The goal of this project is to investigate the perceptual and motor contributions of the superior colliculus (SC) to choice behavior;specifically, we propose to do this in the context of an urgent decision-making task that allows us to dissociate perceptual and motor performance with great, unprecedented effectiveness. Neuroscientists have intensely studied choice behavior with numerous tasks in which a perceptual judgement is made and is followed by a motor report. While highly successful, this approach has limitations. Most notably, various covert factors such as attention, anticipation, motivation or task difficulty can be traded against each other, creating ambiguities that cannot be resolved via standard psychophysical metrics, i.e., reaction time and choice accuracy. Thus, fundamental questions about the timing of choices remain poorly understood. How long does it take to make a perceptual judgement versus executing a motor action to report it? At what point in time is a subject committed to a particular option? When the success rate increases, is it because stimuli are processed more effi- ciently, or faster, or because more time is dedicated to their analysis? The PIs in this project recently developed an urgent-choice task with which these questions - and the underlying neural mechanisms - can be exam- ined in a direct, unambiguous way. It produces a new psychophysical measure, the tachometric curve, which characterizes a subject's perceptual performance over time independently of motor execution. The PIs have also constructed a computational model that relates neuronal responses to the subjects'behavior with great quanti- tative detail. In the proposed experiments, neuronal activity in the SC will be recorded from monkeys trained to perform this novel saccadic-choice task. Two problems will be addressed. First, the participation of the SC in perceptual processing, which in contrast to the SC's firmly established role in saccade execution, remains less well characterized. The unique properties of the tachometric curve will be exploited to determine when per- ceptual information informs the SC activity, and how this correlates with the temporal evolution of the subject's percept during task performance (Aim 1). In addition, subthreshold microstimulation current will be injected into the SC using various spatial and temporal configurations to determine how the artificially evoked activity impacts perceptual and motor performance (Aim 2). The second problem is how choice behavior changes when it is driven by multisensory stimuli;that is, by the simultaneous presentation of visual and auditory cues. The SC is known to respond most effectively to such stimulus combinations, but their impact during choice behav- ior has not been assessed. A multisensory variant of the task will be used to determine whether multisensory integration alters the speed of the perceptual judgements or of the saccadic motor plans (Aim 3). This work will provide critical insight about how perceptual information is dynamically translated into motor output, will characterize how multisensory integration at the neuronal level leads to enhanced behavioral performance, and will determine how the SC participates in these processes.
When a driver sees a traffic light turn red, it takes about 400 milliseconds for him to react and start stepping on the brake, but it is extremely difficult to dissociate how much of that reaction time is needed to perceive the color of the light (say, red or green) and how much is needed for preparing the corresponding motor action (a foot press). We propose a series of experiments in which such dissociation can be made very accurately, and plan to investigate how the activity of nerve cells in the midbrain contributes to the perceptual analysis of a visual stimulus and to the motor action that is triggered in response to it during a choice task. The results will help us characterize the mechanisms that the brain uses to generate choices, and will help us understand numerous pathologies associated with impairments in such mechanisms, including schizophrenia, Parkinson's disease and unisensory neglect.