(Project 4: The structure of olfactory neural and perceptual spaces) It is not clear how stimulus-driven neural activity leads to perception. While this problem is typically approached from a feed-forward perspective ? stimuli are systematically varied as neural activity and perception are monitored ? the olfactory system poses a particular challenge, as the molecular stimulus space that drives neural responses is not well defined. Here we suggest an alternative approach: rather than working forward from receptors in the nose, we propose to work backwards from perception. That is, we will take advantage of a variety of conceptual and technical innovations to construct a new type of ?ruler? that will allow us to accurately measure perceptual distances between odors, and then use this ruler to comprehensively map out the global structure of perceptual odor space, first in humans and then in mice. In parallel we will take advantage of advances in neural imaging technologies that give us, for the first time, access to cortical responses to odorants in awake mice. Important preliminary data demonstrates that the piriform cortex harbors a systematic and invariant representation of odor space, and that this space correlates with perceptual relationships amongst odors. By assessing both cortical and perceptual responses to the same odorant set, these experiments will reveal how information about odorant chemistry is organized in cortex, and how this information relates to perception, both in mice and in humans. Constraining the dimensionality and structure of neural and perceptual odor spaces, in a systematic and comprehensive fashion, will reveal key aspects of the input-output function of the olfactory system.
