Current Biology Vol 24 No 18 R838
13. Lin, Y.Y., and Gubb, D. (2009). Molecular dissection of Drosophila Prickle isoforms distinguishes their essential and overlapping roles in planar cell polarity. Dev. Biol. 325, 386–399. 14. Bosveld, F., Bonnet, I., Guirao, B., Tlili, S., Wang, Z., Petitalot, A., Marchand, R., Bardet, P.L., Marcq, P., Graner, F., et al. (2012). Mechanical control of morphogenesis by Fat/Dachsous/Four-jointed planar cell polarity pathway. Science 336, 724–727. 15. Mao, Y., Rauskolb, C., Cho, E., Hu, W.L., Hayter, H., Minihan, G., Katz, F.N., and Irvine, K.D. (2006). Dachs: an unconventional myosin that functions downstream of Fat to regulate growth, affinity and gene expression in Drosophila. Development 133, 2539–2551.
16. Sagner, A., Merkel, M., Aigouy, B., Gaebel, J., Brankatschk, M., Julicher, F., and Eaton, S. (2012). Establishment of global patterns of planar polarity during growth of the Drosophila wing epithelium. Curr. Biol. 22, 1296–1301. 17. Wu, J., Roman, A.C., Carvajal-Gonzalez, J.M., and Mlodzik, M. (2013). Wg and Wnt4 provide long-range directional input to planar cell polarity orientation in Drosophila. Nat. Cell Biol. 15, 1045–1055. 18. Brittle, A., Thomas, C., and Strutt, D. (2012). Planar polarity specification through asymmetric subcellular localization of Fat and Dachsous. Curr. Biol. 22, 907–914. 19. Feng, Y., and Irvine, K.D. (2007). Fat and Expanded act in parallel to regulate growth
Attentional Selection: Mexican Hats Everywhere A recent study elegantly shows that allocating attention to a particular color not only enhances perception of the attended color but also suppresses that of similar colors, presumably giving any potentially relevant object in the visual environment a perceptual advantage by increasing its perceptual strength at the expense of similar but different stimuli. Stefan Treue Visual perception resembles the task of an Alaskan bear standing in a rushing stream, trying to catch the salmon on its way up the river: just as the bear needs to detect the fish in the swirling mass of water to ensure his meal, we need to be able to concentrate our visual processing resources on the small fraction of relevant information in the torrent of data delivered by our eyes. This attentional selection can be based on a spatial location (as when the bear concentrates on a place between the rocks that the salmon prefer for their ascent) or on a feature (such as the unique color of the salmon’s scales that the bear has learned to look out for, the salmon’s body shape or the salmon’s orientation). While the neural mechanisms of spatial selection have been in the focus of attention research for decades, feature-based selection has only been in the center of interest much more recently. A study by Sto¨rmer and Alvarez [1] published recently in Current Biology provides a big step towards bringing our understanding of feature-based attention up to par with that of spatial attention. Much of our understanding of spatial attention is well-captured by the
metaphor of the spotlight — attending covertly (without making an eye movement) to one or two particular location(s) in our visual field makes our perception faster, more accurate, of higher spatial resolution and of enhanced sensitivity for fine changes. The physiological correlate of these enhancements is a gain increase of neurons with receptive fields that overlap the attended location, similar to the sensory effect of increasing the salience of a given stimulus. While the perceptual enhancement is often assumed to fall off monotonically with distance from the attended location, there is behavioral and electrophysiological evidence [2–5] for a suppressive zone in the direct vicinity of the spotlight of attention, in line with a computational model where such a ‘Mexican hat’ profile of cortical responsiveness with an excitatory center and an inhibitory surround is a core component [6]. Sto¨rmer and Alvarez’ [1] addressed the question of whether such an inhibitory surround also exists for feature-based attention. Feature-based attention refers to an enhancement of cortical information processing and perception for attended features across the whole visual field, a process particularly useful in visual search, where features
through Warts. Proc. Natl. Acad. Sci. USA 104, 20362–20367. 20. Sing, A., Tsatskis, Y., Fabian, L., Hester, I., Rosenfeld, R., Serricchio, M., Yau, N., Bietenhader, M., Shanbhag, R., Jurisicova, A., et al. (2014). The atypical cadherin Fat directly regulates mitochondrial function and metabolic state. Cell, in press.
