Introduction

Virtual reality (VR), which allows users to experience virtual spaces as if they were real, has recently attracted attention. Self-projection, the reproduction of the features of a real space consistent among different modalities in a virtual space, is an important element in VR1,2,3. Another important element is the reproduction of these consistent sensory stimuli in a virtual space that makes one perceive their whole body as being in the virtual space. Previous studies on bodily self-consciousness in the field of cognitive neuroscience examined how one recognized their own body, which was sense of embodiment that referred to the sensations of being inside, having, and controlling a body4,5,6,7. Humans can experience the sensation of their entire body outside of their physical body via synchronous visual and tactile stimulation. However, how auditory stimulation affects this illusion of the sensation of the entire body remains unclear.

Previous studies on sense of embodiment utilized out-of-body illusion4,5,6,7, the illusion that one’s sense of embodiment was outside the real body. These out-of-body illusions bear some similarities to out-of-body experiences described in patients with neurological problems; particularly, the experience of one’s sense of self being located in a different place from the real body. Out-of-body illusions involve changes in certain aspects of the feeling of embodiment, namely self-location and body ownership. Sense of embodiment refers to the perception that one’s body is one’s own and the experience of having control over it in a specific location, which includes places different from one’s actual physical body. It encompasses three main components: (1) self-location: the feeling that one’s body or a part of it is located in a specific position in space, (2) agency: the sense that one is the cause of their own actions and movements, and (3) body ownership: the sense that one’s body or a part of it belongs to oneself. Sense of embodiment is a crucial aspect in how we interact with the world and our sense of self. It can be affected by various factors, such as neurological conditions, virtual reality experiences, and certain psychological disorders.

Out-of-body experiences can be induced by drug abuse, general anesthesia, and sleep8. Furthermore, a previous study revealed that they could be experimentally induced in healthy participants via VR technology9. This study primarily focused on the phenomenon of an out-of-body illusion being experimentally induced via a VR environment. Ehrsson stated that tactile-visual stimuli caused participants’ sense of embodiment to arise at the location of the camera positioned behind them. Participants observed their back in real time via a head-mounted device (HMD). Subsequently, the experimenter moved a stick in a poking motion toward the bottom of the camera, which was presented as a visual stimulus as it was visible in the video. Simultaneously, another stick poked the participant’s chest as a tactile stimulus. Hence, participants felt as if their chests were being poked as the stick moved toward the bottom of the camera, which corresponded to the location where the stick poked, as if their bodies were in the camera’s position. Importantly, this out-of-body illusion was induced when visual and somatosensory stimuli were synchronized; however, it was not induced when they were asynchronized. Furthermore, Guterstam and Ehrsson conducted a follow-up experiment that involved an out-of-body illusion via both synchronous and asynchronous conditions10. In addition, Petkova and Ehrsson reported a version of an out-of-body illusion induced by synchronous finger movements in a “hand-shaking task”11.

Somatic sensation and proprioception, the sensations of recognizing the positions of one’s joints through movement, are part of somatic sensations1. Integrating information from somatic and visual sensation processing influences sense of embodiment4,12,13,14,15,16. Tactile stimuli are directly inputted to the skin as somatosensory stimuli and could affect sense of embodiment via the somatosensory system, which passes through the actual physical body. However, studies have not examined whether out-of-body illusions could be induced without somatic sensations related to tactile stimulation.

