Erich J. Greene

Refreshing: a minimal executive function.

Executive functions include processes by which important information (e.g., words, objects, task goals, contextual information) generated via perception or thought can be foregrounded and thereby influence current and subsequent processing. One simple executive process that has the effect of foregrounding information is refreshing--thinking briefly of a just-activated representation. Previous studies (e.g., Johnson et al., 2005) identified refresh-related activity in several areas of left prefrontal cortex (PFC). To further specify the respective functions of these PFC areas in refreshing, in Experiment 1, healthy young adult participants were randomly cued to think of a just previously seen word (refresh) or cued to press a button (act). Compared to simply reading a word, refresh and act conditions resulted in similar levels of activity in left lateral anterior PFC but only refreshing resulted in greater activity in left dorsolateral PFC. In Experiment 2, refreshing was contrasted with a minimal phonological rehearsal condition. Refreshing was associated with activity in left dorsolateral PFC and rehearsing with activity in left ventrolateral PFC. In both experiments, correlations of activity among brain areas suggest different functional connectivity for these processes. Together, these findings provide evidence that (1) left lateral anterior PFC is associated with initiating a non-automatic process, (2) left dorsolateral PFC is associated with foregrounding a specific mental representation, and (3) refreshing and rehearsing are neurally distinguishable processes.

Emotional Arousal Can Impair Feature Binding in Working Memory

To investigate whether emotional arousal affects memorial feature binding, we had participants complete a short-term source-monitoring taskremembering the locations of four different pictures over a brief delay. On each trial, the four pictures were all either high arousal, medium arousal, or low arousal. Memory for picture-location conjunctions decreased as arousal increased. In addition, source memory for the location of negative pictures was worse among participants with higher depression scores. Two subsequent functional magnetic resonance imaging experiments showed that relative to low-arousal trials, high- and medium-arousal trials resulted in greater activity in areas associated with visual processing (fusiform gyrus, middle temporal gyrus/middle occipital gyrus, lingual gyrus) and less activity in superior precentral gyrus and the precentral-superior temporal intersect. These findings suggest that arousal (and perhaps negative valence for depressed people) recruits attention to items thereby disrupting working memory processes that help bind features together.

Using fMRI to investigate a component process of reflection: Prefrontal correlates of refreshing a just-activated representation

Using fMRI, we investigated the functional organization of prefrontal cortex (PFC) as participants briefly thought of a single just-experienced item (i.e., refreshed an active representation). The results of six studies, and a meta-analysis including previous studies, identified regions in left dorsolateral, anterior, and ventrolateral PFC associated in varying degrees with refreshing different types of information (visual and auditory words, drawings, patterns, people, places, or locations). In addition, activity increased in anterior cingulate with selection demands and in orbitofrontal cortex when a nonselected item was emotionally salient, consistent with a role for these areas in cognitive control (e.g., overcoming “mental rubbernecking”). We also found evidence that presenting emotional information disrupted an anterior component of the refresh circuit. We suggest that refreshing accounts for some neural activity observed in more complex tasks, such as working memory, long-term memory, and problem solving, and that its disruption (e.g., from aging or emotion) could have a broad impact.

Neuroimaging a single thought: dorsolateral PFC activity associated with refreshing just-activated information.

Neuroimaging studies of human working memory (WM) show conflicting results regarding whether dorsolateral prefrontal cortex (PFC) contributes to maintaining information in consciousness or is recruited primarily when information must be manipulated. Using functional magnetic resonance imaging (fMRI), we looked at a minimal maintenance process--thinking back to a single, just-seen stimulus (refreshing). We found greater activity in left dorsolateral PFC (BA9) when participants refreshed a word compared to reading a word once or a second time. Furthermore, recognition memory was subsequently more accurate and faster for items that had been refreshed, demonstrating that a single thought that maintains activation can have consequences for long-term memory. Our fMRI results call into question any class of models of the functional organization of PFC and WM that associates simple and/or maintenance processes only with ventrolateral PFC or that associates dorsolateral PFC only with more complex processes such as manipulation.

Prefrontal Cortex Activity Associated with Source Monitoring in a Working Memory Task

Using functional magnetic resonance imaging (fMRI), we investigated prefrontal cortex (PFC) activity during remembering specific source information (format, location judgments) versus remembering that could be based on undifferentiated information, such as familiarity (old/new recognition ON, recency judgments). A working memory (WM) paradigm with an immediate test yielded greater activation in the lateral PFC for format and location source memory (SM) tasks than ON recognition; this SM-related activity was left lateralized. The same regions of PFC were recruited in Experiment 2 when information was tested immediately and after a filled delay. Substituting recency for location judgments (Experiment 3) resulted in an overall shift in task context that produced greater right PFC activity associated with ON and recency tasks compared to the format task, in addition to left SM-related activity. These data extend to WM previous findings from long-term memory (LTM) indicating that the left and right PFC may be differentially involved in memory attributions depending on the specificity of information evaluated. The findings also provide evidence for the continuity of evaluative processes recruited in WM and LTM.

