Manipulating The Body Representation: Assessment Of A Novel Platform

Abstract

The Rubber Hand Illusion can be used to induce the illusion that a fake hand is part of one’s own body. Thus, it can be used to alter the body representation. It was also reported that the Rubber Hand Illusion induces a proprioceptive drift of one’s real hand toward the fake hand. The Rubber Hand Illusion can be induced when the fake hand is placed farther in the sagittal plane (distally) compared to the real hand. In this case, the induced update of the body representation is an elongation of the arm. Virtual Reality and haptic technologies can be used to manipulate the perceived scenario in a virtual version of the Rubber Hand Illusion, the Virtual Hand Illusion. We developed a novel platform consisting in a virtual reality application integrating an optical motion capture device and haptic stimulators to study the manipulation of the body representation. We developed two experimental protocols to induce embodiment of an elongated arm: one validated in previous studies, that employs congruent visuo-motor-tactile stimulation, and one reproducing the typical Virtual Hand Illusion where only congruent visuo-tactile stimulation was employed. We tested both protocols with healthy participants.

I. Introduction

The sense of body ownership refers to the particular perceptual status when part of one’s own body is identified as self. The Rubber Hand Illusion (RHI) is a method employed to study the sense of body ownership. It is performed by congruently brush-stroking a visible fake hand and the hidden real hand of a participant, while the participant is observing the fake hand. It is used to induce the illusion that a fake hand is part of a participant’s own body. Thus, the illusion that arises can be used to alter the body representation. It has been reported that the RHI also induces in participants a shift of perceived location of the real hand towards the fake hand, called the proprioceptive drift (PD) [1].

Virtual Reality technologies have proved to be able to easily modify the visual stimuli in experimental environments [2][3]. Furthermore, the deception of a subject is even stronger when the visual VR technology is mixed with additional technologies such as motion capture and haptic stimulation devices. Thus, VR, motion capture and haptic stimulation can be employed to manipulate the perceived scenario in a virtual version of the RHI, the Virtual Hand Illusion (VHI), aimed at studying the alteration of the body and the plasticity of its representation.

Usually, the fake hand of the RHI is placed closer along the frontal plane with respect to the real hand (medially), and the illusion induces a latero-medial drift of the perceived real hand’s position. However, it has been shown that the RHI can also be induced when the fake hand is placed farther along the sagittal plane with respect to the real hand (distally) [4]. A similar experiment performed in VR [5] in which participants perceived congruent visuo-motor and visuo-tactile stimulation with an elongated virtual arm showed that it induces a proximo-distal PD towards the distal virtual hand (up to three times the length of the participants forearm).

We developed a novel platform consisting in a highly immersive VR application integrating an optical motion capture device and haptic stimulators in order to study the manipulation of the body representation. We tested two experimental protocols aiming at inducing embodiment of an elongated arm. one is the VHI paradigm that reproduces the typical RHI where only congruent visuo-tactile stimulation was employed to induce the embodiment (VT-VHI). The second one is reproduced from the experiment performed in [5] employing congruent visuo-motor-tactile stimulation (VMT-VHI).

II. Material and Methods

A. Real environment

In the real environment, the participant is seated comfortably in a chair in front of a table and is equipped with the VR head mounted device (HMD) covering the eyes. They are positioned at the center of the motion capture workspace while VR infrared cameras are placed outside of the workspace, oriented towards the HMD. On the table is placed a large paper goniometer to perform the angle estimation used to evaluate the proprioceptive drift.

B. VR application

The VR application is developed with Unity 2018.3.0f2 (Unity Technologies). The VR hardware equipment used is the HTC Vive kit consisting in the HMD and two infrared cameras. All the virtual avatars used in this application were created with the open source software MakeHuman™. In the virtual environment, participants see their virtual body coincident with their real body in a first-person perspective. The physical appearance of the avatar can be switched between a male and female appearance, and the length of the left arm and left forearm can be modified to match the morphology of the subject. The participants’ avatar is seated in front of a table.

Two different virtual environments were developed to test the different experimental protocols. For the VT-VHI experiment, a virtual experimenter is seated in front of the subject on the other side of the table, in a room similar to the real one. The experimenter is holding a virtual paint brush in the right hand to perform the stroking on the participant’s left hand. When animated, the experimenter avatar performs the brush-stroking of the subject’s avatar’s left hand’s index. The brushing movement is reproduced from a previous motion capture recording of a real experimenter performing the brush-stroke. The forearm of the subject’s avatar can be elongated up to twice or three times the size of the participants real forearm.

