Modulation of working-memory maintenance by directed attention
Introduction
The ability to maintain task-relevant representations in working memory (WM) is of critical importance for many cognitive tasks. Adaptive behaviour relies on flexible access and manipulation of these representations, and also requires prioritizing of particular items in WM in light of changing task goals and expectations. Recent studies have demonstrated that such prioritization can be controlled flexibly by means of directing attention to the internal representation of an item stored in WM (Bays and Husain, 2008, Griffin and Nobre, 2003, Landman et al., 2003).
Many models of WM propose a close relationship between WM and attention (Awh and Jonides, 2001, Courtney, 2004, Cowan, 1988, Curtis and D’Esposito, 2003, Fuster, 2000, Jonides et al., 2008, McElree, 2006, Passingham and Sakai, 2004, Postle, 2006). On the neural level, a prevailing account suggests that top-down signals from prefrontal areas bias activity in posterior regions to maintain, monitor, and/or manipulate information in WM (Awh and Jonides, 2001, Curtis and D’Esposito, 2003, Harrison and Tong, 2009), in close analogy to attentional top-down signals biasing activity in visual areas during perception (Desimone and Duncan, 1995, Kastner and Ungerleider, 2001, Treue, 2003). Directly supporting this notion, a recent study demonstrated that the activity during maintenance in sensory areas was modulated as a consequence of attentional orienting towards the type of information represented in this area. These effects were linked to improved retrieval of attended relative to unattended items, and suggest a direct influence of attentional orienting on the maintenance of the items themselves (Lepsien & Nobre, 2007).
However, other brain areas beyond the sensory cortices may also be involved in representing task-relevant items in WM. Representations in higher order areas may be expected to be less tied to the specific perceptual nature of the stimuli and to be flexibly modulated according to task goals. Key candidate regions are the posterior parietal (PPC) and the prefrontal cortices (PFC), which have both been shown to be sensitive to WM load (Linden et al., 2003, Rypma et al., 2002, Todd and Marois, 2004, Xu and Chun, 2006). However, it can be difficult to tease apart whether these areas play a role in WM maintenance or in complementary selective attention or executive-control functions. For example, activity in the PPC is sustained during working-memory delays (Chafee and Goldman-Rakic, 1998, Courtney et al., 1998), and is correlated with the amount of information stored in WM (Todd and Marois, 2004, Xu and Chun, 2006) and with individual differences in WM capacity (Todd & Marois, 2005). This has led to some authors to ascribe to the PPC a direct role in WM maintenance (e.g., Todd & Marois, 2004). However, other authors (Magen, Emmanouil, McMains, Kastner, & Treisman, 2009) have suggested a more attentional role for the PPC and proposed that the load-related activity in PPC reflects different attentional demands on rehearsal of information in WM rather than storage. Similarly, the PFC has been implicated in a wide range of WM-related processes, such as maintenance of task-relevant information (Courtney, 2004, Funahashi et al., 1989, Fuster and Alexander, 1971, Fuster, 1973, Goldman-Rakic, 1987), manipulation of information in WM (Petrides, 1994, D’Esposito et al., 1998, Owen et al., 1999), or other executive-control functions like monitoring multiple mnemonic representations (Petrides, 2000). PFC has also been suggested to support WM through attention-related functions. For example, Rowe, Toni, Josephs, Frackowiak, and Passingham (2000) have suggested that area 46 is associated with the selection of relevant items in WM via attention (Bledowski et al., 2009, Lebedev et al., 2004), and other authors emphasized the role of this region in protecting WM representations against distraction (Knight et al., 1999, Miller et al., 1996, Sakai et al., 2002; also referred to as ‘active maintenance’, see Miller & Cohen, 2001).
The present experiment investigated the extent to which brain activity in perceptual and high-level brain areas involved in maintaining items in WM is flexibly modulated by changes in the task relevance of the memoranda. Attentional orienting to specific categories of stimuli being maintained in WM (faces or scenes) was manipulated by presenting retro-cues during the maintenance period, which indicated the category of stimuli that would be relevant to perform a subsequent comparison judgement to a probe item (Griffin and Nobre, 2003, Lepsien and Nobre, 2007). In each trial, participants viewed four items to be held in WM, from a variable and complementary (1–3) number of faces and scenes. Upon the presentation of a retro-cue indicating the relevant stimulus category, the effective load of task-relevant memoranda was reduced to 1–3. If it is possible to adapt WM maintenance flexibly by orienting attention to the relevant items, the load-related activity before and after attentional allocation should differ, and should be expressed in brain areas coding for the effective load of relevant items in WM (‘effective WM load’).
In addition, investigating the interaction between attention to a stimulus category and different levels of WM load for each category provides an elegant way to demonstrate more directly the effect of attentional orienting on WM maintenance. Lepsien and Nobre (2007) previously demonstrated that orienting attention to one stimulus category modulates stimulus-specific activity during the maintenance period. However, it is increasingly clear that the mere expectation of a probe stimulus of the relevant stimulus category can also modulate activity in higher order visual areas processing these stimuli (Esterman and Yantis, 2010, Puri et al., 2009, Stokes et al., 2009). Accordingly, it could be argued that the retro-cue manipulation used by Lepsien and Nobre (2007) could also have caused anticipation of an upcoming probe in addition to or instead of modulation of WM maintenance. If, however, attentional orienting were to result in graded changes of activity that tracked the effective WM load, this would be difficult to explain by mere anticipation.
Section snippets
Participants
Eighteen healthy volunteers took part in two experimental sessions (aged 18–33 (mean 22.4), six female, normal or corrected-to-normal vision). Data sets of six participants were partly lost due to technical failures of the fMRI scanner, thus only the data of twelve participants were entered into the analysis. Participants gave written informed consents. The study protocol had approval from the Oxfordshire Research Ethics Committee (06/Q1606/70).
