- Email: [email protected]
Exploiting Cognitive Brain Maps
Brain and Cognition 42, 64–67 (2000) doi:10.1006/brcg.1999.1163, available online at http://www.idealibrary.com on
Exploiting Cognitive Brain Maps Michael I. Posner Sackler Institute, Weill Medical College of Cornell University, New York
Until late in this century it did not seem possible that we would be able to see brain areas active while humans performed cognitive tasks. Most of those who entered psychology in the mid twentieth century learned from the work of Lashley (1931) that cognitive activity was generally a property of the whole brain not localized in any region. One of the tenets of Cognitive Psychology was that cognition was about software and thus the study of brain activity was largely irrelevant. However, in the past few years the methods of positron emission tomography (PET) and functional magnetic resonance imaging (fMRI) have transformed the cognitive landscape. Despite the previous views of the irrelevance of brain activity to cognition and many very appropriate cautions about the resolution, consistency, and utility of these methods, we find ourselves looking at maps of brain areas active during a variety of cognitive tasks like reading, listening, searching, learning sequences, comparing shapes or numbers, and many others. The accuracy and reliability with which we will be able to visualize this anatomy is bound to improve as new scanners and methods are developed. What are we to do with these maps? One step, which has been taken however incompletely, is to transform maps into circuits. To do this it was necessary to see when the specific anatomy involved in the task was active. Because PET and fMRI measured changes in blood flow that lag behind neuronal activity by a second or so it is not obvious how to use these methods to measure directly the time course of mental operations which are often in the millisecond range. However, by relating the generators found in imaging studies to electrical activity recorded from the scalp (e.g., Abdullaev & Posner, 1998; Hillyard & Anllo-Vento 1998) or by depth electrodes (Abdullaev, Bechtereva, & Melnichuk, 1998) it has proven possible to see when Address correspondence and reprint requests to Michael Posner, Department of Psychiatry, Sackler Institute, Weill Medical College of Cornell University, Box 140, 1300 York Ave., New York, NY 10021. 64 0278-2626/00 $35.00 Copyright 2000 by Academic Press All rights of reproduction in any form reserved.
MILLENNIUM ISSUE
65
in time many of these brain areas become active. The use, accuracy, and reliability of these methods are almost certain to improve over the next period. Another important step will be to determine the plasticity of brain circuits that serve cognition functions. When in infant and/or child development do particular circuits come on line? New adaptations of magnetic imaging may be able to trace non invasively myelination of specific pathways between brain areas (Le Bihan, 1995). We will then be able to try to predict when particular behaviors should emerge. I suspect the visual system and particular visually guided eye movements may be studied first, but higher level cognitive activity is sure to follow. We would then have a disciplined approach to understanding when a human brain becomes ready to learn a cognitive skill. Development is a particularly important and obvious way to study plasticity, but of course learning also changes the brain. We already have good examples of how practice in generating particular word associations change the circuitry involved (Raichle et al., 1994) and how priming tunes the number of neurons involved in a computation (Buckner & Koustaal, 1998). We should be able to observe at the circuit level what educational experiences can do to organize specific brain areas and their order of activation (Posner & McCandliss, 1999). Another important application will be in the study what various brain injuries or pathologies do to the circuitry that supports cognitive processes and how they recover spontaneously or with drug, practice, or insight therapies. We are just at the very start of guiding our therapeutic interventions with imaging methods. A quite different direction for the use of imaging is to begin to understand the reasons that certain computations occur in particular brain areas. There has been spectacular progress in efforts to relate imaged human brain areas to the visual maps obtained from cellular recording in primates (Tootell et al., 1998). This information could illuminate our understanding of why some areas are used to perform some functions. An important example is the somewhat controversial finding of the St. Louis PET group of a brain area that chunks the letters of visually presented into units which they termed the visual word form