4 Vestibular contributions to various aspects of the bodily self

4.5 The spatial self: Self-location

4.5.1 Behavioural studies in humans

Self-location is the experience of where “I” am located in space and is one of the (if not the) crucial aspects of the bodily self (Blanke 2012). Recently, self-location has been systematically investigated in human behavioural and neuroimaging studies using multisensory conflicts (Ionta et al. 2011; Lenggenhager et al. 2007; Lenggenhager et al. 2009; Pfeiffer et al. 2013). While we usually experience ourselves as located within our own bodily borders at one single location in space, the sense of self-location can be profoundly disturbed in psychiatric and neurological conditions, most prominently during out-of-body experiences (Bunning & Blanke 2005). Based on findings in neurological patients that revealed a frequent association between vestibular illusions (floating in the room, sensation of lightness or levitation) and out-of-body experiences, Blanke and colleagues proposed that the illusory disembodied self-location was due to a dis-integration of vestibular signals with signals from the personal (tactile and proprioceptive signals) and extrapersonal (visual) space (Blanke et al. 2004; Blanke & Mohr 2005; Blanke 2012; Lopez et al. 2008). The authors proposed that this multisensory disintegration is mostly a result of abnormal neural activity in the temporo-parietal junction (Blanke et al. 2005; Blanke et al. 2002; Heydrich & Blanke 2013; Ionta et al. 2011). In this section, we review experimental data in healthy participants that may account for the tight link between vestibular disorders and illusory or simulated changes in self-location. While the most direct evidence of such a link comes from the finding that artificial stimulation of the vestibular organs induces an illusory change in self-location[13] (Fitzpatrick & Day 2004; Fitzpatrick et al. 2002; Lenggenhager et al. 2008), we focus on three experimental set-ups that have been used to alter the experience of self-location in healthy participants.

4.5.1.1 Illusory change in self-location during full-body illusions

Full-body illusions have increasingly been used to study the mechanisms underlying self-location (see Blanke 2012, for a review). No study has so far investigated the influence of artificial vestibular stimulation on such illusions. Nevertheless, there is some experimental evidence suggesting a vestibular involvement in illusory changes in self-location. While the initial full-body illusion was described in a standing position (Lenggenhager et al. 2007), the paradigm has later been adapted to a lying position (Ionta et al. 2011; Lenggenhager et al. 2009; Pfeiffer et al. 2013), mainly because the frequency of spontaneous out-of-body experiences is higher in lying position than in standing or sitting positions (Green 1968). It has been speculated that this influence of the body position on the sense of embodiment is related to the decreased sensitivity of otolithic vestibular receptors and decreased motor and somatosensory signals in the lying position (Pfeiffer et al. 2013). We hypothesized that under such conditions of reduced vestibular (and proprioceptive) information, visual capture is enhanced in situations of multisensory conflict, thus resulting in a stronger change in self-location during the full-body illusion. So far, the full-body illusion has not been directly compared in standing versus lying positions. However, the application of visuo-tactile conflicts in a lying position not only alters self-location but also evokes sensations of floating (Ionta et al. 2011; Lenggenhager et al. 2007). This finding hints toward a reweighting of visual, tactile, proprioceptive, and vestibular information during the illusion, plausibly in the temporo-parietal junction and human PIVC. In line with this finding, the changes in self-location and perspective have been associated with individual perceptual styles of visual-field dependence (Pfeiffer et al. 2013), i.e., weighting of visual as compared to vestibular information in a subjective visual vertical task, suggesting an individually different contribution and weighting of the various senses for the construction of the bodily self (for a similar finding regarding the rubber hand illusion, see David et al. 2014).

