Neuroscience Letters 339 (2002) 243–247 www.elsevier.com/locate/neulet

Spatial invariance in anticipatory orienting behaviour during human navigation Pascal Pre´vosta,*, Ivanenko Yurib, Grasso Renatob, Berthoz Alaina a

LPPA, Colle`ge de France, 11 place Marcelin Berthelot, 75005 Paris, France Section of Human Physiology, Scientific Institute S. Lucia, 00179 Rome, Italy

b

Received 10 September 2002; received in revised form 25 November 2002; accepted 25 November 2002

Abstract We have recently reported that the head systematically deviates toward the future direction of the trajectory about 500 ms before attaining a turning point of 908 corner trajectories both in light and in darkness. Here, we investigated how this anticipatory strategy is modified whilst varying visual conditions (Experiment 1) and walking speed (Experiment 2). Exp. 1 showed similar anticipatory behaviour when walking with or without vision. Exp. 2 (that varied walking speed; eyes open) showed that the head started to deviate at a constant distance rather than at a constant time to the corner. The results appear inconsistent with optic flow theories of the guidance of walking direction and might highlight the role of landmarks and/or egocentric direction in anticipatory orienting behaviour. q 2002 Elsevier Science Ireland Ltd. All rights reserved. Keywords: Locomotion; Steering; Head movements; Anticipation; Feed-forward control; Reference frame

One of the main tasks of the brain is to anticipate future events. This behaviour is obviously present in many motor tasks. In everyday life, when one plans a goal-oriented locomotor path, the brain has to use information relative to the path and relative to the place where one has planned to go to. In this locomotor task, several kinds of information are used (visual, vestibular, proprioceptive and efferent copy of the motor command) to control all sequences of the trajectory towards the objective. In the event, ascending pathways conveying multi-sensory information attain specific areas in the brain related to the representation of space [19]. Although there is no denying that vision is very important in navigation, it is not the only source of spatial information. The vestibular system provides a significant contribution to the orientation and localisation of the body in space during displacement [3,4]. The head has to be controlled relative to space in order to guarantee stabilisation of vision and reconstruction of our environment. This facilitates the interpretation of external (mainly visual) and internal (inertial and proprioceptive) cues during locomotion [16]. This points out the interest of investigating predictive * Corresponding author. Tel.: þ 33-1-4427-1298; fax: þ33-1-4427-1382. E-mail address: [email protected] (P. Pre´vost).

mechanisms in navigational tasks, which are essential to the feed-forward control of displacements during goal-oriented locomotion. In this context, it has been recently demonstrated that the head anticipates the changing direction of the entire body during circular [5] and curvilinear trajectories [3,4,7]. On the basis of these observations, we suggested [7] that the main purpose of anticipatory head-orienting strategies is to provide a change of reference frame by acquiring in advance information about the environment in the new direction of heading. Both optic flow [13] and egocentric direction [18] might guide locomotion. If optic flow has a major contribution to steering manoeuvres, then the dynamics of head movements would be significantly affected by the absence of visual information and vice versa. We demonstrated previously [7] that the head and eyes deviate in the direction of turn even in darkness, however, we did not characterise these dynamics quantitatively. Here, to assess the contribution of visual information for triggering orienting movements, we tested whether the onset and the magnitude of head deviation depends on visual conditions (Experiment 1). Furthermore, to understand which features of the optic flow or of the trajectory formation drive steering, we varied walking speed. Time to contact theories [13] would reasonably predict that subjects start orienting reactions at progress-

0304-3940/02/$ - see front matter q 2002 Elsevier Science Ireland Ltd. All rights reserved. doi: 1 0 . 1 0 1 6 / S 0 3 0 4 - 3 9 4 0 ( 0 2 ) 0 1 3 9 0 - 3