area (Petersen et al., 1990). Clearly any such brain system must be organized by the specific learning experiences that children have during their early education. Indeed these areas begin in childhood with responding to specific words the child has learned and only after training become more related to the orthography of the language than to the familiarity of learned exemplars (Posner & McCandliss, 1999). This area in the left ventral occipital lobe appears to be in or near areas active during the processing of color. We know that the color areas are late visual adaptations and that they have input from the high spatial frequency parvocellular systems of the geniculate and primary cortex. If we were to follow the exact computations performed by homologous areas in higher nonhuman primates and ob-
66
MILLENNIUM ISSUE
serve the role of spatial frequency and learning in their use we might have an empirical approach to the affinity of this area for the computation needed to organize visual letters. A more distant goal for the study of imaging, but one that needs careful thought, is the development of general principles of how closely related computations are expressed both in brain tissue and in performance. Kinsbourne & Hicks (1978) proposed that the more densely connected two brain areas were the more interference they would show when simultaneously active and the more priming when successively active. At the time this was proposed we were not able to test these ideas except by experiments using different motor outputs. Now, if one assumes that being closer in brain space relates to connectivity, there are many opportunites to test links between performance and neural distance. In neuroscience a very fundamental unit is the cortical column which is thought to form the basic computational unit of the brain. No similar principle is yet available for saying how similar cognitive operations should be mapped within the brain. The mechanisms that allow for covert shifts in orienting of attention (spotlight) are based in the parietal lobe, how do these relate to selection of the scale (zoom function)? One might expect two such related computations to be close together in comparison with each of them and mechanisms involved in deciding how much to like or prefer something. While mechanisms within a single cognitive domain may sometimes be close together (e.g., visual color and form) it often appears that both frontal and posterior areas are involved in the same domain (e.g., language and working memory). The progress made in mapping cognitive functions in the human brain has been swift and startling. However, there is still a long way to go to exploit these maps into a more general understanding of the role of brain areas in thought and behavior. REFERENCES Abdullaev, Y. G., Bechtereva, N. P., & Melnichuk, Y. 1998. Neuronal activity of human caudate nucleus and prefrontal cortex in cognitive tasks. Behavioural Brain Research, 159–177. Abdullaev, Y., & Posner, M. I. 1998. Event related potential imaging of semantic encoding during single word processing. Neuroimage, 7, 1–13. Buckner, R. L., & Koustaal, W. 1998. Functional neuroimaging studies of encoding, priming and explicit memory retrieval. Proceedings of the National Academy of Science USA, 95, 891–898. Hillyard, S. A., & Lourdes, Anllo-Vento 1998. Event related potentials in the study of visual attention. Proceedings of the National Academy of Science USA, 95, 781–787. Kinsbourne, M., & Hicks, H. 1978. Functional cerebral space: A model for overflow, transfer and interference effects in human performance. In J. Requin (Ed.), Attention and performance VII (pp. 345–363). New York: Wiley. Lashley, K. S. 1931. Mass action in cerebral function. Science, 73, 245–254.
MILLENNIUM ISSUE
67
LeBihan, D. 1995. Diffusion and perfusion magnetic resonance imaging. New York: Raven Press. Petersen, S. E., Fox, P. T., Snyder, A., & Raichle, M. E. 1990. Activation of extrastriate and frontal cortical areas by words and word-like stimuli. Science, 249, 1041–1044. Posner, M. I., & McCandliss, B. D. 1999. Brain circuitry during reading. In R. Klein & P. McMullen (Eds.), Converging methods for understanding reading and dyslexia (pp. 305– 337). Cambridge: MIT Press. Posner, M. I., Petersen, S. E., Fox, P. T., & Raichle, M. E. 1988. Localization of cognitive functions in the human brain. Science, 240, 1627–1631. Raichle, M. E., Fiez, J. A., Videen, T. O., Pardo, J. V., Fox, P. T., & Petersen, S. E. 1994. Practice-related changes in human brain functional anatomy during nonmotor learning. Cerebral Cortex, 4, 8–29. Tootell, R. B. H., Mendola, J. D., Hadjikhani, N. K., Liu, A. K., & Dale, A. M. 1998. The representation of the ipsilateral visual field in human cerebral cortex. Proceedings of the National Academy of Science USA, 95, 818–824.
