Decoding Visual Long-Term Memory Traces During Working Memory Maintenance In Early Visual Cortex - Info and Reading Options

The Internet Archive:

In recent years, researchers have proposed that the same neural regions responsible for processing sensory information are also recruited to actively store it for short durations (Curtis & D’Esposito, 2003; D’Esposito, 2007; D’Esposito & Postle, 2015; Postle, 2006), a framework known as the sensory recruitment hypothesis (Serences et al., 2009). Supporting evidence comes from multivoxel pattern analysis (MVPA) studies, which have successfully decoded stimulus-specific visual features from activity in the visual cortex, including orientation (Harrison & Tong, 2009; Ester et al., 2015; Pratte & Tong, 2014), colour (Serences et al., 2009), static patterns (Christophel et al., 2012), and motion characteristics like speed and direction (Emrich et al., 2013; Riggall & Postle, 2012). Similar decoding has also been achieved for more complex visual features, such as flow-field motion (Christophel & Haynes, 2014) and saccade goals (Rahmati et al., 2018). These studies support a distributed model of working memory, where sensory areas maintain low-level representations, while associative regions are responsible for higher-order, abstract information (Christophel et al., 2017; Rademaker et al. 2019; Iamshchinina et al. 2021). While these studies demonstrate the role of sensory regions in visual working memory (vWM), they typically rely on simple, non-semantic stimuli to ensure experimental control. However, this approach limits ecological validity. Recent research has shown that vWM performance benefits from visual long-term memory (vLTM), with factors like meaning, familiarity, and expertise enhancing vWM (Brady & Störmer, 2022; Jackson & Raymond, 2008; Xie & Zhang, 2018). These findings suggest that pre-existing vLTM traces play a critical role in the encoding and maintenance of vWM. Yet, studying vWM and vLTM together remains challenging due to the lack of suitable stimuli. Simple visual features used in vWM tasks are not ideal for vLTM studies, while naturalistic images often lack experimental control. This study aims to clarify how pre-existing vLTM traces influence the relationship between vWM and sensory representations. Recent advances in generative neural networks have made it possible to create synthetic, naturalistic images with controlled perceptual variability (Goetschalckx et al., 2021; Son et al., 2022). Using stable diffusion, we developed a custom stimulus set of high-quality, naturalistic images that vary along a perceptual scale while maintaining consistent semantic content (Pettini et al., 2024). To investigate the interaction between vWM and vLTM, we conducted a two-session fMRI experiment. In the first session, participants performed a delayed match-to-sample task. They memorised a complex, naturalistic target image, memorized it for an 8-second delay and then were required to identify it when presented together with a distractor. Task difficulty (easy, medium, hard) was manipulated by varying the level of perceptual similarity between the target and distractor. Repeated and non-repeated targets are included to examine the influence of pre-existing vLTM traces on vWM. Importantly, target and distractor images differ perceptually but retain the same semantic meaning, making them well-suited for a working memory task with naturalistic stimuli. In the second session, participants view the same images while performing a simple orthogonal fixation task. This session provides independent sensory encoding data for the subsequent MVPA analysis. By linking perceptual and memory-related activity, this study addresses how vLTM influences vWM processes in the early visual cortex (V1, V2, V3).

Search for “Decoding Visual Long-Term Memory Traces During Working Memory Maintenance In Early Visual Cortex” downloads:

Visit our Downloads Search page to see if downloads are available.

Find “Decoding Visual Long-Term Memory Traces During Working Memory Maintenance In Early Visual Cortex” in Libraries Near You:

Read or borrow “Decoding Visual Long-Term Memory Traces During Working Memory Maintenance In Early Visual Cortex” from your local library.

• The WorldCat Libraries Catalog: Find a copy of “Decoding Visual Long-Term Memory Traces During Working Memory Maintenance In Early Visual Cortex” at a library near you.