How does the brain determine the colors we see? The retina contains three cone types, but each cone is actually colorblind: the red-sensitive cone cannot tell the difference between a dim yellow-green and a bright blue. Theoretically color can be calculated by comparing the activity of the different cone types, but how the brain does this remains mysterious. This project will investigate how cone signals are processed by specialized brain regions thought to be important in color perception. It will employ functional magnetic resonance imaging (fMRI) to generate an overview of regions involved in color in the macaque monkey, a model of human trichromatic color vision. Microelectrodes will then be targeted to specific regions identified by fMRI. By correlating the patterns of neural activity of single neurons with controlled visual stimuli, this project will test hypotheses about the function of the neurons in encoding color percepts. One important question that will be addressed is how color-selective neurons sample the cone signals to achieve color selectivity. This project aims to reveal not only the neural mechanisms underlying color, but more generally the principles that govern how neural circuits transform sensory stimuli into perception and behavior. It will provide a unique opportunity to train undergraduate women in the use of cutting-edge systems research technology, and will foster inter-institutional networks and partnerships. The results will have broad impact: they will be useful to engineers interested building artificial vision systems, and will engage the general public by tackling important questions such as how and why color is compelling. Finally, by combining single-unit electrophysiology and fMRI in the same subjects, this project will further our knowledge of the neural basis for fMRI, which is still poorly understood, despite the fact that fMRI is the most widely used clinical tool for assessing brain function.
How do we see color? This ancient question is compelling because it probes how we obtain sensory knowledge about the world. Indeed, color is a powerful tool of neuroscience: a classical paradigm for exploring consciousness that affords the rare possibility of linking many aspects of perception/cognition with neural mechanisms. But two impediments have held back our understanding of color. First, a lack of knowledge about the functional organization of color circuits in the brain; and second, a paucity of information on the extent to which rhesus macaque monkeys (the predominant animal model of human vision/cognition) possess the same color abilities as humans. Work supported by the National Science Foundation (grant #0918064) has overcome both of these hurdles, and made substantial progress towards addressing the principal question. Careful behavioral tests were carried out on macaque monkeys and humans, under identical experimental conditions, to compare the color abilities of these two primate species. The scientists found that the color saturation required for a target to be visible was very similar for monkeys and humans (Figure 1); if anything, monkeys appeared to have slightly lower color-detection thresholds (suggesting they see color better). This work, conducted predominantly by female undergraduate students, over-turns a century of dogma that monkeys have worse color abilities than humans. The research also provided behavioral evidence that the color-processing circuits downstream of the cones (so-called "post-receptoral chromatic mechanisms") are the same in the two primate species. Together, these studies cement the macaque monkey as an important model of human color perception. To address on a mechanistic level how the brain computes color, the research team used non-invasive brain imaging (fMRI) to determine the parts of the brain most responsive to color. This research yielded a surprising finding: many parts of the cerebral cortex were involved, not just the single area (called area V4) that was previously thought to be the "color area". The color-biased regions were not randomly scattered in the cerebral cortex, but rather arranged in a precise fashion suggestive of a new organizational scheme (Figure 2). The scientists tested this idea with an extensive set of fMRI experiments. The work produced an important discovery of a new organizing scheme for the part of the cerebral cortex that is responsible for object perception. The scientists found that the parts of the brain most responsive to images of faces were physically located in a systematic way next to the parts of the brain most responsive to colors, which themselves were physically located next to the parts most responsive to places. This work suggests that object perception depends upon computations carried out by multi-staged, parallel processing channels, and provides clues to how the cerebral cortex expanded during the course of evolution. Research conducted during the funding period also provides insight into how the brain generates color perception on the basis of the color signals carried by the neurons within the color-biased regions. Computational-theoretic analyses guided by prior physiological recordings suggest that the brain makes a color judgment about a stimulus by listening selectively to those neurons in the entire population that are the most active when the stimulus is presented. This work provides a powerful testable hypothesis that will guide future research. In total, research supported by this four-year grant contributed to 16 publications, 17 abstracts for professional international meetings, and 24 invited public lectures including several streamed live and archived (including a talk during Martha’s Vineyard TEDx event; and The Flame Challenge with Alan Alda at the World Science Festival). The work has had strong intellectual impact and broad impact, both from the scientific discoveries and from complementary activities such as mentorship of undergraduate women and underrepresented minorities engaged in STEM research; expansion of cross-institutional partnerships involving traditional research institutions and undergraduate liberal arts colleges; development and dissemination of undergraduate neuroscience course materials through websites developed by the principal investigator and his undergraduate collaborators; enhancement of public scientific literacy through a large number of public talks and workshops; and strengthening of connections between arts and sciences. The work provided extensive research experiences for over a dozen undergraduate students, including six women whose contributions were recognized by authorship on peer-reviewed publication and twelve women who had the opportunity to present at international scientific conferences. These experiences have provided excellent preparation for careers in medicine, biomedical research and the private sector, and promoted the advancement of women in science and technology
