Multimodal Human-Robot Interaction

—Touch provides an important cue to perceive the physical properties of the external objects. Recent studies showed that tactile sensation also contributes to our sense of hand position and displacement in perceptual tasks. In this study, we tested the hypothesis that, sliding our hand over a stationary surface, tactile motion may provide a feedback for guiding hand trajectory. We asked participants to touch a plate having parallel ridges at different orientations and to perform a self-paced, straight movement of the hand. In our daily-life experience, tactile slip motion is equal and opposite to hand motion. Here, we used a well-established perceptual illusion to dissociate, in a controlled manner, the two motion estimates. According to previous studies, this stimulus produces a bias in the perceived direction of tactile motion, predicted by tactile flow model. We showed a systematic deviation in the movement of the hand towards a direction opposite to the one predicted by tactile flow, supporting the hypothesis that touch contributes to motor control of the hand. We suggested a model where the perceived hand motion is equal to a weighted sum of the estimate from classical proprioceptive cues (e.g., from musculoskeletal system) and the estimate from tactile slip I. INTRODUCTION Cutaneous touch plays an important role in the perception of the physical properties (e.g., shape, texture, weight) and the motion status of external objects [1], [2]. Material properties of the object, such as roughness [1], [3] and compliance [4], [5], are also encoded by the tactile system. In addition to this role in object and material perception, the deformation of the fingertip also contributes to our sense of hand position and motion [6], [7]. Classical studies in physiology showed that receptors in the musculoskeletal system (such as muscle spindle, Golgi tendon organ and joint receptors) and strain patterns on the skin convey information on the static position and the displacement of our limbs [8], [9], [10], [11], [12], [13]. Specific cutaneous stimuli can also produce the illusory sensation of hand motion in perceptual tasks. We recently showed that the deformation of the fingertip's skin occurring when we push the finger against a soft surface generates the illusory sensation of finger displacement [6]. The tactile motion generated by a rotating disk produces the sensation of the hand rotating in the opposite direction of the disk [14]. In this study, we tested the hypothesis that touch provides auxiliary cues to guide hand displacement [15], [16]. In our daily-life experience, whenever we slide the hand over a stationary surface, like a working desk or a table, the velocity of tactile motion is equal and opposite to the hand velocity (Fig. 1). If the sensorimotor system uses touch as a motion cue, a stimulus that decouples the tactile and the kinesthetic velocity estimates would provide a biased motion signal and produce a systematic error in hand movement. Here, we used a well-established tactile phenomenon (previously investigated by some of the coauthors of the present study), to decouple the two velocity estimates. In [17], participants kept the hand world-stationary and a textured plate with parallel raised ridges slid under their fingertip. The perceived movement of the plate was strongly biased towards a direction perpendicular to the orientation of the ridges, in accordance with the tactile flow model [17]. Fig. 1: Sliding the hand over a stationary surface, for example a working desk, the relative movement sensed from cutaneous touch (black) is equal and opposite to hand motion (grey). Therefore, touch can be a strong cue to hand motion. Here, we used this perceptual phenomenon to decouple the tactile and proprioceptive feedback to active hand motion. We asked blindfolded participants to slide the finger on a surface with parallel ridges, trying to move the hand along a straight direction. According to our hypothesis, in the absence of other sensory feedback, tactile feedback should lead to a systematic error in hand displacement, towards a direction opposite to the one predicted by the tactile flow. Finally, we suggested a model (to be further evaluated in future studies) where the perceived hand motion is equal to a weighted sum of the estimate from classical proprioceptive cues (e.g., from receptors in the musculoskeletal system) and the estimate from tactile slip. Previous studies in psychophysics supported the hypothesis of an integration of multiple cues for the perception of hand displacement (see for example [14], [7], [6]), however, this was never evaluated for the on-line control of the hand and the limb movement. II. METHODS A. Participants Six naive healthy participants took part in the experiment (4 males and 2 female, age: 25.1 ± 1.2867, mean ± standard deviation). Participants were all right-handed. Participants reported no medical condition that could have affected the experimental outcomes.The testing procedures were approved by the Ethical Committee of the University of Pisa, in