Department of Zoology, University of Wisconsin, Madison, WI 53706 USA. E-mail: [email protected]
http://dx.doi.org/10.1016/j.cub.2014.08.008
of the searched item, but not its location, are known, and correspondingly perception of the attended feature is enhanced across the whole visual field [7–9], or where one of multiple items needs to be attended at a given location [10]. Just as in spatial attention, feature-based attentional effects are known to exist in other sensory domains too, such as somatosensory and auditory perception [11–13]. In Sto¨rmer and Alvarez’s study [1] human subjects had to detect brief periods of coherent motion in dot patterns of one target color embedded amongst randomly moving dots of another (distractor) color. This had to be done simultaneously for such two-color motion patterns in the left and in the right visual hemifield. The target color on the left and right stimulus could differ. Not surprisingly the best performance was observed when the target color was the same in the left and right field, presumably because subjects could use a single feature (the one target color) as a selection criterion across the whole visual field. The novel finding of the study was made when the difference in hue between the two target colors was systematically varied: the subjects’ performance was worst when the two target colors were different but similar, indicating that attending to one color suppressed similar but different colors across the whole visual field. In a second experiment Sto¨rmer and Alvarez [1] looked directly at the effect of feature-based attention on neural activity by measuring steady-state visual evoked potentials (SSVEPs). The SSVEP is the oscillatory response of the visual cortex to flickering stimuli: it has the same frequency as the driving
Dispatch R839
stimulus, and its amplitude is larger for attended relative to unattended stimuli [14]. Sto¨rmer and Alvarez [1] asked subjects to pay attention to the same two-color stimulus as in the first experiment, but without having to simultaneously pay attention in the opposite hemifield. They used the SSVEPs to provide a measure of the transfer of the feature-based attention subjects paid to the target color in one hemifield on the cortical response to an unattended stimulus of the same or a different color in the opposite hemifield. Given the known global spread of feature-based attention across the visual field they found that responses in the opposite visual field were most strongly suppressed, not for those colors most different from the attended one, but for colors close to the attended color. Both of these experiments document a spotlight of attention in feature-space that combines an enhanced representation of the attended feature with a penumbra of suppression for features that are similar but different from the attended feature [1]. Visual perception in primates is a carefully balanced combination of a powerful sensory periphery (the retinae of the two eyes) with an equally powerful central processing unit (the visual cortex with its array of distinct areas). The retinae are highly specialized, firstly using the over 100 million photoreceptors to capture the photons carrying information about the visual environment; secondly converting these rays into neural signals; and finally compressing the captured information to pass through the bottleneck of just one million ganglion cell axons making up the optic nerve. As far as we know, this is an entirely feedforward, hardwired system, implementing a series of algorithms that have been honed by millions of years of evolution. One such algorithm is that implemented by the center-surround receptive field organization that has been known since the earliest recordings from retinal ganglion cells. This receptive field profile performs a differencing operation, subtracting the illumination falling in the surround from the illumination in the center of the receptive field. As a consequence, retinal ganglion cells carry little information about the general light level but are maximally activated by a spatial