A previous study suggested that audio-visual stimuli affected sense of ownership, a part of sense of embodiment, in a mannequin’s ear17. A microphone was inserted into the left ear of a mannequin-like dummy head and played the sound of a brush tickling that ear. The participant’s actual ear felt ticklish when this sound was presented via headphones close to their actual ear, even though the ear of ​​the actual body was not being physically tickled. This suggested that simultaneous and synchronized audio-visual stimuli elicited an illusion of sense of embodiment in the mannequin’s ears. Other studies reported that when one made a sound when they tapped the floor, which seemed to come from a distance, one felt as if their arms were longer than they were. This suggested that hearing could affect the recognition of one’s body2. Lesur et al.18 investigated the induction of audio-induced autoscopy. Participants heard their own sounds and those of a stranger from various perspectives. When they listened to their own sounds allocentrically, participants reported that they sensed a presence around them. Additionally, they perceived their own sounds as being closer to themselves compared with those of a stranger. Furthermore, the importance of synchronization between visual and auditory stimuli was suggested in out-of-body illusion experiments that involved the hand, a body part, rather than the entire human body. The rubber-hand illusion was an important psychological phenomenon where a person, through synchronous visual and tactile stimuli, perceived a fake rubber hand as their own. Radziun and Ehrsson19 demonstrated how the rubber hand illusion was boosted by synchronous auditory and visuo-tactile stimuli compared with asynchronous ones. They reported that the congruent auditory stimuli could influence the rubber hand illusion, which indicated its potential to elicit an illusion on our own body cognition.

Therefore, it was hypothesized that audio-visual stimuli could induce out-of-body illusions and affect sense of embodiment. However, previous studies have not investigated out-of-body illusions induced by audio-visual stimuli. If out-of-body illusions could be induced by synchronized audio-visual stimuli, they could be influenced by sensory stimuli not directly related to the body even though they were bodily sensations. This was important for improving VR experiences.

This study was the first to investigate the induction of an out-of-body illusion via synchronized audio-visual stimuli and compare it with that induced by tactile-visual stimuli. We hypothesized that significantly more out-of-body illusions would be induced in the synchronized tactile-visual stimuli condition than asynchronized condition, as with audio-visual stimuli.

We conducted two tasks. The first task provided audio-visual stimuli. This replaced the tactile stimuli with auditory sensations from the experimental design of a previous study9. The second provided tactile-visual stimuli that replicated that of a previous study9. This was done to compare the out-of-body illusion induced by the proposed synchronized audio-visual stimuli with that induced by previously reported synchronized tactile stimuli. In addition, two conditions were applied for each task: synchronized or asynchronized stimuli from two modalities. Thus, four conditions were performed for each participant based on the combination of task and synchronicity differences.

This study prepared two questionnaires as evaluation methods based on a previous study9. Questionnaire 1 enquired the experience during the task and participants responded to items related to the induction of an out-of-body illusion. Questionnaire 2 was supplemental and enquired participants’ degree of anxiety regarding the threat posed to the camera after the task. We analyzed the degree of induction of out-of-body illusion and investigated whether audio-visual stimuli induced out-of-body illusion and differences existed between synchronized audio-visual and tactile-visual stimuli.

Methods

Apparatus

An apparatus (Fig. 1(a)) was constructed and used to allow participants to observe their backs in real time via the HMD.

Fig. 1
figure 1

Experiment description. (a) Apparatus. Participants wore a head-mounted device (HMD) that presented the visual stimuli. In addition, a stereo camera was placed 1 m behind the participant, which captured images behind their back. A program, created in Unity, was executed to present the left and right camera images captured by the stereo camera to the HMD’s left and right eye, respectively. The projected image is presented. (b) Experimental task. The experimenter moved a non-sounding bell in front of the camera at 1 Hz. This movement was also reflected in the image projected on the HMD. The participant saw the bell move as a visual stimulus. Synchronized with this movement, a sounding bell was moved in front of the participant as an audio stimulus. This made it seem as if the sound of the bell could be heard from that position by the movement of the bell in the video. (c) Experimental protocol. In each condition, task and threat to the camera for Questionnaires 2 and 1 were administered. Questionnaire 2 was administered after all the conditions. (d) Overall experiment. (Written informed consent was obtained from the participants in the figures for the publication of identifying information/images in an online open-access publication.).

Full size image

This experimental set-up was based on the set-up used in Ehrsson’s study9. Participants wore a Meta Quest 2 (Meta Platforms, Inc.) as the HMD to present the visual stimuli. In addition, a stereo camera was placed 1 m behind each participant. The camera was mounted on a tripod to adjust the height of the participant’s eyes parallel to the ground and in the direction they faced. It captured images behind the participant’s back, and a program, created in Unity, was executed to present the left and right camera images captured by the stereo camera to the left and right eye of the HMD, respectively. Latency for displaying images was approximately 100 ms, which was the result of recording and comparing the times taken from both an actual stopwatch and stopwatch viewed through a stereo camera. Since the threshold for synchronized and asynchronized discrimination between the visual, tactile, and auditory sensations was approximately 250 ms, the delay was considered acceptable20.