Stimuli
The stimuli comprised 120 faces and 120 scenes
Behavioural results
The behavioural results are summarized in Fig. 2. To investigate the effects of the retro-cue on the performance in the delayed-match-to-sample task, mean reaction time (RT) and accuracy (AC) were analysed with repeated-measures analyses of variance (ANOVAs) with the factors cue (face, scene) and array type (1F/3S, 2F/2S, 3F/1S). Only trials with correct responses were entered into the RT analysis. Trials with RTs deviating more than 3 standard deviations from the condition mean were considered
Discussion
The present experiment probed the flexibility of WM representations, and investigated if activity in perceptual and high-level brain areas involved in WM maintenance can be modulated according to the task-relevance of the items held on-line. Cueing attention to a subset of items in WM, tagging them as relevant for subsequent probe comparison, resulted in an enhancement in WM performance, which increased linearly as the number of items in the focus of attention decreased. On the neural level,
Conclusion
The present study provides behavioural and neural evidence that subsets of items held on-line in WM can be highlighted as task-relevant by orienting attention towards them. A network of prefrontal and parietal areas directly reflected the changes of the effective WM load in a systematic fashion, suggesting that these areas directly contribute to the (active) maintenance of information in WM.
Acknowledgements
We would like to thank the OCMR for the help with data acquisition. The research was funded by 21st Century Research Award: Bridging Brain, Mind and Behaviour to Anna C. Nobre by the James S. McDonnell Foundation.
References (83)
- et al.
Unified segmentation
NeuroImage
(2005) - et al.
Overlapping mechanisms of attention and spatial working memory
Trends in Cognitive Sciences
(2001) - et al.
The role of the inferior frontal junction area in cognitive control
Trends in Cognitive Sciences
(2005) - et al.
Persistent activity in the prefrontal cortex during working memory
Trends in Cognitive Sciences
(2003) - et al.
Cognitive control in the posterior frontolateral cortex: evidence from common activations in task coordination, interference control, and working memory
NeuroImage
(2004) - et al.
Functional MRI studies of spatial and non-spatial working memory
Cognitive Brain Research
(1998) - et al.
Activity in fusiform face area modulated as a function of working memory load
Brain Research. Cognitive Brain Research
(2001) - et al.
Event-related fMRI: characterizing differential responses
Neuroimage
(1998) - et al.
How many subjects constitute a study?
Neuroimage
(1999) - et al.
Thresholding of statistical maps in functional neuroimaging using the False Discovery Rate
Neuroimage
(2002)
A brief thought can modulate activity in extrastriate visual areas: top-down effects of refreshing just-seen visual stimuli
Neuroimage
The neural basis of biased competition in human visual cortex
Neuropsychologia
Prefrontal cortex regulates inhibition and excitation in distributed neural networks
Acta Psychologica
Large capacity storage of integrated objects before change blindness
Vision Research
Load response functions in the human spatial working memory circuit during location memory updating
Neuroimage
Cortical capacity constraints for visual working memory: dissociation of fMRI load effects in a fronto-parietal network
Neuroimage
Attentional demands predict short-term memory load response in posterior parietal cortex
Neuropsychologia
The hazards of time
Current Opinion in Neurobiology
Variational Bayesian Inference for fMRI time series
NeuroImage
Frontal lobes and behaviour
Current Opinion in Neurobiology
Working memory as an emergent property of the mind and brain
Neuroscience
The where and how of attention-based rehearsal in spatial working memory
Brain Research. Cognitive Brain Research
Seeking the neural substrates of visual working memory storage
Cortex
Category expectation modulates baseline and stimulus-evoked activity in human inferotemporal cortex
Brain Research
Functional topography of working memory for face or voice identity
Neuroimage
Category-specific modulation of inferior temporal activity during working memory encoding and maintenance
Brain Research. Cognitive Brain Research
Neural system for updating object working memory from different sources: sensory stimuli or long-term memory
Neuroimage
Working memory for location and time: activity in prefrontal area 46 relates to selection rather than maintenance in memory
Neuroimage
Functional topography of a distributed neural system for spatial and nonspatial information maintenance in working memory
Neuropsychologia
Visual attention: the where, what, how and why of saliency
Current Opinion in Neurobiology
Neural components of topographical representation
Proceedings of the National Academy of Sciences of the United States of America
Face recognition in human extrastriate cortex
Journal of Neurophysiology
Dynamic shifts of limited working memory resources in human vision
Science
What “works” in working memory? Separate systems for selection and updating of critical information
Journal of Neuroscience
Selection for cognitive control: a functional magnetic resonance imaging study on the selection of task-relevant information
Journal of Neuroscience
Matching patterns of activity in primate prefrontal area 8a and parietal area 7ip neurons during a spatial working memory task
Journal of Neurophysiology
Attention and cognitive control as emergent properties of information representation in working memory
Cognitive, Affective & Behavioral Neuroscience
An area specialized for spatial working memory in human frontal cortex
Science
Evolving conceptions of memory storage, selective attention, and their mutual constraints within the human information-processing system
Psychological Bulletin
Neural mechanisms of selective visual attention
Annual Review of Neuroscience
Dissecting contributions of prefrontal cortex and fusiform face area to face working memory
Journal of Cognitive Neuroscience
Cited by (86)
Neurobiological characteristics of individuals who have committed sexual offenses: A systematic review
2023, Aggression and Violent BehaviorBehavioural and neurophysiological differences in working memory function of depressed patients and healthy controls
2021, Journal of Affective DisordersThe Relation Between Attention and Memory
2024, Annual Review of Psychology