For the VMT-VHI, the subject in seated in the same position as previously mentioned, but without any virtual experimenter. A virtual piece of cardboard is also present on the table. The possible elongations are the same as the ones in the VHI experiment.

C. Motion Capture

To track the movement of the subject, an Optitrack motion capture system (Natural Point, Inc.) has been used. The motion capture hardware consists in 4 Prime 13W optical cameras placed on the top corners of a cubic metallic structure. Optical markers are attached to the participants’ left arm, left forearm and chest, to track their movements and reproduce them accordingly on the virtual avatar. This is done to provide the subjects with the sense of agency over the virtual avatar. The markers positions are processed through Optitrack’s Motive software (version 2.1.1 Final), and live streamed to the VR application in Unity.

The left hand and its fingers movements are also tracked, but using a Leap Motion device (Leap Motion, Inc.) mounted on the HTC Vive HMD. Their movements are also live streamed to the VR application and reproduced on the subject avatar’s left hand and its fingers.

D. Haptic stimulation

The haptic stimulation is performed thanks to a small 3D printed robotic arm holding a paintbrush. The robotic arm has one degree of freedom and is actuated by a brushed DC motor (Pololu Corporation, USA). It is designed in order to reproduce the typical brushing pattern of a human experimenter. The motor is controlled by prototyping board (STMicroelectronics, Switzerland) and an Olimex controller (Olimex, Bulgaria). The brush of the robot is equipped with a piezo-electric sensor that detects the contact between the brush and the index finger of the subject in order to synchronize the movement of the robot with the movement of the virtual brushing. The haptic stimulation can be performed congruently or incongruently with respect to the visual virtual brushing.

In the case of the VMT-VHI, a piece of cardboard is placed on the table, under the hand of the subject, and the haptic stimulation is performed through the tactile feedback felt by the participant when exploring the piece of cardboard with the left hand.

III. Experimental Procedure

A. Participants

Eight volunteers (3 females, age = 28.5± 1.69 [mean ± standard error]) naïve to the RHI or VHI paradigm, participated in the study (one for each condition). Seven were right handed. All participants were healthy and claimed to have normal hand sensation and normal or corrected to normal vision. Informed consent according to the Declaration of Helsinki (BMJ 1991; 302:1194) and to the Ethical Committee of the Campus Bio-Medico University, was obtained before conducting the experiments.

B. Visuo-motor-tactile VHI

In order to reproduce the experiment performed in [5], we decided to test four different condition: one incongruent condition with virtual forearm length equal to the real one (I), and three congruent conditions with virtual forearm lengths equal to (C1), twice (C2), or three times the real one (C3).

In all conditions, the right arm of the subject was not tracked, and the subject was asked to keep it still alongside the body. The right arm of the avatar was positioned accordingly. The left arm of the subject was initially placed as rest on the table, with the left hand placed on the piece of cardboard.

In the familiarization phase, the participants were first asked to look around the room and especially at the virtual body. The virtual forearm was not yet elongated, and its length was equal to the real one in all conditions. Subjects were then asked to explore the material they felt under the left hand by touching it with the left hand. The movements performed by the subject were accordingly reproduced by the avatar. In congruent conditions, the subjects were provided with visuo-motor and visuo-tactile correlation, as they saw the virtual hand touching the virtual piece of cardboard while feeling its texture. In the incongruent condition however, subjects were only provided with visuo-motor correlation. The virtual piece of cardboard was placed out of reach, farther in front of the avatar’s left hand, hence providing the subject with a visuo-tactile mismatch.

In the experimental phase, the virtual left forearm of the avatar was then elongated accordingly to the current condition (I – C1: no elongation, C2: twice the real length, C3: three times the real length). They were then asked to stroke again the surface of the cardboard they felt for ninety seconds.

Finally, the subjects were told the experiment was over and showed the virtual forearm at the real length for a few seconds before removing the HMD and filling a questionnaire.

C. The visuo-tactile VHI

The first part of the experimental protocol consists in a familiarization phase with the virtual avatar to provide the subjects with the sense of agency over the virtual avatar. In the virtual environment the subject’s avatar is seated on a chair in front of a table and the experimenter is seated on the other side, arms crossed. The right arm of the subject is not tracked, and the subject is asked to keep it still and alongside the body. It would remain in this position during the whole experiment, except for the angle estimation tasks. The right arm of the avatar was positioned accordingly. The motion tracking of the chest and left arm, hand and fingers of the subject was activated, and the subjects were asked to explore the surrounding area by moving their chest, left arm, left hand and its fingers freely while looking at the virtual body moving accordingly. In particular, subjects were asked to pay close attention to the virtual left hand and fingers.