Experiment tasks

We prepared two tasks. First, a task with audio-visual stimuli. This replaced tactile stimuli with audio stimuli from the experimental design of a previous study9. The second task involved a task with tactile stimuli, which replicated that of a previous study9. This was done to compare the out-of-body illusion induced by the proposed synchronized audio-visual stimuli with that induced by previously reported synchronized tactile stimuli.

The audio-visual stimuli task (Fig. 1(b)) utilized a bell that rang and one that did not, with the experimental setup shown in Fig. 1(c). The experimenter moved the non-sounding bell in front of the camera at 1 Hz. This movement was also reflected in the image projected onto the HMD, and the participants viewed the bell movement as a visual stimulus. Synchronized with this movement, a sounding bell was moved in front of the participants as audio stimulus. Hence, it appeared as if the bell’s sound could be heard from that position by its movement in the video. The sounding bell was not shown in the video.

Tactile-visual stimuli was provided to replicate the previous study’s experiment and compare it with the experimental design of audio-visual stimuli in this study9. This task utilized two sticks. A stick was moved toward the bottom of the camera in a poking motion at 1 Hz. Its movement was reflected in the image projected on the HMD, and participants saw the stick move toward the bottom of the image as a visual stimulus. Synchronized with this movement, the other stick poked their chest as a tactile stimulus. The stick that poked the participant’s chest was not shown in the video.

Experiment conditions

Synchronized conditions were conducted in which visual and tactile stimuli or visual and audio stimuli were presented synchronously (Fig. 2). Furthermore, asynchronized conditions were conducted in which the two stimuli were presented asynchronously. In the synchronous conditions, images captured by a stereo camera were streamed to the HMD in real time. Consequently, the movements of the bells and sticks in the video appeared to be synchronized with the sound of the bells and poking of the chest, respectively. In the asynchronized condition, the video was captured by a stereo camera and streamed to the HMD with a 500 ms delay. Consequently, the movement of the bells and sticks in the video was delayed by 500 ms compared with the sound and poking of the chest, respectively. A previous study reported that out-of-body illusion was not induced under asynchronized conditions in which visual information was delayed by 500 ms. The synchronized/asynchronized discrimination threshold between visual and tactile sensations and between visual and auditory sensations was approximately 250 ms. Therefore, a 500 ms delay was considered sufficient as an asynchronized condition20.

Fig. 2
figure 2

Task descriptions. Side view during the task (left column) and image projected on the HMD (right column). Each stimulus is presented such that the upper and lower images are repeated. (a) The audio-visual stimuli task utilized a bell that could ring and one modified to not ring. The experimenter moved the non-sounding bell in front of the camera at 1 Hz. This movement was also reflected in the image projected on the HMD. The participant saw the bell move as a visual stimulus. Synchronized with this movement, a sounding bell was moved in front of the participant as an audio stimulus. This made it seem as if the sound of the bell could be heard from that position by the movement of the bell in the video. The sounding bell was not shown in the video. (b) Task of providing tactile-visual stimuli replicated the previous study’s experiment and compared it with the experimental design of the audio-visual stimuli in this study. In this task, two sticks were prepared. A stick was moved toward the bottom of the camera in a poking motion at 1 Hz. This movement was reflected in the image projected on the HMD, and participants saw the stick move toward the bottom of the image as a visual stimulus. Synchronized with this movement, the other stick poked the participant’s chest as a tactile stimulus. The stick poking the participant’s chest was not shown in the video. (Written informed consent was obtained from the participants in the figures for the publication of identifying information/images in an online open-access publication.).

Full size image

The experimental condition combinations were synchronized audio-visual, asynchronized audio-visual, synchronized tactile-visual, and asynchronized tactile-visual. The four combinations of synchronicity and experimental tasks were performed in a random order.