244

P. Pre´vost et al. / Neuroscience Letters 339 (2002) 243–247

ively earlier locations as the speed of motion increases. Yet, the experiment in a virtual reality environment [7], where subjects were passively moved along a multi-legged virtual corridor at different speeds, indicates that spatial rather than temporal features may be used as thresholds for triggering motor acts. However, passive transport along a virtual corridor in standing subjects could differ substantially from what happens in a real locomotor task. For example, there is no active translation of the body, and the pattern of inertial forces acting on the body is different from that occurring in real locomotion. Therefore, in the present study we investigated a spatio-temporal pattern of head orienting movements by performing an actual walking task at different speeds (Experiment 2). Twelve naive subjects (eight men – four women, age ¼ 33 ^ 3) participated in this experiment. They were asked to perform a navigational task similar to that observed in everyday life while turning around a street corner. It consisted in walking towards an objective along a 908 corner trajectory from an initial position to a final one (materialised on the floor with drawn crosses). Starting from the same position (6 m prior to the corner) in each trial, subjects were instructed to stop exactly on the end-point (6 m after the corner) after turning around an obstacle (tripod, 1.8 m high). No practice trials were allowed. Subjects had to wait 2 min before doing a trial. Subjects were divided in two groups (of four men and two women) to perform the task in two protocols. The first group (Exp. 1) realised a left and right turn in the two visual conditions (blindfolded eyes, EC, and eyes open, EO). In this case, no specific instructions were given with regard to the walking speed, thus subjects were expected to walk at ‘normal’ (‘natural’) speed. The second group (Exp. 2) performed a left turn condition at three different (randomised) walking speeds (SLOW, NORMAL and FAST) with EO. In NORMAL condition, the subject was asked to walk at his/her natural speed (it was 1.20 ^ 0.05 m/s (mean ^ SEM)). In SLOW and FAST conditions, the subject was instructed to walk about 1.5 times slower or faster, respectively. Velocities in all conditions were within the normal range of walking (0.4 – 1.8 m/s). Each subject performed five trials in each experimental condition. The general procedure of the experiment has been described previously [5,7]. Briefly, the instantaneous head direction (a) and walking direction (b) were computed by using instantaneous position of two head markers (recorded by means of an ELITE system) fixed on a helmet along the naso-occipital axis of the head. By subtracting those two directions, we obtained head angular deviation relative to the walking direction, that we called theta angle (u) (Fig. 1A). Time series of the relevant kinematic variables were reconstructed at the time resolution of 0.01 s. Kinematic data were low-pass filtered by means of a linear-phase digital filter with a cut-off frequency of 5 Hz. Due to lateral trunk displacements (within one step

Fig. 1. Methods. (A) Schematic projection of the head position (midpoint of the two markers on the helmet) in the horizontal plane in two successive points ( pt, ptþ1). a angle corresponds to the head direction in space, whereas b gives the direction of walking using two successive head positions. u corresponds to the angle between those two directions (a 2 b). (B) Identification of the interception point between initial rectilinear part of the path before turning and the line dropped from the tripod position perpendicular to this direction. Head orientation is represented by sticks (every 100 ms) superimposed on the trajectory drawn by the head midpoint.

cycle), head orientation oscillates relative to the walking trajectory (b angle) step by step by about ^ 4 –88, while in space (a angle) it oscillates to a much lesser extent (less than 28). Therefore, we used an absolute angle of head orientation (a) in space to characterise the onset of head deviation relative to the initial linear portion of the curvilinear trajectory. Special care was taken to put the markers along the head naso-occipital axis. Indeed, the mean head orientation at the beginning of walking was 1 ^ 0.88. We computed both the distance and the time when subjects started to reorient their head relative to an absolute reference. Head movement was considered to start when head rotation (a) exceeded 98 toward the direction of turn. Time and distance of onset were calculated relative to the point when subjects passed the intersection point (IP) between rectilinear part of the path before turning and the line perpendicular to this direction (Fig. 1B). All parameters were then averaged (across five trials) for each subject. Two way ANOVA with repeated measures and Newman –Keuls post-hoc test (when significant main effects were found) were used to analyse the effect of turning direction and walking speed in both conditions of vision,

P. Pre´vost et al. / Neuroscience Letters 339 (2002) 243–247

with a significant P value fixed at 0.05. Results are reported as mean ^ 1 SEM. In Exp. 1, we studied the influence of visual information (EO versus EC) on anticipatory head orientation behaviour. The walking speed was slightly higher in EO than EC condition (1.20 ^ 0.03 and 1.02 ^ 0.03 m/s, respectively; P ¼ 0:019). However, neither the onset nor the maximal angle of head deviation depended on turning direction (P . 0:6) or visual condition (P . 0:4) (Fig. 3A). In Exp. 2, we studied the effect of walking speed. Mean walking speeds were the following: 0.8 ^ 0.07, 1.2 ^ 0.05 and 1.6 ^ 0.06 m/s at slow, normal and fast speed conditions. Fig. 2 is an example of the trajectory and orientation of the head in each velocity condition in one subject. The big open circle represents the tripod and the small field circle the distance at which head deviation occurs. The head starts to deviate before the corner position is attained and realigns with walking direction afterwards. As it can be seen, the head deviates towards the inner part of the curved trajectory in a similar way in all three speed conditions. The analysis of the maximum angle of head deviation relative to the trajectory (u) showed that head deviated to the same extent (F2;10 ¼ 0:98; P ¼ 0:39) whatever the walking speed (Fig. 3B, right panel). In contrast, the onset of head deviation displayed a significant effect of speed (F2;10 ¼ 14:86; P , 0:001). Post-hoc analysis revealed that the time to the IP decreased as the walking speed