distribution of illumination differing between the center and surround, such as luminance edges, bars or small spots of light. Center surround algorithms therefore enhance contrast at sharp luminance boundaries. This approach seems so beneficial that it has been repeated in receptive fields in cortical neurons, such as the direction-selective cells of area MT/V5, specialized for processing visual motion. In contrast to the feedforward system of the retina of primates, their visual cortex is characterized by a massive back and forth of signal pathways. This allows for a highly adaptive approach in which the processing of information is dynamically modulated based on its behavioral significance — that is, by attention. Interestingly, focusing attention on a particular spatial location in our visual environment generates a center-surround modulation where signals and their processing in the center of our attentional spotlight are enhanced, while the immediate vicinity is suppressed — the attentional equivalent to the center-surround structure of retinal ganglion cell receptive fields. Here, the differencing operation serves to boost the processing of signals emanating from the spatial center while simultaneously suppressing nearby information, a push-pull effect that maximally biases processing of information towards the center locations. The new results of Stoermer and Alvarez [1] document the same center-surround algorithm for feature-based attention, driving home the wide presence of this approach in all aspects of visual information processing. The presence of center-surround operations throughout so many aspects of the visual system is an example of the power and efficiency of using a small set of fundamental operations across a whole neural system, all working together in creating an ‘integrated saliency map’ [15] or ‘priority map’ [16], a central representation of the visual environment that is both efficient and carefully constructed to enhance the most (potentially) relevant aspects of the sensory input at the expense of other information. These Mexican hat profiles of enhancing sensory contrasts, combined with the same algorithms for feature-based attention, presumably
allow not only the hungry Alaskan bear to detect the salmon despite all of its efforts to visually blend into its surrounding, but also enable us to find the proverbial needle in a haystack. References 1. Sto¨rmer, V.S., and Alvarez, G.A. (2014). Feature-based attention elicits surround suppression in feature space. Curr. Biol. 24, 1985–1988. 2. Mounts, J.R.W. (2000). Attentional capture by abrupt onsets and feature singletons produces inhibitory surrounds. Percept. Psychophys. 62, 1485–1493. 3. Mu¨ller, N.G., and Kleinschmidt, A. (2004). The attentional ‘spotlight’s’ penumbra: center-surround modulation in striate cortex. NeuroReport 15, 977–980. 4. Hopf, J.M., Boehler, C.N., Luck, S.J., Tsotsos, J.K., Heinze, H.J., and Schoenfeld, M.A. (2006). Direct neurophysiological evidence for spatial suppression surrounding the focus of attention in vision. Proc. Natl. Acad. Sci. USA 103, 1053–1058. 5. Niebergall, R., Khayat, P.S., Treue, S., and Martinez-Trujillo, J. (2011). Multifocal attention filters out distracter stimuli within and beyond receptive field boundaries of primate MT neurons. Neuron 72, 1067–1079. 6. Tsotsos, J.K., Culhane, S.M., and Cutzu, F. (2001). From foundational principles to a hierarchical selection circuit for attention. In Visual Attention and Cortical Circuits, J. Braun, C. Koch, and J.L. Davies, eds. (Cambridge, MA: MIT Press), pp. 285–306. 7. Treue, S., and Martinez-Trujillo, J.C. (1999). Feature-based attention influences motion processing gain in macaque visual cortex. Nature 399, 575–579. 8. Saenz, M., Buracas, G.T., and Boynton, G.M. (2002). Global effects of feature-based attention in human visual cortex. Nat. Neurosci. 5, 631–632. 9. Maunsell, J.H.R., and Treue, S. (2006). Feature-based attention in visual cortex. Trends Neurosci. 29, 317–322. 10. Patzwahl, D., and Treue, S. (2009). Combining spatial and feature-based attention within the receptive field of MT neurons. Vis. Res. 49, 1188–1193. 11. Shomstein, S., and Yantis, S. (2006). Parietal cortex mediates voluntary control of spatial and nonspatial auditory attention. J. Neurosci. 26, 435–439. 12. Paltoglou, A.E., Sumner, C.J., and Hall, D.A. (2011). Mapping feature-sensitivity and attentional modulation in human auditory cortex with functional magnetic resonance imaging. Eur. J. Neurosci. 33, 1733–1741. 13. Schweisfurth, M.A., Schweizer, R., and Treue, S. (2014). Feature-based attentional modulation of orientation perception in somatosensation. Front. Hum. Neurosci. 8, 1–8. 14. Andersen, S.K., Mueller, M.M., and Hillyard, S.A. (2011). In Cognitive Neuroscience of Attention, M.I. Posner, ed. (New York: Guilford), pp. 197–216. 15. Treue, S. (2003). Visual attention: the where, what, how and why of saliency. Curr. Opin. Neurobiol. 13, 428–432. 16. Bisley, J.W., and Goldberg, M.E. (2010). Attention, intention, and priority in the parietal lobe. Annu. Rev. Neurosci. 33, 1–21.