Experiment protocol

First, the stimuli were provided randomly for between 50 and 80 s. Subsequently, a threat was provided by a hammer after the task, as described below. Participants answered questionnaire 1 as an evaluation. These sequences were performed for each condition. At the end of the experiment, participants answered questionnaire 2. Figure 1(d) presents the experimental protocol.

Participants

This study recruited 20 healthy native Japanese speakers (10 males and 10 females) aged 22.6 ± 1.17 years (mean ± standard deviation). Before the experiment, the participants were briefed on the precautions and signed an informed consent form. The study was conducted in accordance with the principles set by the Declaration of Helsinki and its future amendments. This study was approved by the Ethics Committee of the Tokyo Institute of Technology (approval number: A23046). Of the participants, four had errors in the experimental procedures, and data from a re-run of the experiment were used. Furthermore, one participant was excluded from questionnaire 1’s analysis as his comments in the unstructured interview clarified that he had misinterpreted the meaning of the questions.

Evaluations

Evaluation of the induction of out-of-body illusion by stimuli

Questionnaire 1 (Table 1) was administered after each condition to evaluate whether an out-of-body illusion was induced. It was a Japanese translation of a questionnaire developed in a previous study9. As indices for evaluation, the mean values of Q1–Q3 and Q4–Q10 were calculated as “Out-of-Body Illusion Index (OBI Index)” and “Control Index,” respectively21,22,23,24. In total, three main results were expected. First, the OBI Index would be significantly higher in the synchronized conditions than in the asynchronized conditions. Second, the OBI Index would be significantly higher than the Control Index in the synchronized conditions. Third, the Control Index would be negative in all the conditions, regardless of whether an out-of-body illusion was induced.

Table 1 Questionnaire 1 on induction of out-of-body illusion.
Full size table

Evaluation of anxiety regarding the threats

Similar to in the previous study9, participants were threatened after the task to evaluate their anxiety and investigate whether both synchronized audio-visual and tactile-visual stimuli resulted in a higher degree of localization of the body in extra-personal space. Immediately after the task, the bell or stick was replaced with a rubber hammer that swung toward the bottom of the camera. Participants viewed the image of the hammer through the HMD (Fig. 3).

Fig. 3
figure 3

Threat description. Side view during the threat (left column) and image projected on the HMD (right column). A hammer was swung as a threat, as shown in the bottom panel from the top image. To investigate whether participants’ sense of embodiment was generated by the camera’s position, they were provided a threat after the task to assess their anxiety. Immediately after the task presented the stimuli, the bell or stick was replaced by a rubber hammer that swung toward the bottom of the camera. Participants simultaneously viewed the image of this hammer through the HMD. If participants’ sense of embodiment was generated by the position of the stereo camera rather than their own bodies, they should feel more anxious regarding a threat directed at the camera. Thus, we expected significantly higher anxiety responses in the synchronized conditions than asynchronized conditions. (Written informed consent was obtained from the participants in the figures for the publication of identifying information/images in an online open-access publication.).

Full size image

After all the experimental conditions were completed, questionnaire 2 (Table 2) was administered. Participants reported their anxiety regarding the threat posed by the hammer on a scale from 1 (I was not anxious at all) to 10 (I was the most anxious I could be). If participants’ sense of embodiment was generated by the position of the stereo camera, rather than their own bodies, they would have felt more anxious regarding the threat directed at the camera. Thus, we expected significantly higher anxiety responses in the synchronized conditions than asynchronized conditions.

Table 2 Questionnaire 2 on the threat posed by the hammer.
Full size table

Results

Evaluation results of the induction of out-of-body illusion by stimuli

We evaluated two aspects of whether an out-of-body illusion was induced in the two conditions, synchronized audio-visual and tactile-visual stimuli, via questionnaires on “out-of-body illusion” and “threat anxiety,” as shown in Fig. 4.

