A macaque was trained to fixate while viewing a grid of dots that moved within a circular aperture for 800 msec. The dots were arrayed in evenly spaced columns. At fixed intervals, all the dots were displaced perpendicular to the columns by a fraction of the inter-column spacing, creating a uniform apparent motion. If the displacement was less than half the spacing of the columns, the direction of the apparent motion was unambiguous, but if the displacement was exactly half the spacing, the direction was perceptually bistable from trial to trial. After a 200-msec delay, the animal viewed a second motion presentation (S2), and then indicated with a two-way lever whether the direction was the same or opposite that of the first presentation (S1). The direction of S2 was always unambiguous. Thus, the animal's perceived direction during S1 could be inferred - Even if S1 was perceptually bistable. The animal was rewarded at random on perceptually bistable trials (50%). On each trial, the direction and fractional displacement for S1 were randomly chosen. On average, S1 was unambiguous on 6/7 trials, to ensure that the animal reported his perception accurately on the remaining, perceptually bistable trials. Neuronal recordings were made in parietal areas MT, MST, and LIP. For each unit, the stimulus grid was placed in the receptive field, oriented such that motion would be in either the preferred or null direction. Neuronal responses were sorted with respect to the animal's report of direction in S1. For all three areas most units were direction selective if the direction was unambiguous. However, on bistable trials 42/82 LIP cells were significantly modulated by the animal's report of direction, while only 9/55 MST cells and 0/34 MT cells were significantly modulated (t-test; p<0.05). All modulated units had larger responses for preferred-direction reports. These data support a transition within parietal cortex toward a representation that reflects subjective perception.
ASJC Scopus subject areas
- Sensory Systems