An Age-Related Deficit in Prefrontal Cortical Function Associated With Refreshing Information

Older adults are slower than young adults to think of an item they just saw, that is, to engage or execute (or both) the simple reflective operation of refreshing just-activated information. In addition, they derive less long-term memory benefit from refreshing information. Using functional magnetic resonance imaging (fMRI), we found that relative to young adults, older adults showed reduced refresh-related activity in an area of dorsolateral prefrontal cortex (left middle frontal gyrus, Brodmanns Area 9), but not in other refresh-related areas. This provides strong evidence that a frontal component of the circuit that subserves this basic cognitive process is especially vulnerable to aging. Such a refresh deficit could contribute to poorer performance of older than young adults on a wide range of cognitive tasks.

Neuroimaging Evidence for Agenda-Dependent Monitoring of Different Features During Short-Term Source Memory Tests

A short-term source monitoring procedure with functional magnetic resonance imaging assessed neural activity when participants made judgments about the format of 1 of 4 studied items (picture, word), the encoding task performed (cost, place), or whether an item was old or new. The results support findings from long-term memory studies showing that left anterior ventrolateral prefrontal cortex (PFC) is engaged when people make source attributions about reflectively generated information (cognitive operations, conceptual features). The findings also point to a role for right lateral PFC in attention to perceptual features and/or familiarity in making source decisions. Activity in posterior regions also differed depending on what was evaluated. These results provide neuroimaging evidence for theoretical approaches emphasizing that agendas influence which features are monitored during remembering (e.g., M. K. Johnson, S. Hashtroudi, & D. S. Lindsay, 1993). They also support the hypothesis that some of the activity in left lateral PFC and posterior regions associated with remembering specific information is not unique to long-term memory but rather is associated with agenda-driven source monitoring processes common to working memory and long-term memory.

Refreshing one of several active representations: Behavioral and functional magnetic resonance imaging differences between young and older adults

We explored age-related differences in executive function during selection of a target from among active representations. Refreshing (thinking briefly of a just-activated representation) is an executive process that foregrounds a target relative to other active representations. In a behavioral study, participants saw one or three words, then saw a cue to refresh one of the words, saw one word again and read it, or read a new word. Increasing the number of active representations increased response times (RTs) only in the refresh condition for young adults but increased RTs equally in all conditions for older adults, suggesting that they experienced interference from activated irrelevant information during perception and reflection. Consistent with this interpretation, in a functional magnetic resonance imaging study, young adults showed two areas of the left dorsolateral frontal cortex and a medial area of frontal cortex, including anterior cingulate, that were relatively more sensitive to number of active representations during refresh than read trials; for older adults these areas were equally sensitive to number of active items for refresh and read trials. Young and older adults showed activity associated with refreshing on trials requiring selection in left mid-ventral frontal cortex (an area associated with selection from active representations); older adults also showed activity in left anterior ventral frontal cortex (an area associated with controlled semantic activation). Our results support the hypothesis of an age-related decrease in ability to gate out activated but currently irrelevant information, and are consistent with a dissociation of function between eft mid-ventral and left anterior ventral frontal cortex.

Age-related differences in agenda-driven monitoring of format and task information

Age-related source memory deficits may arise, in part, from changes in the agenda-driven processes that control what features of events are relevant during remembering. Using fMRI, we compared young and older adults on tests assessing source memory for format (picture, word) or encoding task (self-, other-referential), as well as on old-new recognition. Behaviorally, relative to old-new recognition, older adults showed disproportionate and equivalent deficits on both source tests compared to young adults. At encoding, both age groups showed expected activation associated with format in posterior visual processing areas, and with task in medial prefrontal cortex. At test, the groups showed similar selective, agenda-related activity in these representational areas. There were, however, marked age differences in the activity of control regions in lateral and medial prefrontal cortex and lateral parietal cortex. Results of correlation analyses were consistent with the idea that young adults had greater trial-by-trial agenda-driven modulation of activity (i.e., greater selectivity) than did older adults in representational regions. Thus, under selective remembering conditions where older adults showed clear differential regional activity in representational areas depending on type of test, they also showed evidence of disrupted frontal and parietal function and reduced item-by-item modulation of test-appropriate features. This pattern of results is consistent with an age-related deficit in the engagement of selective reflective attention.

Brain Mechanisms Underlying Reality Monitoring for Heard and Imagined Words

Using functional MRI, we investigated reality monitoring for auditory information. During scanning, healthy young adults heard words in another person’s voice and imagined hearing other words in that same voice. Later, outside the scanner, participants judged words as “heard,” “imagined,” or “new.” An area of left middle frontal gyrus (Brodmann’s area, or BA, 6) was more active at encoding for imagined items subsequently correctly called “imagined” than for items incorrectly called “heard.” An area of left inferior frontal gyrus (BA 45, 44) was more active at encoding for items subsequently called “heard” than “imagined,” regardless of the actual source of the item. Scores on an Auditory Hallucination Experience Scale were positively related to activity in superior temporal gyrus (BA 22) for imagined words incorrectly called “heard.” We suggest that activity in these areas reflects cognitive operations information (middle frontal gyrus) and semantic and perceptual detail (inferior frontal gyrus and superior temporal gyrus, respectively) used to make reality-monitoring attributions.