Cognitive Neuroscience Laboratory, German Primate Center, Goettingen and Faculty of Biology and Psychology, Goettingen Unversity, Goettingen, Germany. E-mail: [email protected] http://dx.doi.org/10.1016/j.cub.2014.08.019
13. Lin, Y.Y., and Gubb, D. (2009). Molecular dissection of Drosophila Prickle isoforms distinguishes their essential and overlapping roles in planar cell polarity. Dev. Biol. 325, 386–399. 14. Bosveld, F., Bonnet, I., Guirao, B., Tlili, S., Wang, Z., Petitalot, A., Marchand, R., Bardet, P.L., Marcq, P., Graner, F., et al. (2012). Mechanical control of morphogenesis by Fat/Dachsous/Four-jointed planar cell polarity pathway. Science 336, 724–727. 15. Mao, Y., Rauskolb, C., Cho, E., Hu, W.L., Hayter, H., Minihan, G., Katz, F.N., and Irvine, K.D. (2006). Dachs: an unconventional myosin that functions downstream of Fat to regulate growth, affinity and gene expression in Drosophila. Development 133, 2539–2551.
16. Sagner, A., Merkel, M., Aigouy, B., Gaebel, J., Brankatschk, M., Julicher, F., and Eaton, S. (2012). Establishment of global patterns of planar polarity during growth of the Drosophila wing epithelium. Curr. Biol. 22, 1296–1301. 17. Wu, J., Roman, A.C., Carvajal-Gonzalez, J.M., and Mlodzik, M. (2013). Wg and Wnt4 provide long-range directional input to planar cell polarity orientation in Drosophila. Nat. Cell Biol. 15, 1045–1055. 18. Brittle, A., Thomas, C., and Strutt, D. (2012). Planar polarity specification through asymmetric subcellular localization of Fat and Dachsous. Curr. Biol. 22, 907–914. 19. Feng, Y., and Irvine, K.D. (2007). Fat and Expanded act in parallel to regulate growth
Attentional Selection: Mexican Hats Everywhere A recent study elegantly shows that allocating attention to a particular color not only enhances perception of the attended color but also suppresses that of similar colors, presumably giving any potentially relevant object in the visual environment a perceptual advantage by increasing its perceptual strength at the expense of similar but different stimuli. Stefan Treue Visual perception resembles the task of an Alaskan bear standing in a rushing stream, trying to catch the salmon on its way up the river: just as the bear needs to detect the fish in the swirling mass of water to ensure his meal, we need to be able to concentrate our visual processing resources on the small fraction of relevant information in the torrent of data delivered by our eyes. This attentional selection can be based on a spatial location (as when the bear concentrates on a place between the rocks that the salmon prefer for their ascent) or on a feature (such as the unique color of the salmon’s scales that the bear has learned to look out for, the salmon’s body shape or the salmon’s orientation). While the neural mechanisms of spatial selection have been in the focus of attention research for decades, feature-based selection has only been in the center of interest much more recently. A study by Sto¨rmer and Alvarez [1] published recently in Current Biology provides a big step towards bringing our understanding of feature-based attention up to par with that of spatial attention. Much of our understanding of spatial attention is well-captured by the
metaphor of the spotlight — attending covertly (without making an eye movement) to one or two particular location(s) in our visual field makes our perception faster, more accurate, of higher spatial resolution and of enhanced sensitivity for fine changes. The physiological correlate of these enhancements is a gain increase of neurons with receptive fields that overlap the attended location, similar to the sensory effect of increasing the salience of a given stimulus. While the perceptual enhancement is often assumed to fall off monotonically with distance from the attended location, there is behavioral and electrophysiological evidence [2–5] for a suppressive zone in the direct vicinity of the spotlight of attention, in line with a computational model where such a ‘Mexican hat’ profile of cortical responsiveness with an excitatory center and an inhibitory surround is a core component [6]. Sto¨rmer and Alvarez’ [1] addressed the question of whether such an inhibitory surround also exists for feature-based attention. Feature-based attention refers to an enhancement of cortical information processing and perception for attended features across the whole visual field, a process particularly useful in visual search, where features
through Warts. Proc. Natl. Acad. Sci. USA 104, 20362–20367. 20. Sing, A., Tsatskis, Y., Fabian, L., Hester, I., Rosenfeld, R., Serricchio, M., Yau, N., Bietenhader, M., Shanbhag, R., Jurisicova, A., et al. (2014). The atypical cadherin Fat directly regulates mitochondrial function and metabolic state. Cell, in press.
