Abstract
Visual working memory provides an essential link between perception and cognition, allowing for active maintenance of information no longer visible. Interestingly, a single visual feature such as orientation can be precisely maintained in working memory over many seconds, with delayed discrimination performance nearly matching that of immediate discrimination (Pasternak & Greenlee, 2005). The neural substrate of this highly faithful memory for specific visual features is not well understood; early visual areas show little evidence of sustained activity during working memory (Offen et al., 2008), but are the most selective for these low-level features. Here, we investigated whether early visual areas might be important for maintaining basic visual features in working memory. Specifically, we used fMRI decoding methods (Kamitani & Tong, 2005) to assess whether orientation-selective activity is present in early visual areas during a delayed orientation discrimination task. Two sequentially presented orientation gratings (∼25º and ∼115º) were followed by a cue indicating which grating to hold in memory. After an 11-second delay, a test grating was presented for participants to discriminate relative to the cued orientation. Decoding analyses revealed that activity patterns in areas V1–V4 could reliably predict the orientation maintained in memory, even when overall activity levels were low. These orientation-selective activity patterns were sustained throughout the entire delay, and were similar to responses evoked by unattended, task-irrelevant gratings. Additional control experiments ruled out alternative cognitive accounts such as visual expectancy, selection from memory, or reliance on long-term memory. For example, random orientations held in working memory could also be decoded, indicating that this sustained orientation-selective activity found in early visual areas did not depend on long-term memory. Our results demonstrate that early visual areas can retain specific information about visual features held in working memory, over periods of many seconds when no physical stimulus is present.
This work was supported by National Institutes of Health grant R01 EY017082 to F.T. and a postgraduate scholarship from the Natural Sciences and Engineering Research Council of Canada to S.A.H.