Abstract: Individuals prioritize certain important information based on task demands, which facilitates the processing of prioritized Visual Working Memory (VWM) items. However, the role of sustained visual attention in VWM prioritization remains unclear. Some evidence suggests that visual attention is involved in VWM prioritization, while other evidence indicates a dissociation between visual attention and VWM prioritization. Considering the variability in the number of items across VWM tasks in previous studies, this study examined whether the role of sustained visual attention in VWM prioritization was modulated by working memory resources. When working memory resources were insufficient, sustained visual attention was required to maintain the prioritized item, while it was not required when working memory resources were sufficient. Additionally, we investigated the neural basis of VWM prioritization using the simultaneous acquisition technique of event-related potentials and event-related optical signals (ERP-EROS).
Our study included five experiments, with 24, 23, 24, 24, and 17 Chinese participants in Experiments 1a, 1b, 1c, 2, and 3, respectively. In Experiments 1 and 2, a recall report paradigm was used. Participants performed VWM tasks with one prioritized item in Experiment 1 and two prioritized items in Experiment 2. Additionally, across Experiments 1 and 2, participants performed a visual attention task during the VWM maintenance phase. In Experiment 3, we employed a change detection paradigm. Participants completed VWM tasks with no prioritized item and with one prioritized item, and their brain activity under these different prioritization conditions was recorded.
In Experiment 1, participants were instructed to prioritize one VWM item. Visual attention was consumed by single visual attention task (Exp. 1a) and double visual attention task (Exp. 1b). Furthermore, visual attention was consumed by manipulating the durations of task presentation (early vs. late; Exp. 1c). The results consistently indicated that the memory accuracy of non-prioritized items was disrupted by the visual attention task, while the prioritized item remained unaffected. In Experiment 2, participants were required to prioritize two VWM items and performed a single visual attention task. We found that both prioritized and non-prioritized items were disrupted by the visual attention task. Results from Experiment 3 revealed that compared to neutral cues without VWM prioritization, informative cues indicating prioritization of one item elicited a larger late positive component (LPC) and negative slow wave (NSW), as well as higher activation in the prefrontal and occipital cortices during the maintenance phase.
The results indicated that the role of visual attention in VWM prioritization was modulated by working memory resources. When working memory resources were insufficient, sustained visual attention was required to maintain the prioritized item, whereas it was not necessary when working memory resources were sufficient. The neural basis of VWM prioritization likely involves the activation of the frontal and occipital cortices during the maintenance phase, as well as the allocation of more working memory resources to enhance the stability of the representation of the prioritized item. Our study offers a reasonable explanation for the controversy surrounding the role of visual attention in VWM prioritization in previous research. Furthermore, it sheds further light on the neural mechanisms underlying VWM prioritization from a perspective of high temporal and spatial resolution.

Key words: visual working memory, visual attention, prioritization, event-related optical signals, simultaneous acquisition