Hysteresis of haptic vertical and straight ahead in healthy human subjects

Abstract

Background: The subjective haptic vertical (SHV) task requires subjects to adjust the roll orientation of an object, mostly in the roll plane, in such a way that it is parallel to perceived direction of gravity. Previously we found a tendency for clockwise rod rotations to deviate counter-clockwise and vice versa, indicating hysteresis. However, the contributing factors remained unclear. To clarify this we characterized the SHV in terms of handedness, hand used, direction of hand rotation, type of grasping (wrap vs. precision grip) and gender, and compared findings with perceived straight-ahead (PSA). Healthy subjects repetitively performed adjustments along SHV (n = 21) and PSA (n = 10) in complete darkness.

Results: For both SHV and PSA significant effects of the hand used and the direction of rod/plate rotation were found. The latter effect was similar for SHV and PSA, leading to significantly larger counter-clockwise shifts (relative to true earth-vertical and objective straight-ahead) for clockwise rotations compared to counter-clockwise rotations irrespective of the handedness and the type of grip. The effect of hand used, however, was opposite in the two tasks: while the SHV showed a counter-clockwise bias when the right hand was used and no bias for the left hand, in the PSA a counter-clockwise bias was obtained for the left hand without a bias for the right hand. No effects of grip and handedness (studied for SHV only) on accuracy were observed, however, SHV precision was significantly (p < 0.005) better in right-handed subjects compared to left-handed subjects and in male subjects.

Conclusions: Unimanual haptic tasks require control for the hand used and the type of grip as these factors significantly affect task performance. Furthermore, aligning objects with the SHV and PSA resulted in systematic direction-dependent deviations that could not be attributed to handedness, the hand used, or the type of grip. These deviations are consistent with hysteresis and are likely not related to gravitational pull, as they were observed in both planes tested, i.e. parallel and perpendicular to gravity. Short-term adaptation that shifts attention towards previous adjustment positions may provide an explanation for such biases of spatial orientation in both the horizontal and frontal plane.

Figure 1

Photographs in the three principle planes (frontal, sagittal, and axial) for both the SHV (panels A-C) and the PSA (panels D-F) task when holding the tactile bar that was positioned in front of the subject in the median sagittal plane. Rotations in the frontal plane for the SHV (panel A, frontal view) correspond to rotations in the axial plane for the PSA (panel D, top view). The view from top for the SHV (panel B, axial plane) matches the frontal view for the PSA (panel F, frontal plane) and the view from the side for the SHV (panel C, sagittal plane) parallels the side view for the PSA (panel E, sagittal plane).

Figure 2

Photographs of the two tactile devices (panel A: tactile rod; panel C: tactile plate) used in experiment 1 and 2. For the tactile rod, firm grasp using a wrap grip with the entire palm is mandatory (panel B), whereas for the tactile plate slight touch with the fingertips is needed (panel D). Grasping either device was done in such a way that the Velcro is located between the thumb and the index finger, as shown in Figure 2B (Velcro partially covered by the grasping hand) and Figure 2D.

Figure 3

Average (± 1 StdDev) errors of SHV (haptic vertical) for both the tactile rod (panel A) and the tactile plate (panel B) are illustrated. Adjustments for right- and left-handed subjects are illustrated separately and errors of CW rotations are compared with those of CCW rotations. The starting position of the device and the required direction of rotation (CW vs. CCW) are also illustrated in inlets on the top. Black circles refer to adjustments with the right hand, gray squares to adjustments with the left hand. Trial types with significant errors relative to zero (t-tests) are marked (**).

Figure 4

Here average (± 1 StdDev) trial-to-trial variability of SHV is presented; results when using the tactile rod (panel A) and the tactile plate (panel B) are shown separately. For description of the symbols and trial conditions see legend of Figure 3.

Figure 5

Average (± 1 StdDev) errors of PSA (perceived straight-ahead) for both the tactile plate and the tactile rod in 10 right-handed subjects are shown. For description of the symbols and trial conditions see legend of Figure 3.

Figure 6

Average (± 1 StdDev) trial-to-trial variability of PSA in right-handed subjects is presented; results when using the tactile rod and the tactile plate, respectively, are shown separately. For description of the symbols and trial conditions see legend of Figure 3.