Differences in gaze anticipation for locomotion with and without vision

Abstract

Previous experimental studies have shown a spontaneous anticipation of locomotor trajectory by the head and gaze direction during human locomotion. This anticipatory behavior could serve several functions: an optimal selection of visual information, for instance through landmarks and optic flow, as well as trajectory planning and motor control. This would imply that anticipation remains in darkness but with different characteristics. We asked 10 participants to walk along two predefined complex trajectories (limaçon and figure eight) without any cue on the trajectory to follow. Two visual conditions were used: (i) in light and (ii) in complete darkness with eyes open. The whole body kinematics were recorded by motion capture, along with the participant's right eye movements. We showed that in darkness and in light, horizontal gaze anticipates the orientation of the head which itself anticipates the trajectory direction. However, the horizontal angular anticipation decreases by a half in darkness for both gaze and head. In both visual conditions we observed an eye nystagmus with similar properties (frequency and amplitude). The main difference comes from the fact that in light, there is a shift of the orientations of the eye nystagmus and the head in the direction of the trajectory. These results suggest that a fundamental function of gaze is to represent self motion, stabilize the perception of space during locomotion, and to simulate the future trajectory, regardless of the vision condition.

Keywords: OKN; anticipation; eye movement; human locomotion; nystagmus.

Figure 1

Apparatus, trajectories, and definition of coordinates and studied body segments. (A) Left: the eye-tracker camera used in the experiment. Right: the fabric screen used to prevent any source of light during the experiment. (B) Two trajectories were shown and required to be followed by the participants: a limaçon (left) and an elongated eight-like shape (right). (C) Chosen reference frames (gray): laboratory (R_L), head (R_H) and trajectory (R_Tr). The considered segments are: gaze (blue), head (gray dotted line), shoulders (green), torso (yellow), pelvis (orange). (D) Direction of the various segments in the horizontal plane, with respect to the trajectory and its tangent. (E) Definition of the studied eye movements (visual axis, blue), in azimuth and yaw.

Figure 2

Typical trajectories followed by participants. Trajectory of the pelvis of participant n°1 in the ground plane, for different experimental conditions [(A) eight shape in light, (B) eight shape in darkness, (C) limaçon in light, (D) limaçon in darkness]. The four trials are presented (in order of realization: red, green, blue, black).

Figure 3

Angle of gaze (blue), head (gray), torso (yellow) and pelvis (orange) while walking on an elongated eight shape (top) and a limaçon (bottom) for the same participant. The left and right figures correspond to light and darkness conditions, respectively. Blowups (X5) are shown as well.

Figure 4

Spatial and temporal anticipation of the body segments. (A,B) Averaged anticipations in both visual conditions are represented (left: in light, right: in darkness). All differences are statistically significant (significance threshold: p < 0.05). The vertical bars represent the between-participant standard deviation. (A) Spatial anticipation of gaze (blue) and head (gray) with respect to the tangent to the pelvis orientation. (B) Time shift of the different body segments with respect to the pelvis movement. (C,D) Average spatial anticipation (C) and time shift (D) of the body segments for each participant. Each color corresponds to one participant.

Figure 5

Elevation of the eye and the head. Average elevations of the head (A) and of the eye in the head (B) depending on the visual condition. Negative values indicate a downward orientation (below horizon for the head and below primary eye position for the eye). Large head and eye elevations lead to a shorter distance between gaze/floor intersection and the projection onto the ground of the right orbit [i.e., gaze distance, (C), (D)].

Figure 6

Rotations of the eye in the orbit of one participant walking along the limaçon in light (left) and in darkness (right). The upper part of the figure corresponds to horizontal movements and the lower part to vertical movements.

Figure 7

Cumulative distribution of eye quick phases amplitude. 85% of the quick phases have an amplitude lower than 14°. We did not observe any change in amplitude between visual conditions.

Figure 8

Distributions of the eye position with respect to the head at the beginning (A) and at the end (B) of quick phases. The gray spectrum represent data taken in light and the black spectrum the data in darkness. The dashed vertical lines indicate the median of the distributions.

Figure 9

Characteristics of reverse-saccades occurrence. The reverse-saccade occurrence depends on different parameters for (A) elongated eight and (B) limaçon shapes. (A) Distributions of reverse-saccades as a function of the time onset from one of the two curvature sign inversion present in the extended height shape (black and gray for darkness and light conditions, respectively). A large proportion of reverse-saccades are elicited in a period of 2 s before the curvature sign inversion. Stars (^*) indicate the bins with a significant difference (i.e., with a value higher than the statistical error $\sqrt{N}$). (B) Distribution of reverse-saccades occurrence as a function of the evolution of participant's progression on the overall limaçon shape. (C) The quick phases (without reverse-saccades) distribution as a function of the participant progression on the trajectory is also presented, for darkness (black) and light (gray) conditions. On the limaçon, a large proportion of reverse-saccades are elicited in the last 10% of the trajectory (39.13% and 23.49% of the reverse-saccades in light and darkness conditions, respectively). Stars (^*) indicate the bins with a significant difference.

Figure 10

Gains in space of nystagmi slow phases. Two different gains were considered, both in the horizontal plane: a “fixation” gain corresponding to the tendency of gaze during a slow phase to stay aligned with a location on the ground (left part of the figure), and a “parallel” gain assessing the stability of gaze direction in the horizontal plane (right part of the figure). Both gains were computed from a ratio between the real gaze direction (β_r, from the head direction) and a predicted gaze direction (β_p, from the head direction). The prediction could either be to consider gaze directed toward a point on the ground determined from the previous experimental gaze direction (bottom left) or to consider a gaze parallel to its previous direction (bottom right). A close-to-one gain means that the predicted and real angles were close to each other, as a gain larger than one means that the real angle is wider than the predicted angle. These gains were computed for limaçon (top) and elongated eight (middle) shapes, and separately in darkness (black) and in light (gray). Green and red segments represent gaze direction at the beginning and at the end of a slow phase, respectively.