in WM. Importantly, these patterns allow future studies to assess if functional communication dysfunction relates to an inability to remove information in WM with individual differences and clinical populations.