Abstract
During visual search, the working memory (WM) representation of the search target guides attention to matching items in the visual scene. However, we can hold multiple items in WM. Do all these items guide attention at the same time? Using a new functional magnetic resonance imaging visual search paradigm, we found that items in WM can attain two different states that influence activity in extrastriate visual cortex in opposite directions: whereas the target item in WM enhanced processing of matching visual input, other "accessory" items in memory suppressed activity. These results imply that the representation of task-relevant and (currently) task-irrelevant representations in WM differs, revealing new insights into the organization of human visual WM. The suppressive influence of irrelevant WM items may complement the attention-guiding influence of task-relevant WM items, helping us to focus on task-relevant information without getting distracted by irrelevant memory content.
Introduction
Working memory (WM) is often described as a "search template." When you look for your keys, your brain maintains a representation of those keys that primes your visual system to spot them. However, humans can hold multiple items in WM simultaneously—typically up to four. This raises a critical question: If all items in WM guided our attention at once, wouldn't we be constantly distracted by the things we are simply "holding for later"?
This study by Peters et al. explores the neural dynamics of this problem, investigating whether the brain distinguishes between current search targets and "accessory" items held for future use. The researchers hypothesized that the brain employs different states for these items, specifically within the extrastriate cortex—regions specialized for processing complex objects like faces and houses.
Experimental Design and Methods
To test this, the researchers developed a "memory-loaded temporal search task" performed during functional Magnetic Resonance Imaging (fMRI).
- Encoding Phase: Participants were shown two images (e.g., a face and a house). A cue indicated which was the first Search Target (ST1) and which was the Memory Item (MI1) to be stored for a subsequent task.
- Search Phase: Participants viewed a rapid-fire stream of superimposed face-house images. They were tasked with identifying ST1 while ignoring the accessory item (MI1), which also appeared in the stream as a distractor.
- The Switch: After the first search, the roles reversed: MI1 became the new target (ST2), and the previous target became the accessory item (MI2).
- Neural Mapping: Functional localizers identified the Fusiform Face Area (FFA) and the Parahippocampal Place Area (PPA) for each participant to measure how these specific regions responded to target vs. accessory representations.
Key Insights and Neural Signatures
The study revealed that targets and accessory items have strikingly opposite effects on brain activity. When a participant searched for a house, activity in the PPA (the house-processing area) increased. However, when a house was merely being held in memory as an accessory item for a later task, activity in the PPA actually dropped below baseline levels.
This suppression suggests that accessory items are not just "dormant" in memory; they are actively managed. By suppressing the cortical regions tuned to the accessory item, the brain creates a "template for rejection." This mechanism likely prevents task-irrelevant memories from accidentally capturing attention during the current search.
Results
- Sustained Activity in the PPA: Searching for a house (ST1) significantly increased PPA activity compared to searching for a face ($F_{(1,8)} = 64.8, p < 0.00005$). Conversely, holding a house as an accessory item (MI1) significantly decreased PPA activity ($F_{(1,8)} = 15.9, p < 0.005$).
- FFA Patterns: The FFA showed a mirrored pattern for faces: enhancement for targets ($F_{(1,8)} = 150.2, p < 0.000002$) and suppression for accessory items ($F_{(1,8)} = 12.6, p < 0.01$).
- Match Enhancement: When the actual target appeared in the visual stream, it triggered a transient spike in activity ("match enhancement"). In contrast, the accessory item was treated by the brain exactly like a random distractor, triggering no such boost.
Discussion and Conclusion
The findings demonstrate that working memory is a dynamic filter rather than a passive storage bin. The brain can rapidly flip the state of a memory representation from suppressed (accessory) to enhanced (target) as task demands change. This dual-state system allows us to maintain complex information for the future without compromising our ability to focus on the present.
While this study identifies the effects of these states in the visual cortex, future research is needed to determine the exact source of these signals. The authors suggest the prefrontal cortex (PFC) may be responsible for partitioning these memory states, but the precise inhibitory mechanisms remain a subject for further investigation.
Reference: Peters, J. C., Roelfsema, P. R., & Goebel, R. (2012). Task-Relevant and Accessory Items in Working Memory Have Opposite Effects on Activity in Extrastriate Cortex. The Journal of Neuroscience, 32(47), 17003–17011.
Summary updated