Cortical Neurons Encoding Path and Place: Where You Go Is Where You Are Michael T. Froehler and Charles J. Duffy

To characterize the contributions of visual and nonvisual cues, we presented optic flow video simulations on a sled-mounted rear-projection display while recording the activity of MST neurons. Translational sled movement, on the same circular paths used in the preceding experiments, was presented with the optic flow video display or in darkness (no visual stimulus other than the stationary fixation point). When translational movement accompanied a visual display, the optic flow simulated the same direction and speed of movement that was being presented by the sled. The optic flow displays contained white dots on a black background and moved to simulate the visual scene during observer self-movement either in front of a single wall or a three-dimensional cloud (Fig. 1). In the wall display, the number of dots distributed on the screen decreased to simulate movement toward the wall and increased to simulate movement away from the wall; a dot-density distance cue. In the cloud display, dot speed increased as they came closer to the observer in the simulation, and their speed decreased as they went further from the observer in the simulation; a motion parallax distance cue.

Supplemental Figure 1. The wall display (A) simulated the visual motion pattern presented by movement in the room; movement in front of a stationary array of small, white lights. The number of dots in the display varied as a dot-density distance cue, decreasing with simulated movement toward the wall. The cloud display (B) used motion parallax as a distance cue. The dots occupied a volume in front of the monkey, and the relative motion of near and far dots varied as a function of observer distance. The number of dots on the screen was approximately triple that of the wall display, and dot density did not change throughout the simulation. We recorded the activity of 44 neurons during randomly interleaved presentations of both video displays, video displays plus translational sled movement, or sled movement alone. Responses to the cloud optic flow video display show the same amplitudes and directionality as responses to the wall video display but were stronger and more selective for heading direction than responses to sled movement in darkness (Fig. 2, A and B). Responses to the cloud and wall video displays were also not substantially changed by the addition of directionally congruent sled movement (Fig. 2, C and D). Thus, neither the three-dimensional depth cues in the cloud and wall displays nor the sled movement stimuli had much effect on the responses.

Medium version | Full size version

Supplemental Figure 2. Cloud video displays evoked responses comparable to wall displays in amplitude [(A): slope = 0.97, r² = 0.91] and directionality [(B): slope = 1.02, r² = 0.94], but stronger than sled movement alone [(A): slope = 0.34, r² = 0.56; (B): slope = 0.10, r² = 0.33]. Combined video-plus-movement stimuli yielded responses similar to video alone for wall [(C): slope = 1.06, r² = 0.89; (D): slope = 0.89, r² = 0.84] and cloud [(C): slope = 0.93, r² = 0.84; (D): slope = 0.96, r² = 0.81] displays.

Medium version | Full size version