4.5.1.2 Mental own-body transformation and perspective taking

Another way to investigate bodily self-consciousness has been to use experimental paradigms requiring participants to put themselves “into the shoes” of another individual, that is to mentally simulate an external self-location (own-body, egocentric, mental transformation tasks) and a third-person visuo-spatial perspective. Typically, participants are instructed to make left-right judgments about a body, for example, to judge whether this other shown person is wearing a glove on his right or left hand (Blanke et al. 2005; Lenggenhager et al. 2008; Parsons 1987; Schwabe et al. 2009). Other tasks require that participants adopt the visual perspective of another person to decide whether a visual object is to the right or left of the other person (David et al. 2006; Lambrey et al. 2012; Vogeley & Fink 2003). Early studies have shown that the time needed for own-body mental transformations correlates with the distance or angle between the participant’s position in the physical space and the position to be simulated (Parsons 1987). It is largely admitted that own-body mental transformation is an “embodied” mental simulation that can be influenced by various sensorimotor signals from the body (e.g., Kessler & Thomson 2010). In line with this view, various experiments demonstrated that the actual body position influences mental own-body transformation of body parts (e.g., Ionta et al. 2012). Importantly, next to proprioceptive and motor mechanisms, visuo-spatial perspective taking and own-body mental transformation also require the integration of vestibular information (active or passive body motion). Thus, while most of this research looked at how body parts’ posture (e.g., of the hand) influences mental own-body (part) transformation, some recent research investigated how mental own-body transformation is influenced by vestibular cues (Candidi et al. 2013; Dilda et al. 2012; Falconer & Mast 2012; Lenggenhager et al. 2008; van Elk & Blanke 2014). All these studies revealed that vestibular signals influence mental (full) own-body transformation, confirming again the influence of the vestibular system in the sense of self-location and perspective taking.[14]

4.5.1.3 Change in self-location and the feeling of presence

The development of immersive virtual environments has launched a powerful research area where the mechanisms of self-location can be investigated and manipulated by the feeling of presence. The term “presence” stems from virtual reality technologies and commonly refers to the feeling of being immersed (“being there”) in the virtual environment. Yet, it has been argued that “presence” also reflects a more general and basic state of consciousness (Riva et al. 2011). The study of presence has thus been suggested to provide useful tools to study (self-)consciousness, with the advantage of precise experimental control (Sanchez-Vives & Slater 2005).

Similar to previously mentioned full-body illusions, a participant who is immersed in a virtual environment receives contradicting multisensory information about his or her self-location: while visual information suggests that s/he is located in a virtual world, proprioceptive information suggests that s/he is located in the real world, for example, by indicating a different body position between the physical body and the avatar. Furthermore, and contrary to the full-body illusion, the visual information often indicates that the participant is moving, whereas the proprioceptive and vestibular information suggests that he or she is sitting still. The compelling feeling of presence in virtual environments indicates that participants rely strongly on visual cues. Of note, some authors have proposed that a sort of bi-location is possible in such a situation, by which one feels to a certain degree being localized simultaneously in both the real and virtual environments (Furlanetto et al. 2013; Wissmath et al. 2011), which has also been described in a clinical condition called heautoscopy (e.g., Blanke & Mohr 2005; Brugger et al. 1994).

Neuroimaging studies in healthy participants showed that self-identification with—and self-localization at—a position of a virtual avatar seen from a third-person perspective activates the left inferior parietal lobe (Corradi-Dell’acqua et al. 2008; Ganesh et al. 2012). Corroboratively, people who are addicted to video-games show altered processing in a left posterior area of the middle temporal gyrus (Kim et al. 2012). These studies converge in their conclusion that multimodal areas in the temporo-parietal junction are involved in altered self-localization in virtual reality. As mentioned before, the temporo-parietal junction is a main region for vestibular processing. We thus hypothesize that the feeling of presence might be mediated by vestibular signals, which should be directly tested by assessing whether the feeling of presence can be modified by caloric and galvanic vestibular stimulation.

