Abstract
Smooth pursuit combines retinal motion with extra-retinal signals to track a moving object in the world (Young et al., 4th Annual NASA Manual 1968). While the retinal motion signals driving pursuit initiation have been well characterized (e.g., Lisberger and Westbrook, 1985), the existence of extraretinal signals related to ongoing eye motion in the visual motion pathways involved in driving pursuit (e.g., Newsome et al., 1985) makes the nature of the visual command signal less clear. Here, we use the pursuit “oblique effect”, a directional anisotropy, as a fingerprint to examine the coordinate system of the visual motion signals driving the onset of pursuit.
Methods. With their head upright or tilted to the right, observers were asked to pursue a small spot that made an initial step from fixation at a random angle, then moved back through fixation at one of four speeds (5, 10, 20, 30 deg/s) in a Rashbass design. De-saccaded eye-velocity responses were used to measure pursuit direction (near open-loop interval 200–300 ms after motion onset) and to compute direction gain (Krukowski & Stone, Neuron 2005). To compute the head and eye tilt, we used pairs of high-resolution digital photographs in the upright and tilted configurations, and measured head rotation using selected features and eye rotation using a circular cross-correlation of the iris. Results. When the head tilts by 19–21 degrees, the oblique effect for pursuit changed orientation (mean shift: 15.7 ± 1.9 deg, p0.05).
Conclusion. The neural signals for pursuit initiation are linked to an eye-centered reference frame (not world- or head-centered coordinates).
NASA Space Human Factors Engineering.