Visuo-acoustic stimulation that helps you to relax: A virtual reality setup for patients in the intensive care unit


After prolonged stay in an intensive care unit (ICU) patients often complain about cognitive impairments that affect health-related quality of life after discharge. The aim of this proof-of-concept study was to test the feasibility and effects of controlled visual and acoustic stimulation in a virtual reality (VR) setup in the ICU. The VR setup consisted of a head-mounted display in combination with an eye tracker and sensors to assess vital signs. The stimulation consisted of videos featuring natural scenes and was tested in 37 healthy participants in the ICU. The VR stimulation led to a reduction of heart rate (p = 0. 049) and blood pressure (p = 0.044). Fixation/saccade ratio (p < 0.001) was increased when a visual target was presented superimposed on the videos (reduced search activity), reflecting enhanced visual processing. Overall, the VR stimulation had a relaxing effect as shown in vital markers of physical stress and participants explored less when attending the target. Our study indicates that VR stimulation in ICU settings is feasible and beneficial for critically ill patients.


After prolonged stay in intensive care units (ICU), patients often complain about cognitive impairment which can lead to a reduction of health-related quality of life as compared to before ICU admission1,2. Such patients may initially have suffered from severe infection, multiple trauma injuries, heart failure and other diseases leading to multiple organ failure that requires intensive or emergency treatment. In recent years, intensive care medicine has seen a culture change in that the focus on improving survival rates has been expanded to include long-term health-related quality of life after discharge3. Health disorders seen in patients, collectively referred to as post intensive care syndrome, include persistent impairments across physical, psychiatric and cognitive domains4. Long-term cognitive impairment affects between 50% and 75% of patients, independent of cognitive status prior to admission, involvement of brain injury and age of the patients4,5,6,7. Long-term cognitive impairment frequently manifests first as acute brain dysfunction (delirium and coma) when patients are being treated for critical illness in the ICU8. Thereafter, long-term cognitive impairment is known to affect neurocognitive functions such as information processing, concentration, attention, learning and memory as well as executive functions6,9.

Current treatment modalities that address cognitive impairment in ICU survivors include preventive and remedial measures as well as drug therapy during and after the course of the ICU stay. Preventive strategies, on the one hand, are aimed at reducing various risk factors for cognitive dysfunction and psychological morbidity during ICU stay. These include early physical rehabilitation, sleep promotion, minimization of use of sedatives and psychological support (e.g. counselling)10. These approaches have been reinforced with findings from studies that have shown that early mobilization and physical training, and choice of sedation can indeed reduce cognitive impairment5. Furthermore, the “ABCDE” strategy (awakening, breathing, choice of sedatives, daily delirium monitoring, and early mobilization) has been specifically targeted at reducing sedation and delirium that are suspected to cause post-ICU cognitive impairment11. While these strategies were shown to reduce delirium, it has not yet been demonstrated that they actually prevent cognitive impairment4. Remedial strategies, on the other hand, use cognitive interventions aimed at retraining impaired cognitive functions and thus improve everyday functions. Several authors have highlighted the potential benefit of early cognitive and physical therapy during the ICU stay. However, most studies were conducted in critically ill patients after the stay in the ICU5,6,12,13 and therefore evidence from interventional studies in critically ill patients remains scarce particular during ICU stay4. It is assumed that early onset of intervention will have the biggest impact on later cognitive and psychological morbidity by both preventing cognitive impairments from developing and improving cognitive functions after discharge4,10.

A critical factor in critically ill patients is the typical ICU environment characterized by constant exposure to unpatterned stimuli such as light, noise, lack of daily living routine and limited mobility14. Patients are exposed to the extremes of both sensory overload and sensory deprivation and are at risk of suffering alterations in sensory perception that may result in cognitive dysfunction15. Sensory overload can be produced by stimulation at above normal levels with continuous and excessive unfamiliar sensory input16. Sensory deprivation is caused by reduced amounts or variability of stimulation with decreased or monotonous, unpatterned and meaningless sensory input17. It is the resulting combination of both stimulus overload and simultaneous stimulus deprivation that is believed to lead to a loss of differentiated perception, orientation and cognitive and attentional fatigue in critically ill patients18,19,20. This perception disorder is further reinforced by medication and therapeutic measures such as analgesia or sedation. These environmental stressors also cause considerable psycho-physiological stress, especially in older patients.