4.5.2 Physiological and vestibular mechanisms of self-location

4.5.2.1 Categories of cells coding self-location and self-orientation

Electrophysiological investigations in rodents have identified three categories of neurons encoding specifically where the animal is located, how its head is oriented, and how the animal moves in its environment (see Barry & Burgess 2014, for a recent review). These neurons are referred to in the literature as “place cells,” “head-direction cells,” and “grid cells”. In rats, place cells have been recorded as early as the 1970s in the hippocampus, and later in the subiculum and entorhinal cortex (O’Keefe & Conway 1978; O’Keefe & Dostrovsky 1971; Poucet et al. 2003). The firing rate of these neurons increases when the animal is located at a specific position within the environment. This activity is strongly modulated by allocentric signals (visual references in the environment) and vestibular signals (Wiener et al. 2002). Place cells have later been identified in several other animal species including mice (McHugh et al. 1996), bats (Ulanovsky & Moss 2007), monkeys (Furuya et al. 2014; Ludvig et al. 2004; Matsumura et al. 1999; Ono et al. 1993) and humans (Ekstrom et al. 2003; Miller et al. 2013). Head-direction cells were first recorded in the rat postsubiculum and later in several nuclei constituting the Papez circuit, such as the dorsal thalamic nucleus and lateral mammillary nuclei (Taube 2007). They were also found in the retrosplenial and entorhinal cortex. Electrophysiological recordings revealed that head-direction cells “discharge allocentrically as a function of the animal’s directional heading, independent of the animal’s location and ongoing behavior” (Taube 2007). Head-direction cells have also been identified in the monkey hippocampus (Robertson et al. 1999). Finally, grid cells have been identified in the rat medial entorhinal cortex, but also in the pre- and parasubiculum (Boccara et al. 2010; Sargolini et al. 2006). Grid cells fire for multiple locations of the animal within its environment. Altogether, these locations form a periodic pattern, or “grid,” spanning the entire surface of the environment. More recently, electrophysiological recordings have shown grid cells in mice (Fyhn et al. 2008), bats (Yartsev et al. 2011) and monkeys (Killian et al. 2012), and even probable homologues of grid cells in the human hippocampus (Doeller et al. 2010; Jacobs et al. 2013).

4.5.2.2 Place cells in the human hippocampus and “virtual” self-location

We can only speculate about the neural mechanisms of place and head-direction specific coding in the human brain. With the non-invasive neuroimaging techniques available to date (fMRI, PET, scalp electroencephalography (EEG), near-infrared spectroscopy (NIRS)), it remains difficult to investigate neural activity of potential human homologues of place cells, head-direction cells and grid cells (for fMRI identification of grid cells, see Doeller et al. 2010). Single-unit recordings can only be achieved during rather rare intracranial EEG carried out for presurgical evaluations of drug refractory epilepsy.

In a seminal intracranial EEG study conducted in 7 epileptic patients, Ekstrom et al. (2003) identified neurons with place selectivity in the hippocampus. Patients were immersed in a virtual environment and played a taxi driver computer game, picking up customers at one location in the virtual town and delivering them to another location of the town. As illustrated in figure 6, a neuron recorded in the right hippocampus had a significantly stronger firing rate when the patient was virtually “located” in the upper left corner than in any other location of the virtual town, showing its place selectivity. The authors found that 24% of neurons recorded in the hippocampus displayed a pattern of place selectivity, a proportion that was significantly larger than in the other brain structures they explored. Using a very similar procedure in a virtual environment in patients with intracranial electrodes, a recent study identified probable grid-like cells in humans (Jacobs et al. 2013). They were predominantly located in the entorhinal cortex and anterior cingulate cortex.

Interestingly, in both studies, patients did not physically move but moved virtually using button presses on a keyboard or a joystick. Nevertheless, the firing rate of these neurons changed as a function of the “virtual” location of the participants within the virtual environment. This observation indicates that both hippocampal “place cells” as well as entorhinal and cingulate “grid-cells” were coding the patient’s location in the virtual word on the basis of allocentric visual signals, rather than the patient’s position in the real world. Although the findings about these properties of the hippocampus have been mostly interpreted in the research field of spatial navigation and memory Burgess & O’Keefe 2003), we make a new proposition that they can also shed light on the neural underpinnings of bodily self-consciousness, especially on how the brain localizes the self both in everyday life as well as in situations of multisensory conflicts.

Figure 6: Map illustrating the firing rate of one cell in the right hippocampal showing a pattern of place selectivity. The rectangular map represents the virtual town explored by the participant using key presses on a keyboard and the red line represents the participant’s trajectory within the virtual town. The nine white boxes indicate the location of buildings in the virtual town (SA, SB, and SC represent three shops that were “visited” by the participant). Colors from blue to red in the background represent the firing rate of the hippocampal cell as a function of the participant’s location in the virtual town. This neuron displays a significantly higher firing rate when the participant was located in the left upper part of the virtual environment (location showed by a black square). Reproduced from Ekstrom et al. (2003).