Exploiting Cognitive Brain Maps Michael I. Posner Sackler Institute, Weill Medical College of Cornell University, New York
Until late in this century it did not seem possible that we would be able to see brain areas active while humans performed cognitive tasks. Most of those who entered psychology in the mid twentieth century learned from the work of Lashley (1931) that cognitive activity was generally a property of the whole brain not localized in any region. One of the tenets of Cognitive Psychology was that cognition was about software and thus the study of brain activity was largely irrelevant. However, in the past few years the methods of positron emission tomography (PET) and functional magnetic resonance imaging (fMRI) have transformed the cognitive landscape. Despite the previous views of the irrelevance of brain activity to cognition and many very appropriate cautions about the resolution, consistency, and utility of these methods, we find ourselves looking at maps of brain areas active during a variety of cognitive tasks like reading, listening, searching, learning sequences, comparing shapes or numbers, and many others. The accuracy and reliability with which we will be able to visualize this anatomy is bound to improve as new scanners and methods are developed. What are we to do with these maps? One step, which has been taken however incompletely, is to transform maps into circuits. To do this it was necessary to see when the specific anatomy involved in the task was active. Because PET and fMRI measured changes in blood flow that lag behind neuronal activity by a second or so it is not obvious how to use these methods to measure directly the time course of mental operations which are often in the millisecond range. However, by relating the generators found in imaging studies to electrical activity recorded from the scalp (e.g., Abdullaev & Posner, 1998; Hillyard & Anllo-Vento 1998) or by depth electrodes (Abdullaev, Bechtereva, & Melnichuk, 1998) it has proven possible to see when Address correspondence and reprint requests to Michael Posner, Department of Psychiatry, Sackler Institute, Weill Medical College of Cornell University, Box 140, 1300 York Ave., New York, NY 10021. 64 0278-2626/00 $35.00 Copyright 2000 by Academic Press All rights of reproduction in any form reserved.
MILLENNIUM ISSUE
65
in time many of these brain areas become active. The use, accuracy, and reliability of these methods are almost certain to improve over the next period. Another important step will be to determine the plasticity of brain circuits that serve cognition functions. When in infant and/or child development do particular circuits come on line? New adaptations of magnetic imaging may be able to trace non invasively myelination of specific pathways between brain areas (Le Bihan, 1995). We will then be able to try to predict when particular behaviors should emerge. I suspect the visual system and particular visually guided eye movements may be studied first, but higher level cognitive activity is sure to follow. We would then have a disciplined approach to understanding when a human brain becomes ready to learn a cognitive skill. Development is a particularly important and obvious way to study plasticity, but of course learning also changes the brain. We already have good examples of how practice in generating particular word associations change the circuitry involved (Raichle et al., 1994) and how priming tunes the number of neurons involved in a computation (Buckner & Koustaal, 1998). We should be able to observe at the circuit level what educational experiences can do to organize specific brain areas and their order of activation (Posner & McCandliss, 1999). Another important application will be in the study what various brain injuries or pathologies do to the circuitry that supports cognitive processes and how they recover spontaneously or with drug, practice, or insight therapies. We are just at the very start of guiding our therapeutic interventions with imaging methods. A quite different direction for the use of imaging is to begin to understand the reasons that certain computations occur in particular brain areas. There has been spectacular progress in efforts to relate imaged human brain areas to the visual maps obtained from cellular recording in primates (Tootell et al., 1998). This information could illuminate our understanding of why some areas are used to perform some functions. An important example is the somewhat controversial finding of the St. Louis PET group of a brain area that chunks the letters of visually presented into units which they termed the visual word form area (Petersen et al., 1990). Clearly any such brain system must be organized by the