Beyond Localization Parametric Manipulation of and Dynamic Interactions in Working Memory

Black Belt Memory

10x Your Memory Power

Get Instant Access

Brain imaging studies of working memory and divided attention paint a fairly consistent picture. On-line maintenance of information in support of cognitive performance, whether in specific processing domains such as verbal, object, or spatial processing or more generally as in executive processes, activates specific neural circuits in the frontal cortex. The first generation of these imaging studies was primarily concerned with localization. As such, they have value in pointing to the cortical regions putatively associated with resources and, therefore, in describing the cortical expression of mental workload. However, localization per se is only partially informative with respect to the functional aspects of the neural mechanisms underlying mental workload. Localization is only a first step in a deeper understanding of such mechanisms. More important would be evidence that shows that neural activity in specific cortical regions exhibits systematic variation with task or subject variables or, as previously discussed, with exogenous or endogenous drivers of mental workload.

An early example of such parametric manipulation of workload was reported in an fMRI study of language processing in which the computational complexity of comprehending different sentences was varied. Sentence comprehension requires working memory processing resources first to compute the comprehension operations (word integration, syntactic and semantic relations, etc.) and second to maintain these representations actively during the period of processing, which can be considerable for a long, complex sentence. The study examined sentences that varied in structural complexity by comparing simple conjoined sentences, including subject and object clauses. Important controls were the number ofwords, word frequency, and the complexity of individual words, so that the sentences differed only in overall structural complexity. Left hemisphere areas associated with sentence processing were analyzed for the volume of activation in the superior temporal cortex (Wernicke's area) and inferior frontal gyrus (Broca's area). Activation volumes increased with sentence structural complexity in both cortical areas and in a graded manner. Thus, the quantitative extent of neural activity evoked by sentence processing was directly related to the associated computational demand. These findings suggest that manipulation of processing complexity unconfounded with other factors results in increased brain activation in a manner consistent with the allocation of specific processing resources.

Similar parametric studies of working memory have also been reported. Such studies have shown that, as mental workload increases with an exogeneous driver, the cortical areas associated with the type of process manipulated exhibit corresponding increases in activity. PET studies of the n-back task have shown monotonic increases in activity in a number of areas, including the DLPFC, VLPFC, posterior parietal cortex, and Broca's area, as the size of n or the amount of information retained in working memory increases. In addition, monotonic deactivation with increasing verbal working memory demands has also been reported for certain brain regions. Although no clear consensus has been reached as to the relevance of these deactivations to the functioning of working memory systems, the consistent parametric activation of process-specific "slave''modules, coupled with parametric deactivation of other brain regions, suggests a reallocation of mental resources to compensate for increased task demands.

The working memory and dual-task studies indicate first that activation of specific frontal cortical regions can be identified as cortical signs of processing resources associated with task performance. Second, activity in these regions increases, or the volume of activated tissue increases, in response to demand for greater processing resources with exogenous drivers of workload. Such increases probably also reflect endogenous sources of workload, but exogenous and endogenous effects have not yet been differentiated by brain imaging studies. An important additional question is whether such cortical signs of processing resources are necessary for efficient task performance. For example, is the working memory "signature" in the frontal cortex sustained across time when a target object that has to be kept in mind occurs repeatedly? Specifically, how can processing resources be deployed to sustain attention to a specific object over time in the presence of distractors? fMRI studies have shown that neural responses are changed as stimuli become more familiar or as new associations are learned. However, such alterations in activation cannot distinguish a specific object that is currently the focus of attention from equally familiar distracters that should be ignored.

The neural responses that mediate repeated target identification during working memory were examined in an fMRI study. On a given trial subjects were required to study a sample object (a face) and then to identify it repeatedly with a button press in a stream of objects containing distracters (there were a total of 13 items plus the sample in a single trial). Subjects were familiarized with every object, and a specific object was used as a target on some trials and as a distracter on other trials. Distracters were also repeated within a trial. Thus, neither familiarity nor simple repetition could be used to signal the presence of a target, but rather a representation of the target on that trial had to be kept in mind for the duration of the trial. Under these conditions, enhanced neural responses were found primarily in bilateral inferior-middle frontal and left insular cortices (see Fig. 6). These responses signaled the identification and maintenance in working memory of the currently attended target item. Targets were also responded to faster with each repetition within a trial. The neural response to targets remained constant with repetition in frontal-insular areas, whereas those to targets or distracters in posterior regions declined (see Fig. 7). Activity associated with repeated distracters in frontal regions also remained at a constant but low level throughout the trial. Thus, the enhanced frontal neural responses and the speeded response to targets signaled the maintenance of the target object in working memory, supporting a role for the active maintenance component of working memory in the selection of targets among distracters. Effective selective attention requires that the neural response to a target stimulus is enhanced and maintained during the period of time the target remains behaviorally relevant. The sustained enhancement of frontal responses across target repetitions might reflect such top-down allocation of attentional resources to target detection.

Figure 6 fMRI activation patterns showing enhanced neural responses to targets. Labels 4, 5, and 6 indicate frontal-insular areas, the supplemental motor area, and the left motor cortex, respectively. Reprinted with permission from Jiang et al. (2000). Complementary neural mechanisms for tracking items in human working memory. Science 287,643-646. Copyright 2001 American Association for the Advancement of Science.

Figure 6 fMRI activation patterns showing enhanced neural responses to targets. Labels 4, 5, and 6 indicate frontal-insular areas, the supplemental motor area, and the left motor cortex, respectively. Reprinted with permission from Jiang et al. (2000). Complementary neural mechanisms for tracking items in human working memory. Science 287,643-646. Copyright 2001 American Association for the Advancement of Science.

Was this article helpful?

0 0
Unraveling Alzheimers Disease

Unraveling Alzheimers Disease

I leave absolutely nothing out! Everything that I learned about Alzheimer’s I share with you. This is the most comprehensive report on Alzheimer’s you will ever read. No stone is left unturned in this comprehensive report.

Get My Free Ebook


Post a comment