As mentioned earlier, the experience of self-location can be manipulated by creating conflicts between visual cues about the location of one’s own body (or an avatar) in the external word and tactile or other somatosensory signals (Ehrsson 2007; Lenggenhager et al. 2011; Lenggenhager et al. 2009; Lenggenhager et al. 2007). These visuo-tactile conflicts can induce the perception of being located closer to the avatar. The recent use of these visual-tactile conflicts during fMRI recordings showed that the apparent changes in self-location and visuo-spatial perspective were related to signal changes in the temporo-parietal junction, not in the hippocampus (Ionta et al. 2011). It is not clear whether hippocampal place cells’ activity can be recorded with the large-scale, non-invasive functional neuroimaging techniques available. Yet, we predict that visuo-tactile conflicts, by modifying the experienced self-location, should also modify the neural activity of place cells and grid cells and their vestibular modulation (see next section), as showed during navigation in immersive virtual environments (Ekstrom et al. 2003; Jacobs et al. 2013). Future research using intracranial EEG recordings in epileptic patients should endeavour to study directly the relation between place cell activity and the experience of human self-location in situations of conflicting multisensory information.

4.5.2.3 Vestibular signals and place cells

In this section, we emphasize the contribution of vestibular signals to the neural coding of self-location in the hippocampus. As mentioned above, the firing rate of place cells is strongly modulated by allocentric signals, a finding replicated in several studies in rodents (Wiener et al. 2002). Vestibular signals have also been shown to modulate the firing pattern of the hippocampal place cells, which is necessary when animals navigate in darkness (O’Mara et al. 1994).

For example, Stackman et al. (2002) temporarily inactivated the vestibular system of rats using bilateral transtympanic injections of tetrodotoxin (TTX). TTX abolishes almost immediately neural activity in the vestibular nerve, producing a temporary vestibular deafferentation, mimicking the situation of patients with a bilateral vestibular loss. Figure 7 illustrates changes in the firing rate of ten hippocampal neurons before and after TTX injection. Before TTX injection, hippocampal neurons displayed a typical pattern of place selectivity when the animal explored the circular environment. A major finding of this study was that as early as one hour after vestibular deafferentation, the location-specific activity of the same hippocampal neurons was strongly disturbed. In particular, the vestibular deafferentation reduced the spatial coherence and spatial information content that usually characterize the place cells. These disorders remained between thirty-six and seventy-two hours after TTX injection, despite the fact that the rats continued to explore their circular environment and had normal locomotor activity twelve hours after TTX injection. These results indicate that place cells are continuously integrating vestibular signals to estimate one’s location within the physical environment and that vestibular signals strongly contribute to one of the most important neural mechanisms of self-location.

The activity of place cells or grid cells has not been recorded after vestibular deafferentation in humans. Nevertheless the neural consequences of vestibular lesions on place cells (Stackman et al. 2002) and head-direction cells (Stackman & Taube 1997) in animal models corroborate the effects of unilateral and bilateral vestibular lesions in humans. Patients with vestibular disorders may experience spatial disorientation as measured during path completion tasks (Glasauer et al. 1994) and navigation in virtual environments (Hüfner et al. 2007; Péruch et al. 1999). We propose that vestibular disorders, by disorganizing the firing pattern of place cells in the human hippocampus (and in other brain regions containing place cells) may strongly disturb the sense of self-location and thus the coherent sense of self, which could eventually even lead to disturbance of the usually very stable feeling of being located at a single place at a given time (see the strong disorganization of the place cells activity in figure 7). Another striking consequence of a bilateral vestibular loss is the induced atrophy of the hippocampus, whose volume is decreased by about seventeen percent (Brandt et al. 2005). Altogether, these data show that one neural mechanism of bodily self-location (place cells encoding of the body location in the environment) strongly relies on vestibular signals.

Figure 7: Modification of spatial selectivity of ten hippocampal place cells before and after inactivation of the vestibular apparatus with TTX injection. The colors ranging from yellow to purple represent the increase in firing rate of the place cells as a function of the location of the rat in the circular arena. From Stackman et al. (2002).