Fig. 4
figure 4

Results of the two questionnaires (*p < 0.05, **p < 0.01, ***p < 0.001). The display of significant differences between modalities is omitted for readability in (a). Errors bars represent the SEM. (a) Graph presents the results of Questionnaire 1’s “Out-of-Body Illusion Index (OBI Index)” (orange) and “Control Index” (purple) scores (each condition is displayed as bars). For both audio-visual and tactile-visual stimuli, the OBI Index was significantly higher in the synchronized conditions than asynchronized conditions (p = 0.010 and p = 0.011, respectively). The OBI Index was significantly higher in the synchronized conditions than the Control Index (ps < 0.001) in both the tasks. (b) Graph presenting the results of Questionnaire 2. Scores for the synchronized (blue) and asynchronized (orange) conditions in the task for each stimulus are displayed as bars. The degree of anxiety was significantly higher in the synchronized condition than asynchronized condition for both audio-visual stimuli (p = 0.006) and tactile-visual stimuli (p < 0.001).

Full size image

A Friedman test was performed on the results of questionnaire 1 and confirmed significant differences (p < 0.001). To select a non-parametric or parametric test as the statistical multiple comparison method for the subtest, the normality of the data to be analyzed was confirmed. The Shapiro-Wilk normality test was used to assess the normality of the distribution of the data-set for each experimental condition. Consequently, the null hypothesis that the distribution would be normal was rejected for some datasets (ps < 0.05). Therefore, in this study, a Wilcoxon signed-rank test with Holm-Bonferroni method, a non-parametric statistical test, was used. Concretely, it was performed for multiple comparisons. For both audio-visual and tactile-visual stimuli, the OBI Index was significantly higher in the synchronized condition than asynchronized condition (p = 0.010, p = 0.011, respectively). The OBI Index was significantly higher than the Control Index in the synchronized condition (ps < 0.001) in both the tasks. Furthermore, the Control Index did not differ significantly between the synchronized and asynchronized conditions. Thus, an out-of-body illusion was induced via synchronized audio-visual and tactile-visual stimuli. In addition, under the same conditions, the OBI and Control indices did not differ significantly between tactile-visual and audio-visual stimuli. Hence, synchronized audio-visual stimuli and tactile-visual stimuli similarly induced an out-of-body illusion.

Evaluation results of anxiety regarding the threats

A Friedman test was performed on the results of Questionnaire 2 and confirmed a significant difference (p < 0.001). Furthermore, a Wilcoxon signed-rank test with Holm-Bonferroni method was performed for multiple comparisons. Participants’ degree of anxiety was significantly higher in the synchronized condition than asynchronized condition for both the audio-visual (p = 0.006) and tactile-visual stimuli (p < 0.001).

Supplementary results

The supplementary datasheet includes further detailed analyses. Table S1 presents assumption checks for the statistical analyses. Table S2 presents details on the multiple comparison results in Questionnaire (1) Table S3 and Figure S1 present the results of all the individual questions in Questionnaire 1 for all the conditions. Table S4 presents details on the multiple comparison results in Questionnaire (2) Figures S2 and S3 illustrate the boxplots in Questionnaires 1 and 2, respectively. Tables S5, S6, S7, and S8 presents the correlation analysis results between the experimental conditions in Questionnaires 1 and 2.

Discussion

This study aimed to investigate whether synchronized audio-visual stimuli could induce an out-of-body illusion. We hypothesized that significantly more out-of-body illusions would be induced in the synchronized audio-visual condition than asynchronized audio-visual condition. Therefore, tasks with tactile-visual and audio-visual stimuli were conducted to investigate this hypothesis. In each task, the stimuli of the two modalities were either synchronized or asynchronized. Thus, four conditions were assigned to each participant, which combined the task and difference in synchronization. Subsequently, two questionnaires were administered to evaluate the induction of an out-of-body illusion.