specific learning experiences that children have during their early education. Indeed these areas begin in childhood with responding to specific words the child has learned and only after training become more related to the orthography of the language than to the familiarity of learned exemplars (Posner & McCandliss, 1999). This area in the left ventral occipital lobe appears to be in or near areas active during the processing of color. We know that the color areas are late visual adaptations and that they have input from the high spatial frequency parvocellular systems of the geniculate and primary cortex. If we were to follow the exact computations performed by homologous areas in higher nonhuman primates and ob-
66
MILLENNIUM ISSUE
serve the role of spatial frequency and learning in their use we might have an empirical approach to the affinity of this area for the computation needed to organize visual letters. A more distant goal for the study of imaging, but one that needs careful thought, is the development of general principles of how closely related computations are expressed both in brain tissue and in performance. Kinsbourne & Hicks (1978) proposed that the more densely connected two brain areas were the more interference they would show when simultaneously active and the more priming when successively active. At the time this was proposed we were not able to test these ideas except by experiments using different motor outputs. Now, if one assumes that being closer in brain space relates to connectivity, there are many opportunites to test links between performance and neural distance. In neuroscience a very fundamental unit is the cortical column which is thought to form the basic computational unit of the brain. No similar principle is yet available for saying how similar cognitive operations should be mapped within the brain. The mechanisms that allow for covert shifts in orienting of attention (spotlight) are based in the parietal lobe, how do these relate to selection of the scale (zoom function)? One might expect two such related computations to be close together in comparison with each of them and mechanisms involved in deciding how much to like or prefer something. While mechanisms within a single cognitive domain may sometimes be close together (e.g., visual color and form) it often appears that both frontal and posterior areas are involved in the same domain (e.g., language and working memory). The progress made in mapping cognitive functions in the human brain has been swift and startling. However, there is still a long way to go to exploit these maps into a more general understanding of the role of brain areas in thought and behavior. REFERENCES Abdullaev, Y. G., Bechtereva, N. P., & Melnichuk, Y. 1998. Neuronal activity of human caudate nucleus and prefrontal cortex in cognitive tasks. Behavioural Brain Research, 159–177. Abdullaev, Y., & Posner, M. I. 1998. Event related potential imaging of semantic encoding during single word processing. Neuroimage, 7, 1–13. Buckner, R. L., & Koustaal, W. 1998. Functional neuroimaging studies of encoding, priming and explicit memory retrieval. Proceedings of the National Academy of Science USA, 95, 891–898. Hillyard, S. A., & Lourdes, Anllo-Vento 1998. Event related potentials in the study of visual attention. Proceedings of the National Academy of Science USA, 95, 781–787. Kinsbourne, M., & Hicks, H. 1978. Functional cerebral space: A model for overflow, transfer and interference effects in human performance. In J. Requin (Ed.), Attention and performance VII (pp. 345–363). New York: Wiley. Lashley, K. S. 1931. Mass action in cerebral function. Science, 73, 245–254.
MILLENNIUM ISSUE
67
LeBihan, D. 1995. Diffusion and perfusion magnetic resonance imaging. New York: Raven Press. Petersen, S. E., Fox, P. T., Snyder, A., & Raichle, M. E. 1990. Activation of extrastriate and frontal cortical areas by words and word-like stimuli. Science, 249, 1041–1044. Posner, M. I., & McCandliss, B. D. 1999. Brain circuitry during reading. In R. Klein & P. McMullen (Eds.), Converging methods for understanding reading and dyslexia (pp. 305– 337). Cambridge: MIT Press. Posner, M. I., Petersen, S. E., Fox, P. T., & Raichle, M. E. 1988. Localization of cognitive functions in the human brain. Science, 240, 1627–1631. Raichle, M. E., Fiez, J. A., Videen, T. O., Pardo, J. V., Fox, P. T., & Petersen, S. E. 1994. Practice-related changes in human brain functional anatomy during nonmotor learning. Cerebral Cortex, 4, 8–29. Tootell, R. B. H., Mendola, J. D., Hadjikhani, N. K., Liu, A. K., & Dale, A. M. 1998. The representation of the ipsilateral visual field in human cerebral cortex. Proceedings of the National Academy of Science USA, 95, 818–824.