Scientific Understanding of Consciousness
Attention and Awareness in Primary Visual Cortex
Science 11 November 2011: 829-831
Attention But Not Awareness Modulates the BOLD Signal in the Human V1 During Binocular Suppression
Masataka Watanabe, Kang Cheng, Yusuke Murayama, Kenichi Ueno, Takeshi Asamizuya, Keiji Tanaka, and Nikos Logothetis
1School of Engineering, University of Tokyo, Tokyo, Japan.
2Max Planck Institute for Biological Cybernetics, Tuebingen, Germany.
3RIKEN Brain Science Institute, Wako, Saitama, Japan.
4Imaging Science and Biomedical Engineering, University of Manchester, Manchester, UK.
Although recent psychophysical studies indicate that visual awareness and top-down attention are two distinct processes, it is not clear how they are neurally dissociated in the visual system. Using a two-by-two factorial functional magnetic resonance imaging design with binocular suppression, we found that the visibility or invisibility of a visual target led to only nonsignificant blood oxygenation level–dependent (BOLD) effects in the human primary visual cortex (V1). Directing attention toward and away from the target had much larger and robust effects across all study participants. The difference in the lower-level limit of BOLD activation between attention and awareness illustrates dissociated neural correlates of the two processes. Our results agree with previously reported V1 BOLD effects on attention, while they invite a reconsideration of the functional role of V1 in visual awareness.
Attention and awareness in everyday life are closely interwoven; when we attend to an object, we become visually aware of its attributes, and being aware of an object may lead to attention directed toward it. However, recent progress in psychophysics indicates that visual attention and visual awareness are two dissociated functions in the visual system. Top-down attention directed toward perceptually invisible stimuli leads to negative aftereffects or priming, whereas the gist of scenes can be perceived despite the near absence of attention. Until now, there has been no clear evidence of differential neural correlates for attention and awareness.
The extent of the modulation of visual awareness on neuronal activity in V1 is crucial because V1 lies at the bottom of the cortical visual hierarchy. Previous studies have shown a large and significant awareness modulation on the BOLD signal in both human and monkey V1, whereas single-unit studies have failed to detect robust perception-locked changes. A similar twist exists for attention; single-unit studies claim only modest attentional effects in monkey V1, whereas numerous functional magnetic resonance imaging studies report a strong attentional modulation on the BOLD signal in human V1. The previously observed discrepancy between single-unit activity and BOLD signal in V1 in visual awareness experiments could be simply explained by attentional modulation, because attention was not thoroughly controlled in these previous studies. This is indeed a major experimental complication because, without adequate control, attention and awareness are intimately interrelated. Our results suggest that this is actually the case, and the previously reported awareness modulation on the BOLD signal in V1 is likely to be an artifact caused by the concurrent attentional modulation.
Previous findings claim a functional role of V1 in awareness, which seems to be immune to the problem of attentional confounds. The phenomenon of blindsight (namely, the manifestation of residual visual processing despite the total loss of visual awareness) caused by lesions in V1 suggests a role for V1 in awareness. However, if we assume the existence of a hypothetical area downstream of V1 as a neural correlate of awareness, a lesion in V1 would also heavily affect this area by removing a large portion of its afferent input. Transcranial magnetic stimulation (TMS) studies suggest that the feedback activity from the middle temporal area to V1 is crucial for visual awareness of TMS-induced motion phosphenes and motion stimuli. These, too, seem to contradict our results, but the claim that feedback activity to V1 is a necessary condition for awareness does not necessarily point to V1 as an area for neural modulation by the contents of conscious percept. As in the case of blindsight, downstream areas can be equally affected by the consequences of TMS application to V1. However, it should be emphasized that our current conclusions, based solely on BOLD measurements, do not completely rule out the involvement of V1 in visual awareness. Future efforts, using nonbiased and fine-scale neuronal measurement techniques that can capture the modulation in local circuits in V1 (such as specific neuronal subtypes), or using neural dynamics instead of firing rates (such as synchronization) that differ from hemodynamic measurements and conventional neural recordings, are needed to further clarify the involvement of V1 in visual awareness.
[end of paraphrase]
Return to — Attention