The space around us can be divided based on what lies within our reach or outside of it. This distinction is of primary importance for defining efficient motor plans that allow us to interact with the environment. The area surrounding our body, which is known as the Peripersonal space (PPS), was first described in monkeys, with the discovery of multi-modal neurons that only fire if a stimulus is placed near the body of the animal [1]. These neurons are located in the ventral premotor cortex F4 [2] and have receptive fields that surround the monkey’s hand and move with it [1]. Interestingly, the electrical stimulation of these neurons elicits defensive-like movements in monkeys, as if the animals are trying to protect the part of the body where the receptive fields of the stimulated neurons are located [3]. The PPS can thus be conceptualized as a system specifically dedicated to the perception of stimuli that are in the immediate surroundings of the body and that may indicate potential risk or interest. In line with that, it has been proposed that the activity of brain areas in which PPS neurons are located aims at maintaining a margin of safety around the body [4].
Neuropsychological studies with neglect patients provide evidence for the presence of a specific area in the human brain dedicated to the perception of stimuli in the PPS. Hemineglect, or simply neglect, is a condition resulting from brain damage that leads to the inability to attend to stimuli presented in the contralesional hemifield [5]. In some cases, neglect patients fail to report the presence of a stimulus presented in the controlesional side when a competing stimulus is simultaneously shown in the ipsilesional side, a phenomenon known as “extinction” [6]. Ladavas et al. [7] demonstrated that the extinction effect also occurs cross-modally: the reduction in sensitivity to a tactile stimulus triggered on the controlesional hand induced by an ipsilesional touch was rather identical to that yielded by a visual stimulus displayed around the ipsilesional hand, suggesting the existence of a cross-modal visuo-tactile extinction. However, this effect was significantly reduced when the visual stimulus was presented outside of the patient’s PPS, indicating that the deficit in the combination of visual and tactile information in neglect was more pronounced within the PPS. A dissociation between PPS and the Extrapersonal space (EPS) is also found in pseudo-neglect, an attentional deficit presented by neurologically healthy individuals. When asked to bisect a horizontal line, most people tend to provide leftward-biased responses [8]. Interestingly, this attentional bias attenuates progressively with distance. When asked to perform the bisection task in EPS, individual responses shift rightward instead of leftward [9], [10].
While the general definition of the size of the PPS is straightforward, its precise measurement poses some challenges. In animal models, the border between PPS and EPS can be measured with accuracy using in-vivo single-cell recording [1]. However, in humans a non-invasive behavioral approach has to be adopted. One of the most widely used methods to capture the boundary between PPS and EPS is the audio-tactile detection task pioneered by Canzonieri et al. [11]. In this task, participants are presented with a looming auditory stimulus that creates the illusion of an approaching sound source. Following a predetermined delay, a tactile stimulus is delivered to the hand as a vibration and participants are instructed to react to this as fast as possible, while ignoring the auditory stimulus. If the vibration is delivered when the sound is perceived as being in the PPS, participants are faster to react than when the sound is perceived as being in the EPS. Thus, the operational definition of PPS using this auditory-tactile detection task can be summarized as the maximum distance from the participant’s body at which the auditory stimulus can still facilitate the detection of the tactile stimulus. Notably, one of the characteristics of PPS neurons is that they are multimodal [1] and thus capable of firing for both visual and auditory impulses. This has opened up to the possibility of measuring detection facilitation within the PPS with visual instead of auditory stimuli.