Department of Zoology, University of Wisconsin, Madison, WI 53706 USA. E-mail: [email protected]
http://dx.doi.org/10.1016/j.cub.2014.08.008
of the searched item, but not its location, are known, and correspondingly perception of the attended feature is enhanced across the whole visual field [7–9], or where one of multiple items needs to be attended at a given location [10]. Just as in spatial attention, feature-based attentional effects are known to exist in other sensory domains too, such as somatosensory and auditory perception [11–13]. In Sto¨rmer and Alvarez’s study [1] human subjects had to detect brief periods of coherent motion in dot patterns of one target color embedded amongst randomly moving dots of another (distractor) color. This had to be done simultaneously for such two-color motion patterns in the left and in the right visual hemifield. The target color on the left and right stimulus could differ. Not surprisingly the best performance was observed when the target color was the same in the left and right field, presumably because subjects could use a single feature (the one target color) as a selection criterion across the whole visual field. The novel finding of the study was made when the difference in hue between the two target colors was systematically varied: the subjects’ performance was worst when the two target colors were different but similar, indicating that attending to one color suppressed similar but different colors across the whole visual field. In a second experiment Sto¨rmer and Alvarez [1] looked directly at the effect of feature-based attention on neural activity by measuring steady-state visual evoked potentials (SSVEPs). The SSVEP is the oscillatory response of the visual cortex to flickering stimuli: it has the same frequency as the driving
Dispatch R839
stimulus, and its amplitude is larger for attended relative to unattended stimuli [14]. Sto¨rmer and Alvarez [1] asked subjects to pay attention to the same two-color stimulus as in the first experiment, but without having to simultaneously pay attention in the opposite hemifield. They used the SSVEPs to provide a measure of the transfer of the feature-based attention subjects paid to the target color in one hemifield on the cortical response to an unattended stimulus of the same or a different color in the opposite hemifield. Given the known global spread of feature-based attention across the visual field they found that responses in the opposite visual field were most strongly suppressed, not for those colors most different from the attended one, but for colors close to the attended color. Both of these experiments document a spotlight of attention in feature-space that combines an enhanced representation of the attended feature with a penumbra of suppression for features that are similar but different from the attended feature [1]. Visual perception in primates is a carefully balanced combination of a powerful sensory periphery (the retinae of the two eyes) with an equally powerful central processing unit (the visual cortex with its array of distinct areas). The retinae are highly specialized, firstly using the over 100 million photoreceptors to capture the photons carrying information about the visual environment; secondly converting these rays into neural signals; and finally compressing the captured information to pass through the bottleneck of just one million ganglion cell axons making up the optic nerve. As far as we know, this is an entirely feedforward, hardwired system, implementing a series of algorithms that have been honed by millions of years of evolution. One such algorithm is that implemented by the center-surround receptive field organization that has been known since the earliest recordings from retinal ganglion cells. This receptive field profile performs a differencing operation, subtracting the illumination falling in the surround from the illumination in the center of the receptive field. As a consequence, retinal ganglion cells carry little information about the general light level but are maximally activated by a spatial
