Interhemispheric communication during haptic self-perception

Abstract

During the haptic exploration of a planar surface, slight resistances against the hand's movement are illusorily perceived as asperities (bumps) in the surface. If the surface being touched is one's own skin, an actual bump would also produce increased tactile pressure from the moving finger onto the skin. We investigated how kinaesthetic and tactile signals combine to produce haptic perceptions during self-touch. Participants performed two successive movements with the right hand. A haptic force-control robot applied resistances to both movements, and participants judged which movement was felt to contain the larger bump. An additional robot delivered simultaneous but task-irrelevant tactile stroking to the left forearm. These strokes contained either increased or decreased tactile pressure synchronized with the resistance-induced illusory bump encountered by the right hand. We found that the size of bumps perceived by the right hand was enhanced by an increase in left tactile pressure, but also by a decrease. Tactile event detection was thus transferred interhemispherically, but the sign of the tactile information was not respected. Randomizing (rather than blocking) the presentation order of left tactile stimuli abolished these interhemispheric enhancement effects. Thus, interhemispheric transfer during bimanual self-touch requires a stable model of temporally synchronized events, but does not require geometric consistency between hemispheric information, nor between tactile and kinaesthetic representations of a single common object.

Keywords: force–geometry illusion; haptic perception; interhemispheric communication; self-touch; sensorimotor integration.

Figure 1.

Experimental set-up and stimuli. (a) Front view of the leader–follower robotic set-up. (b) Top view of the experimental set-up. Participants made two proximo-distal movements with their right hand while holding the leader robot, and simultaneously felt a corresponding stroke on the left forearm from a brush attached to the follower robot. (c) Lateral view of the experimental set-up. Resistance was added to both right-hand movements in order to generate two illusory haptic bumps of different sizes. Participants judged which of the two bumps felt larger. The trajectory of the leader robot was replicated by the follower robot, which stroked the participant's left forearm with a paintbrush. The tactile stroking included one of three task-irrelevant modulations of tactile contact force. These achieved the perception of a ‘tactile bump’, a ‘tactile hole’ and a ‘tactile flat’ contact, corresponding to temporarily increased, decreased or unchanged pressure on the left forearm, respectively (electronic supplementary material, videos S1 and S2). (Online version in colour.)

Figure 2.

Average psychometric curves for (a) Experiment 1, blocked design and (b) Experiment 2, randomized design. The ‘proportion of the test stimulus was larger’ was fitted as a function of the difference between test and reference (test–reference). Grey curve is the control condition that served as a baseline (no tactile pressure change), while coloured curves indicate experimental conditions, with different tactile events contributing to the haptic percept. Shaded area indicates the standard error of the mean. Vertical dashed line indicates the bump size in the reference movement. The reference bump size was always 3 N indicated by the peak of the added resistance, and the test stimulus was randomly selected from five possible sizes (±1 N, ±0.5 N and ±0 N relative to the reference bump). See Methods for details. (Online version in colour.)

Figure 3.

PSEs of Experiment 1 (blocked design, (a)) and Experiment 2 (randomized design, (b)). Bars of different colours denote the four experimental conditions: tactile bump + test, tactile hole + test, tactile bump + ref and tactile hole + ref, respectively. The error bar indicates the standard error of the mean across participants. The size of the reference bump was always 3 N, and the test stimulus was randomly selected from five possible sizes (±1N, ±0.5 N and ±0 N relative to the reference bump). We calculated the difference in PSE for trials where tactile information was added to the test bump versus the reference bump, and then corrected this value by subtracting each participant's PSE in the baseline: no tactile pressure change condition. (Online version in colour.)