Results of questionnaire 1 revealed that the OBI Index, which indicated whether audio-visual and tactile-visual stimuli induced an out-of-body illusion, was significantly higher in the synchronized condition than asynchronized condition. This was consistent with the results of a previous study9,10. Furthermore, the OBI Index was also significantly higher than the Control Index in the synchronized audio-visual and tactile-visual conditions, which was also consistent with previous results9,10. In addition, as hypothesized, the Control Index was not significantly different between the synchronized and asynchronized conditions for either audio-visual or tactile-visual stimuli. In Questionnaire 2, the degree of anxiety was significantly higher in the synchronized conditions than asynchronized conditions for both audio-visual and tactile-visual stimuli, similar to in a previous study9. These results indicated that audio-visual stimuli induced out-of-body illusion, which was consistent with results of previous studies. However, this study was the first to demonstrate that out-of-body illusion could be induced by synchronized audio-visual stimuli without somatic sensation stimuli. Out-of-body illusion was also induced by synchronized tactile-visual stimuli, similar to results of a previous study.

Results of Questionnaire 1 revealed no significant differences between the modalities of stimuli (tactile-visual and audio-visual stimuli) in the OBI and Control indices for the synchronized and asynchronized conditions. Similarly, no significant difference was observed in the anxiety responses in the synchronized and asynchronized conditions in Questionnaire 2, regardless of modality. Therefore, out-of-body illusion was similarly induced via both synchronized tactile-visual and audio-visual stimuli.

Previous studies used tactile stimuli to induce an out-of-body illusion9,11,25. Therefore, stimuli were required to be provided directly to the participant’s body to generate an out-of-body illusion. Conversely, when an out-of-body illusion was induced via synchronized audio-visual stimuli, it could be induced by providing stimuli not required to be directly provided to the participants’ bodies. These results demonstrated that audio-visual stimuli could induce an out-of-body illusion throughout the body. Moreover, previous studies also reported that an out-of-body illusion involved changes in bodily self-location10,26,27. Positive value of the OBI index in this study implied that the study’s variant based on audio-visual synchronized stimulus also involved changes in self-location.

Based on previous studies, what mechanism triggers an out-of-body illusion? This phenomenon may be related to information processing in each sense and their multisensory integration. Multisensory integration is the construction of multiple sensory stimuli, such as what we see, hear, and touch, as coherent rather than separate perceptions. Among the multiple regions of the brain where multisensory integration occurs, the temporoparietal junction is considerably involved in an out-of-body illusion. Previous studies demonstrated that an out-of-body illusion owing to drug abuse, general anesthesia, and sleep were related to the inability to integrate multisensory information from the actual body at the temporoparietal junction8,28. Guterstam et al.29 reported that people experienced an out-of-body illusion in a functional magnetic resonance imaging study, which scanned human brain activity. They suggested that involvement of two brain networks was associated with changes in body ownership and self-location, respectively. Illusory self-location was associated with activity in the hippocampus and posterior cingulate, retro-splenial, and intraparietal cortices. Furthermore, body ownership was associated with premotor-intraparietal activity. These brain regions could also be involved in an out-of-body illusion induced by synchronous audio-visual stimulation.

Previous studies have not thoroughly investigated the neuroscientific aspects of an out-of-body illusion induced by audio-visual synchronous stimulation. Therefore, to address this issue, we considered previous research on out-of-body illusion via visual-somatosensory synchronous stimulation. In an out-of-body illusion based on visual-somatosensory synchronous stimulation, different neural information processing systems in the brain managed visual and somatosensory stimuli. The illusion was elicited from their multisensory integration12,14,16. Therefore, when an out-of-body illusion was induced by synchronous visual-somatosensory stimulation, visual and somatosensory information were initially processed separately in the nervous system and subsequently integrated. Next, we considered an out-of-body illusion based on audio-visual synchronous stimulation according to previous findings. Specifically, visual information is processed in the visual cortex in the occipital lobe, while auditory information is processed in the auditory cortex in the temporal lobe. Furthermore, the temporoparietal junction, which is associated with an out-of-body illusion, integrates visual and auditory inputs30. When visual and auditory stimuli are presented synchronously31 and from the same location32, more multisensory neurons involved in multisensory integration are activated. This suggests that visual and auditory information may be closely related. Thus, in an out-of-body illusion based on audio-visual synchronous stimulation, audio-visual synchronous stimulation may be processed separately and subsequently integrated, as with visual-somatosensory synchronous stimulation.

