Neuropsychologia 49 (2011) 1578–1588 Contents lists available at ScienceDirect Neuropsychologia journal homepage: www.elsevier.com/locate/neuropsychologia Magnetic stimulation of the dorsolateral prefrontal cortex dissociates fragile visual short-term memory from visual working memory Ilja G. Sligte ∗ , Martijn E. Wokke, Johannes P. Tesselaar, H. Steven Scholte, Victor A.F. Lamme Cognitive Neuroscience Group, Department of Psychology, University of Amsterdam, Roetersstraat 15, 1018WB Amsterdam, The Netherlands article info Article history: Received 31 July 2010 Received in revised form 26 November 2010 Accepted 6 December 2010 Available online 17 December 2010 Keywords: TMS Sensory memory Fragile VSTM Working memory Dorsolateral prefrontal cortex Awareness abstract To guide our behavior in successful ways, we often need to rely on information that is no longer in view, but maintained in visual short-term memory (VSTM). While VSTM is usually broken down into iconic memory (brief and high-capacity store) and visual working memory (sustained, yet limited-capacity store), recent studies have suggested the existence of an additional and intermediate form of VSTM that depends on activity in extrastriate cortex. In previous work, we have shown that this fragile form of VSTM can be dissociated from iconic memory. In the present study, we provide evidence that fragile VSTM is different from visual working memory as magnetic stimulation of the right dorsolateral prefrontal cortex (DLPFC) disrupts visual working memory, while leaving fragile VSTM intact. In addition, we observed that people with high DLPFC activity had superior working memory capacity compared to people with low DLPFC activity, and only people with high DLPFC activity really showed a reduction in working memory capacity in response to magnetic stimulation. Altogether, this study shows that VSTM consists of three stages that have clearly different characteristics and rely on different neural structures. On the methodological side, we show that it is possible to predict individual susceptibility to magnetic stimulation based on functional MRI activity. Crown Copyright © 2010 Published by Elsevier Ltd. All rights reserved. 1. Introduction Our brain essentially acts as a filter that reduces the amount of available information at each subsequent step in the neural hierarchy. This mechanism is especially evident when looking at different stages in visual short-term memory (VSTM). Initially, people build up a high-capacity representation in iconic memory (Sperling, 1960) that is related to persistence in retinal photore- ceptors (Sligte, Scholte, & Lamme, 2008) and primary visual cortex (Duysens, Orban, Cremieux, & Maes, 1985) beyond stimulus dura- tion. Iconic memory is usually measured by presenting a memory array containing multiple rows of letters. Then, after offset of the memory array, a partial-report cue is shown that singles out the row to report. When this partial-report cue immediately follows memory array offset, people can report almost all letters from any specific row, suggesting that all or at least a large amount of items are stored (Averbach & Coriell, 1961; Sperling, 1960). Iconic mem- ory traces are short-lived as they can only be measured over the first half second after memory array offset and they are extremely volatile as well (Averbach & Coriell, 1961; Sperling, 1963); each time a new image hits our retina, iconic memory is erased to make way for a new high-capacity internal picture. ∗ Corresponding author. Tel.: +31 20 525 8868; fax: +31 20 639 1656. E-mail address: I.G.Sligte@uva.nl (I.G. Sligte). Still, when we lay our eyes on a pretty person walking by, we are able to retain his/her appearance in mind for some time, even in the face of the continuous arrival of new visual information. This kind of memory – that is resistant to overwriting – is usually called visual working memory. One of the striking aspects of visual working memory is its severe capacity limit and this can be illus- trated beautifully with change detection experiments. In standard change detection experiments performed in the lab (see Figs. 1, bot- tom, and 2, top for examples), people are shown a memory display containing multiple objects or a complex scene and they are asked to memorize the entire image. After a brief retention interval, a test display (or probe/match display) is shown in which one of the objects has changed with respect to the memory display on 50% of the trials and subjects have to indicate whether there was a change between displays or not. In general, people perform badly on change detection tasks, even when changes are as large as a jet engine or a building disappearing (Rensink, O’Regan, & Clark, 1997). This apparent blindness to changes can be well explained; usually, changes in the environment are accompanied by motion signals that automatically capture attention (Rensink, 2002; Simons & Rensink, 2005). However, when capture of attention is prevented by masking the change (in this case by interposing a blank interval), people have to rely on top-down information that is represented in visual working memory. As change blindness is the rule rather than the exception, top-down resources are apparently very sparse; it has been estimated that visual working memory 0028-3932/$ – see front matter. Crown Copyright © 2010 Published by Elsevier Ltd. All rights reserved. doi:10.1016/j.neuropsychologia.2010.12.010