However, PPS should not be considered as a static bubble that surrounds one's body. Not only the PPS has been reported to show different characteristics between healthy and clinical populations [12], [13], but it has also been shown to be malleable to environmental conditions. For instance, as the size of the PPS is robustly based on the extension of our reach, if our reach changes, the PPS gets modulated accordingly. Peripersonal neurons tuned to a specific area within the surrounding space have been shown to quickly reshape their receptive fields to accommodate changes in the extent of the reaching area. Indeed, when monkeys are trained to use a stick to retrieve food from a distance, the receptive field of PPS neurons becomes longer [14]. In humans, there have been reports of amputees [15] and wheelchair users [16] having a smaller PPS compared to controls, whereas blind cane users [17] and long-term computer mouse users [18] have a larger PPS when holding their cane or mouse respectively. Also elite athletes show a remapped PPS. For instance, experienced tennis players exhibit larger PPS when they hold their racket [19], with a similar effect also being found in fencers holding their sword [20].
While modulation resulting from long-term conditioning induces a stable and durable reshaping of the PPS, there have also been reports of PPS reshaping after a short training session. For instance, after performing a task in which participants use the tip of a cane to find objects scattered on the floor, an extension of the PPS was reported. This effect was similar to blind cane-users, however the remodulation in the healthy participants was transient [17]. Indeed, when participants were retested the day after the training, the PPS extension had disappeared. This result supports the idea that the duration of the training is proportional to the stability of the remapping. Indeed, this brief extension of PPS as a result of a short tool-use training has been observed under several different conditions [15], [18], [21], [22]. An enlargement of the PPS has also been observed while walking on a treadmill [23], suggesting that the illusion of moving forward, even when there is no overall displacement, increases the perception of what can be considered within reach. However, as the expansion of PPS has been achieved via quite different motor routines, it is not clear which aspect of the training triggers the reshaping. Is it the forward motion of a body part, the proprioceptive feedback of reaching with a tool, or the motion of the body induced by walking?
Virtual Reality (VR) set-ups offer an ideal environment to study which aspect of the training plays a key role in inducing the reshaping of the PPS as it allows to design ecologically-plausible looming visual stimuli. Indeed, previous reports suggest that PPS can successfully be investigated with VR [24]–[26]. As it has been demonstrated for the audio-tactile task, if the vibration is delivered when the virtual visual stimulus is perceived as being close to the participants body, reaction times (RTs) to the vibration are significantly reduced, and in some cases almost halved [24].
However, some concerns have been presented on whether measuring the PPS in VR provides a true representation of the PPS in the real world. Specifically, Ferroni et al. [27] asked participants to perform a short training in which they had to move small objects from point A to point B with a horizontal dragging motion, with both A and B being landing points placed in EPS. This task triggered an enlargement of the PPS when the task was performed in the real world. However, this was not the case when the same routine was executed in a VR environment. Even though this result seems to question the usefulness of VR environments to study modulations of the PPS, it is important to note that a large variability of the effectiveness of the motor training has also been reported for experiments carried out in real world, so a more systematic investigation is needed to reach a definitive conclusion.
In our study we attempted to induce a reshaping of PPS in a VR environment by carrying out four different types of training. The first group of participants (Experiment 1) was trained using the most widely used task to induce a reshaping of the PPS: pulling an object closer to one’s body with the use of a tool. In a different condition, the same participants were instead instructed to perform a similar motor routine but with the opposite goal: to push an object away from their bodies into the EPS. In Experiment 2 we aimed to determine whether crossing the PPS/EPS border, as in the push-pull conditions in Experiment 1, is essential for reshaping the PPS or whether any repetitive hand movement with a tool can induce such a reshaping. To this end, participants had to repeatedly hit a target in the EPS using a hammer. To further explore this issue, in a final condition, participants were required to interact with the target in the EPS without any direct “physical contact” as they were required to shoot the target with a toy-gun. The first assessment in validating the VR as a tool to investigate PPS would be to replicate the PPS expansion as a consequence of training that involved pulling movements, as this training is the one of the most widely used in previous literature. Then, in case the PPS malleability is bidirectional, we would expect to reduce the PPS space as a consequence of pushing. Finally, with the hammering and shooting conditions, we aimed to investigate whether crossing the PPS/EPS border (hammering) or direct physical contact with a target (shooting) during training is essential for modulating the size of the PPS.