distribution of illumination differing between the center and surround, such as luminance edges, bars or small spots of light. Center surround algorithms therefore enhance contrast at sharp luminance boundaries. This approach seems so beneficial that it has been repeated in receptive fields in cortical neurons, such as the direction-selective cells of area MT/V5, specialized for processing visual motion. In contrast to the feedforward system of the retina of primates, their visual cortex is characterized by a massive back and forth of signal pathways. This allows for a highly adaptive approach in which the processing of information is dynamically modulated based on its behavioral significance — that is, by attention. Interestingly, focusing attention on a particular spatial location in our visual environment generates a center-surround modulation where signals and their processing in the center of our attentional spotlight are enhanced, while the immediate vicinity is suppressed — the attentional equivalent to the center-surround structure of retinal ganglion cell receptive fields. Here, the differencing operation serves to boost the processing of signals emanating from the spatial center while simultaneously suppressing nearby information, a push-pull effect that maximally biases processing of information towards the center locations. The new results of Stoermer and Alvarez [1] document the same center-surround algorithm for feature-based attention, driving home the wide presence of this approach in all aspects of visual information processing. The presence of center-surround operations throughout so many aspects of the visual system is an example of the power and efficiency of using a small set of fundamental operations across a whole neural system, all working together in creating an ‘integrated saliency map’ [15] or ‘priority map’ [16], a central representation of the visual environment that is both efficient and carefully constructed to enhance the most (potentially) relevant aspects of the sensory input at the expense of other information. These Mexican hat profiles of enhancing sensory contrasts, combined with the same algorithms for feature-based attention, presumably
allow not only the hungry Alaskan bear to detect the salmon despite all of its efforts to visually blend into its surrounding, but also enable us to find the proverbial needle in a haystack. References 1. Sto¨rmer, V.S., and Alvarez, G.A. (2014). Feature-based attention elicits surround suppression in feature space. Curr. Biol. 24, 1985–1988. 2. Mounts, J.R.W. (2000). Attentional capture by abrupt onsets and feature singletons produces inhibitory surrounds. Percept. Psychophys. 62, 1485–1493. 3. Mu¨ller, N.G., and Kleinschmidt, A. (2004). The attentional ‘spotlight’s’ penumbra: center-surround modulation in striate cortex. NeuroReport 15, 977–980. 4. Hopf, J.M., Boehler, C.N., Luck, S.J., Tsotsos, J.K., Heinze, H.J., and Schoenfeld, M.A. (2006). Direct neurophysiological evidence for spatial suppression surrounding the focus of attention in vision. Proc. Natl. Acad. Sci. USA 103, 1053–1058. 5. Niebergall, R., Khayat, P.S., Treue, S., and Martinez-Trujillo, J. (2011). Multifocal attention filters out distracter stimuli within and beyond receptive field boundaries of primate MT neurons. Neuron 72, 1067–1079. 6. Tsotsos, J.K., Culhane, S.M., and Cutzu, F. (2001). From foundational principles to a hierarchical selection circuit for attention. In Visual Attention and Cortical Circuits, J. Braun, C. Koch, and J.L. Davies, eds. (Cambridge, MA: MIT Press), pp. 285–306. 7. Treue, S., and Martinez-Trujillo, J.C. (1999). Feature-based attention influences motion processing gain in macaque visual cortex. Nature 399, 575–579. 8. Saenz, M., Buracas, G.T., and Boynton, G.M. (2002). Global effects of feature-based attention in human visual cortex. Nat. Neurosci. 5, 631–632. 9. Maunsell, J.H.R., and Treue, S. (2006). Feature-based attention in visual cortex. Trends Neurosci. 29, 317–322. 10. Patzwahl, D., and Treue, S. (2009). Combining spatial and feature-based attention within the receptive field of MT neurons. Vis. Res. 49, 1188–1193. 11. Shomstein, S., and Yantis, S. (2006). Parietal cortex mediates voluntary control of spatial and nonspatial auditory attention. J. Neurosci. 26, 435–439. 12. Paltoglou, A.E., Sumner, C.J., and Hall, D.A. (2011). Mapping feature-sensitivity and attentional modulation in human auditory cortex with functional magnetic resonance imaging. Eur. J. Neurosci. 33, 1733–1741. 13. Schweisfurth, M.A., Schweizer, R., and Treue, S. (2014). Feature-based attentional modulation of orientation perception in somatosensation. Front. Hum. Neurosci. 8, 1–8. 14. Andersen, S.K., Mueller, M.M., and Hillyard, S.A. (2011). In Cognitive Neuroscience of Attention, M.I. Posner, ed. (New York: Guilford), pp. 197–216. 15. Treue, S. (2003). Visual attention: the where, what, how and why of saliency. Curr. Opin. Neurobiol. 13, 428–432. 16. Bisley, J.W., and Goldberg, M.E. (2010). Attention, intention, and priority in the parietal lobe. Annu. Rev. Neurosci. 33, 1–21.
Cognitive Neuroscience Laboratory, German Primate Center, Goettingen and Faculty of Biology and Psychology, Goettingen Unversity, Goettingen, Germany. E-mail: [email protected] http://dx.doi.org/10.1016/j.cub.2014.08.019