245

Fig. 2. Representative data of head orientation in one subject at slow, normal and fast locomotion speed. The vertical line represents the perpendicular from the tripod position to the initial linear part of the path. The field circle corresponds to the distance at which the head starts to deviate.

Fig. 3. Mean (^SEM) values of computed parameters in each experiment (A) Effects of vision: condition of vision has no effect on any parameter. (B) Effects of speed: time of head deviation onset to intercepting point is influenced by locomotion speed. On the contrary, onset of head deviation occurs at the same distance relative to this point.

246

P. Pre´vost et al. / Neuroscience Letters 339 (2002) 243–247

increased (all P values , 0.05). Thus, the head started to deviate earlier for slow velocities (Fig. 3B, left panel). However, there was no significant effect of speed on the distance to the IP (F2;10 ¼ 0:57; P . 0:05) (Fig. 3B, middle panel). On average, the onset of head deviation occurred at about 0.3 ^ 0.3 m distance to the interception point (Fig. 1B) or 1.1 ^ 0.2 m relative to the new direction of motion (relative to the IP between rectilinear part of the path before and after the corner). The initial portion of the trajectory before the onset of head deviation was linear. Indeed, the direction of the walking trajectory veered on average only by 1.6 ^ 1.28 at the moment when the head started to deviate. In addition, the results did not depend significantly on the threshold (58 or 98) we used for calculating the onset of head deviations. When we used the 58 threshold, obviously the onset occurred earlier (0.5 ^ 0.3 m to the interception point instead of 0.3 ^ 0.3 m for 98 threshold) but displayed the same tendency with speed. Thus, the spatial structure of head orientations seemed to be conserved across different walking speeds in the range 0.8 –1.6 m/s. Our findings confirm previous results [5,7,15] and show that head is controlled with respect to the direction of movement: it steers the future direction of walking. This is observed whatever the condition of vision, turn, or speed. So this anticipatory strategy can be considered as stable and reproducible, i.e. an invariant characteristic of human locomotion. This behaviour is an important aspect of navigation because the head contains essential tools (eyes, neck muscle proprioception, vestibular system) used to obtain information about the place to which we are going and information used to dynamically control gait during locomotion. Indeed, a close relation has been demonstrated [1,10,11] between eye and head movements that are implicated in the control of body movements in space. Furthermore, gaze orientation strategies during locomotion are present both in light and in darkness [7] and may reflect the need to prepare a stable viewer-centred reference frame for navigation. Since, in each visual condition, the subject saw a goal and an obstacle prior to performing the task, the lack of influence of having the eyes closed (Fig. 3A) could be explained by the use of a short-term spatial memory in which information about environment is updating during locomotion in a egocentric manner [17]. Similar to what happens in car driving [12] and in line with the results recently reported in virtual environment [6], we have found that spatial cues drive predictive orienting head movement during real navigation. Head deviation is initiated at a constant distance, which is in opposition with a time-constant strategy [13]. As a consequence, head deviation onset occurred earlier relative to turning point when locomotion speed decreased. The synergies between the head and the rest of the body still remain unclear but a dynamic model of visual guiding of locomotion with heading information could partly explain our results [20]. Nevertheless, this model does not explain why we have

observed the same anticipation without vision. We propose that, in this blindfolded navigation, the subjects relied more on proprioceptive and vestibular information (and efferent copy of the motor program), as already suggested in previous studies [2,9]. They could refresh the internal representation about their position as well as the position of the target and adopt the same strategy for guiding head orientation during the navigation task. The results confirm the hypothesis we have proposed in our previous study [5,7] about the ‘go where we look’ strategy. In accordance with previous studies [14], our findings emphasize that head control in space is essential for stabilizing the reference frame captured by the gaze and for interpreting sensory information. This proposal is also supported by recent results about gaze behaviour associated with changing locomotion direction [8]. Spatial invariance of head orienting movements appears to be inconsistent with popular (optic flow model) theories concerning the role of vision in the guidance of walking direction and might highlight the role of landmarks and/or egocentric direction [17] in anticipatory orienting behaviour.

