Neural correlates of conscious perception in the attentional blink Cornelia Kranczioch, a Stefan Debener, a Jens Schwarzbach, b Rainer Goebel, c and Andreas K. Engel a, * a Institut fu ¨r Neurophysiologie und Pathophysiologie, Zentrum fu ¨r Experimentelle Medizin, Universita ¨tsklinikum Hamburg-Eppendorf, 20246 Hamburg, Germany b F.C. Donders Centre for Cognitive Neuroimaging, 6500 HB Nijmegen, The Netherlands c Faculteit der Psychologie, Universiteit Maastricht, 6200 MD Maastricht, The Netherlands Received 16 April 2004; revised 5 August 2004; accepted 16 September 2004 Available online 24 November 2004 If attending to a target in a rapid stream of visual stimuli within the next 400 ms or so, a second target in the stream is frequently not detected by an observer. This so-called attentional blink can provide a comparison of neural signals elicited by identical stimuli that, in one condition, reach conscious awareness and, in the other, fail to be selected for awareness. In the present study, using event-related functional magnetic resonance imaging (fMRI), differences of neural activation were studied in an attentional blink experiment in order to identify brain regions putatively involved in controlling the access of information to consciousness. Subjects viewed a rapid stream of black letters in which the second target (T2) was either presented within or outside the attentional blink period, or not at all. We observed an increase in activation for detected as compared to missed T2 presented during the attentional blink in frontal and parietal cortices. In contrast, in occipitotemporal regions activation was increased for missed as compared to detected T2. Furthermore, in several frontal and parietal areas, missed targets were associated with increased activity if compared to the condition in which no second target was presented. Finally, a selective decrease in activation for detected T2 presented during the attentional blink was observed in areas associated with emotional and predominantly automatic processing. While activations in occipitotemporal regions might mainly reflect duration of attentive search, the frontoparietal areas seem to be involved in a highly distributed network controlling visual awareness. D 2004 Elsevier Inc. All rights reserved. Keywords: Consciousness; Perception; Brain Introduction In recent years, it has become possible to study neural correlates of consciousness (NCC) with methods of cognitive neuroimaging. The search for NCC is an empirical investigation that remains, at least at present, neutral with respect to philosophical issues of mind–brain relationship or causality. Instead, this search concen- trates on identifying and characterizing neural activity patterns that specifically co-vary with conscious experience, rather than with unconscious perception or action (Crick and Koch, 1990; Engel and Singer, 2001; Rees et al., 2002). Due to convergent results from studies using neuroimaging techniques in normal human subjects, invasive recordings in patients, as well as microelectrode approaches in animals, some progress has been made in recent years, both with respect to identifying mechanisms that may be involved in controlling access to consciousness and with respect to studying activity patterns that correlate with specific contents of conscious mental states. What emerges from these studies is that conscious awareness presupposes a complex set of intertwined functions, including sensory preprocessing, attention, and working memory (Crick and Koch, 1990, 2003; Rees et al., 2002). The NCC, thus, is likely to involve a highly distributed set of brain areas subserving these functions. This network engages, via large- scale dynamic interactions, in globally coherent states (Dehaene et al., 2003; Engel and Singer, 2001; Engel et al., 2001; Varela et al., 2001) that seem required for the establishment of a global workspace carrying the contents of awareness (Newman and Baars, 1993). What is still largely unclear is which areas exactly are involved in the network controlling the selection of information through cooperative interaction and, moreover, what exactly the constraints and mechanisms are that underlie the selection of sensory signals for conscious awareness. Sensory paradigms suited for the study of NCC allow the comparison of brain activation in response to physically identical stimuli that are selected for conscious perception in one exper- imental condition but excluded from conscious perception in a control condition. A paradigm that meets this criterion particularly 1053-8119/$ - see front matter D 2004 Elsevier Inc. All rights reserved. doi:10.1016/j.neuroimage.2004.09.024 * Corresponding author. Institut fqr Neurophysiologie und Pathophy- siologie, Zentrum fqr Experimentelle Medizin, Universit7tsklinikum Ham- burg-Eppendorf, Martinistr. 52, 20246 Hamburg, Germany. Fax: +49 40 428037752. E-mail address: ak.engel@uke.uni-hamburg.de (A.K. Engel). Available online on ScienceDirect (www.sciencedirect.com). www.elsevier.com/locate/ynimg NeuroImage 24 (2005) 704 – 714