Visual working memory (VWM) refers to the capacity to temporarily maintain and manipulate visual information. While VWM has traditionally been associated with frontal and parietal brain regions, recent evidence indicates that representations can also be maintained in early visual cortex (EVC) via activity-silent neural states. These latent signals can be perturbed by high-contrast neutral stimuli. Two mechanisms have been proposed to explain this phenomenon: top-down cortical feedback and short-term synaptic plasticity (STSP) driven by sustained sensory input. In a recent study, Krause and colleagues evaluated these models with respect to two defining properties of VWM: flexible selection and resistance to distraction. Using computational modeling, they showed that the feedback model provides a superior account of EVC activity during VWM maintenance. The present thesis tested these predictions using electroencephalography (EEG) and a retro-cue paradigm, in which visual perturbations in the delay period were used to invoke latent activities correlating to VWM maintenance. We hypothesized that only cued target representations would be decodable at the neural level, and that distractor-related traces would be absent, consistent with the feedback model. Exploratory analyses of behavioral performance were also conducted. Behaviorally, participants showed below-chance accuracy with a bias toward distractor orientations, consistent with existing literature. However, EEG decoding revealed that target and distractor orientations were only recoverable during their respective presentations and were undecodable after impulse perturbations. These results provide an inconclusive account for the competing models of VWM and highlight the need for larger samples and more refined paradigms.
Visual working memory (VWM) refers to the capacity to temporarily maintain and manipulate visual information. While VWM has traditionally been associated with frontal and parietal brain regions, recent evidence indicates that representations can also be maintained in early visual cortex (EVC) via activity-silent neural states. These latent signals can be perturbed by high-contrast neutral stimuli. Two mechanisms have been proposed to explain this phenomenon: top-down cortical feedback and short-term synaptic plasticity (STSP) driven by sustained sensory input. In a recent study, Krause and colleagues evaluated these models with respect to two defining properties of VWM: flexible selection and resistance to distraction. Using computational modeling, they showed that the feedback model provides a superior account of EVC activity during VWM maintenance. The present thesis tested these predictions using electroencephalography (EEG) and a retro-cue paradigm, in which visual perturbations in the delay period were used to invoke latent activities correlating to VWM maintenance. We hypothesized that only cued target representations would be decodable at the neural level, and that distractor-related traces would be absent, consistent with the feedback model. Exploratory analyses of behavioral performance were also conducted. Behaviorally, participants showed below-chance accuracy with a bias toward distractor orientations, consistent with existing literature. However, EEG decoding revealed that target and distractor orientations were only recoverable during their respective presentations and were undecodable after impulse perturbations. These results provide an inconclusive account for the competing models of VWM and highlight the need for larger samples and more refined paradigms.
Do distractors linger in the Early Visual Cortex? AN EEG decoding study of visual working memory.
DANG, NGOC LINH TRANG
2025/2026
Abstract
Visual working memory (VWM) refers to the capacity to temporarily maintain and manipulate visual information. While VWM has traditionally been associated with frontal and parietal brain regions, recent evidence indicates that representations can also be maintained in early visual cortex (EVC) via activity-silent neural states. These latent signals can be perturbed by high-contrast neutral stimuli. Two mechanisms have been proposed to explain this phenomenon: top-down cortical feedback and short-term synaptic plasticity (STSP) driven by sustained sensory input. In a recent study, Krause and colleagues evaluated these models with respect to two defining properties of VWM: flexible selection and resistance to distraction. Using computational modeling, they showed that the feedback model provides a superior account of EVC activity during VWM maintenance. The present thesis tested these predictions using electroencephalography (EEG) and a retro-cue paradigm, in which visual perturbations in the delay period were used to invoke latent activities correlating to VWM maintenance. We hypothesized that only cued target representations would be decodable at the neural level, and that distractor-related traces would be absent, consistent with the feedback model. Exploratory analyses of behavioral performance were also conducted. Behaviorally, participants showed below-chance accuracy with a bias toward distractor orientations, consistent with existing literature. However, EEG decoding revealed that target and distractor orientations were only recoverable during their respective presentations and were undecodable after impulse perturbations. These results provide an inconclusive account for the competing models of VWM and highlight the need for larger samples and more refined paradigms.| File | Dimensione | Formato | |
|---|---|---|---|
|
Dang Ngoc Linh Trang 2102393 - Thesis CN2.pdf
Accesso riservato
Dimensione
1.11 MB
Formato
Adobe PDF
|
1.11 MB | Adobe PDF |
The text of this website © Università degli studi di Padova. Full Text are published under a non-exclusive license. Metadata are under a CC0 License
https://hdl.handle.net/20.500.12608/108179