Acknowledgements This research has been performed in the frame of the European Laboratory of Neurophysiology and Neuropsychology of Action (LENNA) and was partly supported by the French Space Agency. This article is dedicated to Renato Grasso who died in October 2000.

References [1] C. Andre-Deshays, A. Berthoz, M. Revel, Eye-head coupling in humans: I, simultaneous recording of isolated motor units in dorsal neck muscles and horizontal eye movements, Exp. Brain Res. 69 (1988) 399 –406. [2] A. Berthoz, I. Israel, P. Georges-Francois, R. Grasso, T. Tsuzuku, Spatial memory of body linear displacement: what is being stored?, Science 269 (1995) 95–98. [3] S. Glasauer, M.A. Amorim, I. Viaud-Delmon, A. Berthoz, Differential effects of labyrinthine dysfunction on distance and direction during blindfolded walking of a triangle path, Exp. Brain Res. 145 (2002) 489 –497. [4] S. Glasauer, M.A. Amorim, E. Vitte, A. Berthoz, Goal-directed linear locomotion in normal and labyrinthine- defective subjects, Exp. Brain Res. 98 (1994) 323– 335. [5] R. Grasso, S. Glasauer, Y. Takei, A. Berthoz, The predictive brain: anticipatory control of head direction for the steering of locomotion, NeuroReport 7 (1996) 1170–1174. [6] R. Grasso, Y.P. Ivanenko, J. McIntyre, I. Viaud-Delmon, A. Berthoz, Spatial, not temporal cues drive predictive orienting movements during navigation: a virtual reality study, NeuroReport 11 (2000) 775 –778. [7] R. Grasso, P. Prevost, Y.P. Ivanenko, A. Berthoz, Eye-head coordination for the steering of locomotion in humans: an anticipatory synergy, Neurosci. Lett. 253 (1998) 115– 118. [8] M.A. Hollands, A.E. Patla, J.N. Vickers, “Look where you’re going!”:

P. Pre´vost et al. / Neuroscience Letters 339 (2002) 243–247

[9]

[10]

[11]

[12] [13] [14]

gaze behaviour associated with maintaining and changing the direction of locomotion, Exp. Brain Res. 143 (2002) 221– 230. Y.P. Ivanenko, R. Grasso, F. Lacquaniti, Neck muscle vibration makes walking humans accelerate in the direction of gaze, J. Physiol. 525 (3) (2000) 803–814. Y.P. Ivanenko, I. Viaud-Delmon, E. Mayer, N. Valenza, J.M. Annoni, A. Rohr, J.P. Guyot, A. Berthoz, T. Landis, Lack of anticipatory gazeorienting responses in patients with right brain damage, Neurology 54 (2000) 1656–1661. Y.P. Ivanenko, I. Viaud-Delmon, A. Semont, V.S. Gurfinkel, A. Berthoz, Eye movements during torso rotations in labyrinthinedefective subjects, J. Vestibular Res. 9 (1999) 401 –412. M.F. Land, D.N. Lee, Where we look when we steer, Nature 369 (1994) 742–744. D.N. Lee, A theory of visual control of braking based on information about time-to-collision, Perception 5 (1976) 437–459. J. Paillard, Les de´merminants moteurs de l’organisation spatiale, Cahiers de Psychologie 14 (1971) 261– 316.

247

[15] A.E. Patla, A. Adkin, T. Ballard, Online steering: coordination and control of body center of mass, head and body reorientation, Exp. Brain Res. 129 (1999) 629–634. [16] T. Pozzo, A. Berthoz, L. Lefort, Head stabilization during various locomotor tasks in humans: I, normal subjects, Exp. Brain Res. 82 (1990) 97–106. [17] S.K. Rushton, J.M. Harris, M.R. Lloyd, J.P. Wann, Guidance of locomotion on foot uses perceived target location rather than optic flow, Curr. Biol. 8 (1998) 1191–1194. [18] S.K. Rushton, J.M. Harris, J.P. Wann, Steering, optic flow, and the respective importance of depth and retinal motion distribution, Perception 28 (1999) 255– 266. [19] J.S. Taube, J.P. Goodridge, E.J. Golob, P.A. Dudchenko, R.W. Stackman, Processing the head direction cell signal: a review and commentary, Brain Res. Bull. 40 (1996) 477– 484. [20] W.H. Warren Jr, B.A. Kay, W.D. Zosh, A.P. Duchon, S. Sahuc, Optic flow is used to control human walking, Nat. Neurosci. 4 (2001) 213–216.