Further investigation is required to determine the features of audio-visual stimuli that are sufficient to induce an out-of-body illusion. Our results suggested that spatiotemporally matched audio-visual stimuli could have sufficient stimulus features to induce an out-of-body illusion. A previous study suggested that humans perceived time in near and far space differently and subsequently distance to the signal source affected sound33. Hence, humans may experience a sense of embodiment when both the visual and auditory stimulus sources are spatiotemporally consistent. Specific characteristics of simultaneous audio-visual stimuli that elicit an out-of-body illusion will be further elucidated in future research based on this study’s findings.

This study has some limitations. First, only the questionnaire results were included and no objective indirect tests were conducted. Specifically, we did not measure threat-evoked skin conductance responses (SCR) during the threat procedure and solely relied on fear ratings. However, fear ratings were correlated with SCR and brain activity in areas related to fear and anticipation of pain9,29,34,35. Previous studies reported that fear reports, such as when illusory owned body parts were physically threatened in body ownership illusions, were correlated with objective physiological measures of fear and pain anticipation responses. The more intensely people experienced a body ownership illusion, the greater the reported fear when the illusory body was threatened. Moreover, the stronger the fear, the more pronounced the threat-evoked skin conductance responses (SCR) and threat-evoked BOLD activation in areas related to fear and pain anticipation, such as the anterior cingulate cortex and insular cortex9,29,34. Therefore, a fear report was an index of the emotional defense reactions triggered by threats to one’s own body or what was perceived as one’s own body in bodily illusions. Fear ratings are a valid indicator of illusory changes in body ownership and self-location4. Future studies should further verify this phenomenon based on questionnaires and physiological indicators, such as the SCR. Additionally, evaluation of the sense of self-location via measurement methods other than questionnaires may also be useful to further investigate the phenomenon of an out-of-body illusion10,25,26.

This study’s paradigm was an out-of-body illusion based on a first-person perspective. Other studies investigated out-of-body illusions based on a third-person perspective via a VR environment with a VR HMD36,37. These studies demonstrated that visual and tactile sensations induced significantly more out-of-body illusions when stimuli were provided synchronously than asynchronously. Another study reported that sense of embodiment changed when the movements of the virtual body were synchronized with those of the participant’s real body25. Ehrsson’s9 paradigm involved multisensory integration from a first-person perspective, which resulted in a strong perceptual illusion of body ownership and self-location. Conversely, Lenggenhager’s36 paradigm involved visuo-tactile stimulation from a third-person perspective, which led to a weaker cognitive feeling of illusory self-identification. Full-body illusions experienced from the first-person perspective were significantly stronger than those experienced from the third-person perspective38,39,40,41,42. Future studies should verify the experimental conditions based on the third-person perspective.

Conclusion

This study aimed to investigate whether synchronized audio-visual stimuli induced an out-of-body illusion and compare it with synchronized tactile-visual stimuli. Specifically, based on Ehrsson’s experimental design9, participants sat in a chair and observed their own back with a stereo camera of the HMD. In the synchronized audio-visual stimuli condition, a sound-making bell was rung in front of the participant’s body to provide an auditory stimulus. Simultaneously, a silent bell was rung in front of the stereo camera in synchronization to provide a visual stimulus. In the synchronous tactile-visual stimuli condition as the control condition, a tactile stimulus was a stick that poked the participant’s chest, and a visual stimulus was evoked by moving the stick in front of the stereo camera in synchronization to provide a visual stimulus. As further control conditions, asynchronized audio-visual and tactile-visual stimuli conditions were also implemented. Consequently, it was indicated that the synchronized tactile-visual stimuli condition induced an out-of-body illusion, similar to in previous studies. Surprisingly and importantly, an out-of-body illusion was induced under the synchronized audio-visual stimuli condition. This suggested for the first time that the out-of-body illusion, as an illusion related to the sense of embodiment that targeted the entire body, could be induced by synchronized audio-visual stimuli alone, without a tactile stimulus. New findings will contribute to elucidating the mechanism of out-of-body illusions and development of applications based on out-of-body illusions, such as in virtual, augmented, and mixed reality.