Then, participants were asked to put their arm and hand down on the table, and to remain still with the hand palm down, in a relaxed position. When this was done, the motion tracking was disabled. Subjects were then asked to pay close attention to the left hand they were seeing, and the brushing session was started. The virtual forearm was not yet elongated, and its length was equal to the real one. In all conditions, the visual brushing was performed by the virtual experimenter on the index finger of the virtual left hand, and the tactile brushing was performed by the brushing robot on the index finger of the real left hand of the subject. In congruent conditions, the subjects were provided with visuo-tactile correlation, as they saw the virtual left hand index finger being brushed while feeling a brushing on their own left hand index finger. In the incongruent condition however, a temporal delay between the virtual and the real brushing was set, resulting in a visuo-tactile mismatch.

Four conditions similar to the one’s of the VMT-VHI were tested. One congruent (C1) and one incongruent (I) brushing condition were tested for a virtual forearm length equal to the real one. Two other congruent brushing conditions were tested, one for a virtual arm length of twice the real one (C2) and one for a virtual arm length three times the real one (C3).

In the experimental phase, the virtual left forearm of the avatar was then elongated accordingly to the current condition (I & C1: no elongation, C2: twice the real length, C3: three times the real length). Participants were then asked to remain still while looking at the left hand like for the first brushing phase, and the second brushing phase was performed for ninety seconds.

D. Embodiment outcomes

In order to measure the strength of the illusion on different conditions, each participant filled-in a questionnaire (one for each stimulation condition, at the end of the trial).

The questionnaire included the six statements list designed in [5]. Three of the statements (i.e. illusion statements) referred to the extent of sensory transfer into the virtual hand and its self-attribution during the trial; whereas the other three statements (i.e. control statements) served as controls for compliance, suggestibility, and “placebo effect”. Participants were asked to rate the extent to which these statements did or did not apply to their experience, by using a seven-point analogue scale. On this scale, -3 meant “absolutely certain that it did not apply,” 0 meant “uncertain whether it applied or not, ” and +3 meant “absolutely certain that it applied”.

The RHI index, expressed as the difference between the mean score of the illusion statements and the mean score of the control statements [6], was calculated for each condition and employed as illusion outcome.

In addition to the nine statements, participants were asked to rate vividness and prevalence of self-attribution of the rubber hand [7], [8]. The vividness was defined as how life-like and realistic the illusion was when it was experienced; it was rated from 1 to 9. The prevalence rating (from 0% to 100%) reflected the percentage of time that the illusion was experienced (equivalent to the continuance of the illusion).

In all conditions, after every experimental phase, participants were then asked to close their eyes, and the display was turned black. They were helped to place their right arm on the table so that their elbow and forearm were placed on the correct position of the goniometer. They were then asked to point with their right hand’s index towards the position where they felt their left hand’s index, moving only the forearm and keeping the elbow at the same position (i.e. angle estimation task). After that, they were asked to put their right arm back to its position alongside the body. Post-experimental minus pre-experimental phase positive differences of the estimated position of the hand indicate a drift of the perceived location of the real hand toward the virtual hand.

The results of the RHI index, the vividness score and the prevalence score show embodiment results in line with previous studies [5][6] for both experimental protocols, directing towards the possibility to evoke embodiment illusion of an elongated arm with the developed platform.

However, it has to be emphasized that the very limited number of participants tested in this study does not enable to give any statistical relevance to those results.

V. Conclusion and future work

We developed a novel platform aimed towards the manipulation of the body representation. To investigate the feasibility of such manipulation, we tested two different experimental protocols to induce embodiment of an elongated arm: a reproduction of the visuo-tactile-motor experiment performed in [5] and a VHI paradigm that reproduces the typical RHI where only congruent visuo-tactile stimulation is employed.

we were able to collect the first data, and it is now necessary to increase the number of subjects to study the possibility to evoke embodiment illusion of an elongated arm with the developed platform.

This platform could be used for several possible future applications. one could be to study techniques of non-invasive stimulation to enhance embodiment of a limb [9]. Another one could be to help to understand the process behind the formation of the internal image of the body. Finally, one could be a possible therapeutic application for people with an altered image representation of their body caused by amputation or another upper limb trauma.

Fig. 1. Visualization of the VR application environment.

(A) Environment of the VMT-VHI, (B) environment of the VT-VHI.

Fig. 2. Hardware setup of the platform in VT-VHI.

Fig. 3. Proprioceptive Drift, RHI index, Vividness score and Prevalence score.