Kenneth H. Britten - US grants

DESCRIPTION (provided by applicant): The basic "building blocks" of visual perception are starting to become reasonably well understood, and we can make a fairly good account of how simple discriminations are done. What we understand much less is how the visual system solves more realistic, everyday challenges. Visually guided navigation is a particularly good "model system" for studying real-world visual processes in the laboratory. The perception of self-motion from the pattern of motion on the retina has been studied extensively, though we still know very little about where in the brain the critical processing steps occur, and how the complex pattern of motion is converted into effective movement. The present proposal seeks to answer these open questions. First, we will seek direct evidence for the involvement of multiple cortical areas in the perception of self-motion, by using multiple, simultaneous recording techniques while our experimental animals are performing a discrimination of self-motion direction. Secondly, we will seek to ask if the parietal cortical area (the ventral intraparietal area or VIP) is both necessary for self-motion perception and is actually used. We will do this by perturbing the pattern of activity in VIP in the context of the self-motion task, both by reversible inactivation as well as by electrical activation. These complementary methods should greatly extend our understanding of how the parietal cortex participates in self-motion perception. However, to really extend our knowledge of self-motion perception, we need to extend the inquiry into a more active context. Human-factors studies have revealed that guidance of self-motion ("steering") is a very active process, with the direction of gaze being a critical component. However, next to nothing is known about the central nervous system mechanisms used in this active task. So, we propose to establish, characterize and exploit an animal model of active locomotion to study the involvement of brain structures in this task. We will train our subject to direct their "virtual" trajectories by joystick, and characterize how their normal behavior is influenced by cues including target direction, gaze direction, gaze velocity, and visual motion information. We will then record activity in multiple cortical areas while animals are engaged in this task, and explore the signals in visual and parietal cortex to better understand brain mechanisms of visually guided navigation. This information, in the long term, might be useful in helping the disabled to navigate, and in the development of visual prosthetics for the blind.

psychology; eye; Primates; Mammalia; behavioral /social science research tag;

Significance Better understanding of how our visual system develops, adapts, and integrateswith our other sensory systems. Objectives One of the central problems of visual science is to establish how the numerous areas together form our seemingly unified visual experience. In my work, I attempt to decipher the mechanisms and consequences of the hierarchical connection between two specific areas in the visual cortex of the rhesus macaque (M. mulatta). These two areas (MT and MST) form a perfect example of serial processing, and we already know a great deal about how they are orgainized and what kind of information they represent. Both are specialized for the analysis of visual motion, but MST (the later stage of processing) appears to represent much more complex aspects of image motion, such as rotation, expansion, and contraction. Using a combination of behavioral and physiological approaches, we are both exploring the mechanisms by which the simpler representation in MT is transformed into the more complex one in MST, and testing the hypothesis that the more complex representation in MST is uniquely involved in more complex visual tasks. To do this, we train monkeys on tasks demanding the analysis of visual motion, and precisely measure the signals in these cortical areas with single cell recording techniques, or alternatively, we perturb the signals in these areas using very small electrical currents, and measure the effects on the animals behavior. Results We have been exploring a specific complex motion task, recovering self-motion from image motion, and have recently demonstrated large effects of microstimulation of both areas MT and MST on this task. This shows that higher areas in cortex are directly involved in this task. Future Directions Future experiments will be probing the mechanisms of this involvement.

parietal lobe /cortex

motion perception; neural information processing; visual depth perception; Primates; animal colony; behavioral /social science research tag; clinical research;

Abstract Not Provided.

Area; Attention; Brain; CRISP; Cell Nucleus; Computer Retrieval of Information on Scientific Projects Database; Dorsal; Electrodes; Encephalon; Encephalons; Evolution; Funding; Grant; Institution; Investigators; Lesion; Location; Mammals, Primates; Medial; NIH; National Institutes of Health; National Institutes of Health (U.S.); Neocortex; Nerve Cells; Nerve Unit; Nervous System, Brain; Neural Cell; Neurocyte; Neurons; Nucleus; Physiologic; Physiological; Primates; Pulvinar; Pulvinar structure; Recurrence; Recurrent; Regulation; Research; Research Personnel; Research Resources; Researchers; Resources; Role; Source; Structure; Suggestion; Testing; Thalamic Nuclei; Thalamic structure; Thalamus; Thinking; Thinking, function; United States National Institutes of Health; Visual Cortex; Visual attention; directed attention; directs attention; extrastriate; extrastriate visual cortex; homotypical cortex; isocortex; neopallium; neuronal; pulvinar thalami; response; sensory cortex; size; social role; success; thalamic; visual cortical

DESCRIPTION (provided by applicant): The principal aims of this project are to understand the perceptual and physiological mechanisms that allow primates, including man, to navigate using visual motion information. Previous work has elucidated the mechanisms of the passive perception of simulated self-motion, but this is a far cry from actually using motion information t guide an ongoing trajectory. There are two primary cues used in visual navigation: the position of targets (and obstacles) and the optic flow information that provides instantaneous feedback about one's immediate trajectory in the world. The experiments in this proposal wil test how motion-sensitive neurons in the higher levels of visual cortex process, combine, and represent these two cues for the guidance of ongoing behavior. The physiological experiments will be directed at two areas at the highest levels of the motion system in dorsal extrastriate cortex: the medial superior temporal area (MST) and the ventral intraparietal area (VIP). We will test how the signals in these areas correlate with performance on a recently developed active steering task. In this task, nonhuman primates control a virtual trajectory presented in a virtual reality setting. This allows precise control of the cues and also enables the physiological experiments that form the core of the proposal. The specific hypotheses under test in the physiological experiments in Aims 1 and 2 concern the mechanisms by which these signals support performance of a complex, natural task. In Aim 3, we will be developing a biologically realistic, experimentally-constrained computational model of how the nervous system controls steering behavior. Successful conclusion of these aims will allow us to understand both where in the brain such guidance is supported, and how the signals there relate to behavioral capabilities. This information is important from a pure scientific perspective, and will also potentially inform the creation of visual prosthetics to assist navigation by visually impaired patients, and could help the development of improved robotic navigation systems.