accordance with the guidelines of the Declaration of Helsinki for research involving human subjects. Informed consent was obtained from all participants involved in the study. B. Stimulus and Procedure The experimental setup (Fig. 2) included a 3D-printed circular plate (diameter: 15 cm) placed over a load cell (Micro Load Cell, 0 to 780 g, CZL616C from Phidgets, Calgary, AB-Canada). The plate had a textured surface with regularly spaced ridges (ridge height and width: 1 mm; space between ridges: 10 mm), consistently with [17]. A Leap Motion device (Leap Motion Inc., San Francisco, U.S.) was placed above the plate for hand tracking. We centered the reference system of the Leap Motion device in the center of the circular plate. The sampling frequency of the Leap Motion device was equal to 40 Hz, which allows to correctly track hand motion at typical scanning speeds. Fig. 2: The experimental setup including the textured circular plate, the Phidgets Micro Load Cell and the Leap Motion device. Blindfolded participants sat on an office chair in front of the setup, with the center of the plate roughly aligned with their body mid-line. Headphones playing pink noise masked occasional ambient sound. In each trial, participants were required to contact the plate with their right index finger and to move the hand away from them along a straight path for approx. 10 cm (solid arrow in Fig. 3). Participants were instructed to contact the plate with a light touch. Prior to each trial, the plate was rotated by the experimenter to one of the following angular position: -60, -30, 0, 30, 60 deg. A zero angle means that the ridges of the plate were parallel to the frontal plane of the participant, Fig. 3: Participants moved the hand on a plate with oblique ridges, along the direction indicated by the solid arrow. According to the model of tactile flow, the cutaneous feedback produced an illusory sensation of bending towards a direction perpendicular to the ridges (dashed arrow). This eventually led to an adjustment of the motion trajectory towards a direction orthogonal to tactile flow, i.e., parallel to the ridges (dotted arrow). The actual hand trajectory also depended on extra-cutaneous cues, e.g. from musculoskeletal system (not shown in the picture). whereas negative (positive) angles means that the ridges were rotated clockwise (counterclockwise). Each stimulus orientation was presented fifteen times, in pseudo-random order. Additionally, participants replicated the task with a smooth plate without ridges. This aimed at correcting our results for possible biases in perceived direction introduced by extra-cutaneous cues, for e.g. proprioception [18]. Participants received no feedback about their performance during the experiment. At the end of each trial, the experimenter lifted the hand of the participant to place it back to the starting position. Before the experimental session, participants underwent a training phase, where the experimenter instructed them to produce the right amount of force and hand displacement. During training, participants received a feedback whenever the actual force exceeded the threshold value of 2 N. C. Data Analysis The hand trajectory was recorded with the tracking system of the apparatus and saved for the analysis. We linearly interpolated the hand trajectory separately for each trial and participant and estimated the angular deviation from a straight-ahead motion direction (i.e., the deviation from the solid arrow in Fig. 3). Negative (positive) angles means that the motion path rotated clockwise (counterclockwise) with respect to the solid arrow in the figure. Using Linear Mixed Model (LMM), we evaluated whether the orientation of the ridges, X, predicted this angular error, A: A=β0+u0+(β1+u1)X+ε, (1) where β0 and β1 are the fixed-effect intercept and slope, respectively, u0 and u1 are the random-effect intercept and slope of the model (between-participant variability), and ε is the residual error term. In order to account for possible biases produced by extra-cutaneous cues (e.g., proprioceptive cues from the musculoskeletal system), we analyzed the trials with a zero-degree orientation of the ridged plate and with the smooth plate. First, we verified, using the Likelihood Ratio Test, that the angular error was not significantly different between these two experimental conditions. Then, we fitted the following model to estimate the angular deviation from straight direction in the absence of biasing tactile stimuli. A, (2) where A0 is the predicted angle with zero-oriented or no ridges, and β0∗ is the estimate of the possible bias due to extra-cutaneous cues. We used β0∗ to correct the estimate of the tactile bias estimated in model (1). Next, we analyzed by means of LMM whether the orientation of the ridges predicted the final position error along the frontal plane, P. P=η0+u0+(η1+u1)X+ε, (3) In Equa

Récupéré en direct depuis OpenAlex et désinversé. Les résumés ne sont pas conservés dans cette base de données : les index inversés représentent 8,6 Go des 9,3 Go de texte de la base, et le serveur dispose de 13 Go libres.