Chen et al. and Turon et al. suggested that virtual reality (VR) technology might be useful to stimulate critically ill patients during ICU stay21,22. VR using head-mounted displays (HMD) seems to be a promising tool to cognitively stimulate critically ill patients in a safe way with the ability to fully control virtual environments that users can interact with and immerse themselves in. Thus, virtual-reality based cognitive stimulation can be used for improving cognitive functioning in critically ill patients while at the same time preventing stimulus overload by isolating the patient from disturbing external visual and acoustic input. To avoid both sensory overload and sensory deprivation, we selected natural restorative environments with a neutral content.

In VR a distinction is made between virtual environments (VE) and virtual worlds (VW). VE use two or three-dimensional environments that include characters and objects that users can see, hear and interact with in a realistic manner in real time. VW are persistent and social worlds where users can interact with other users using an artificial avatar23,24. Furthermore, one can distinguish in VR between active mode navigation where users interact in real-time with the VE and thus have a sense of control over the scenario (e.g. controlling of an avatar), whereas in passive mode navigation (e.g. watching immersive videos) they have not. In both cases VR permits the users to experience VE that would otherwise be difficult or even impossible to implement in real life. To achieve this goal, two components are needed: an appropriate algorithm to render the moving objects and the surrounding environment and a physical device (e.g. HMD) to present the images to the user25. Thus, one of the core elements of VR is the interface between the user and the virtual world. Furthermore, VR environments need to be carefully designed in order to minimize sensory conflicts (e.g. between visual and vestibular system) to avoid cybersickness26,27.

To our knowledge, there is currently no data on the use of VR for cognitive stimulation in critically ill patients during an ICU stay. One challenge when using VR stimulation in critically ill patients is to select a suitable virtual scenario and ensure that it has a relaxing effect and engages the patient’s attention28. An established line of research has suggested that visual exposure to natural environments (e.g. landscapes, vegetation, and water) has protective effects against environmental stressors and restores physiological, emotional and attentional functions29,30,31. These restorative effects have been explained within two complementary theories: The Stress Recovery Theory posits that natural environments support positive changes in emotional states and recovery from psycho-physiological stress through a relaxing effect in the parasympathetic system32. This is relevant since the ICU environment can create considerable stress in patients that affects well-being and recovery. Furthermore, cognitive or attentional capacity, can become fatigued in stressful environments33. This is crucial for critically ill patients facing stimulus overload and high demand on cognitive capacity in ICU settings that increase the risk for mental fatigue and stress response. The Attention Restoration Theory states that exposure to natural environments is less cognitively demanding and promotes a sense of being away, thus enabling attentional capacity to rest and be restored30,33,34.

In this study, we propose to use video oculography (integrated into the HMD) to measure the visual exploration behaviour (e.g. where the patient “looks at” and whether the eyes are open). To measure participant’s reactions to the VR Stimulation, we use vital parameters that are routinely recorded in ICU settings (e.g. heart rate, blood pressure etc.), while presenting three different nature VR videos, i.e. landscape, water worlds and animals.

We hypothesise that stimulation with the three nature VR videos has a relaxing effect (measured via vital markers of physical stress35), the visual explorative behaviour follows the scene contents (e.g. participants look at a visual target, detailed description in method section), creates a sense of immersion (measured by an 5 scale questionnaire) and the technical implementation meets the hygiene and security requirements for the use in critically ill patients. In the present study, feasibility and proof of concept of VR Stimulation was investigated in 37 healthy participants.


Descriptive analysis of pooled vital sign- and oculomotor measurements

A total of 37 (23 female and 14 male) healthy younger and older adults between the ages of 20 and 85 (M = 48, SD = 17) participated in this study. In the initial measurement before the recovery phase (i.e. when the participants were lying on the bed 10 minutes before the start of the stimulation) the mean arterial pressure (MAP) was M = 86.8 mm Hg (SD = 23.1), the heart frequency (HF) M = 68.7 beats/min (SD = 21.1) and the peripheral capillary oxygen saturation (SpO2) M = 92.3% (SD 22.4). The respiratory frequency (RF) could not be measured accurately since participants moved or the ECG was not placed properly at the beginning. The vital sign measurements MAP (M = 87.67 mm Hg, SD = 1.67), HF (M = 67.02 beats/min, SD = 1.08) and RF (M = 21.08 Imp/min, SD = 2.29) showed a significant decrease during presentation of the three VR scenarios, indicating a negative time effect. The SpO2 (M = 97.05%, SD = 0.09) to measure deterioration of participant’s condition remained constant (Fig. 1). However, oculomotor data did not change much across the VR stimulation. The fixation duration per second across stimulations was M = 170.15 ms (SD = 55.42), the number of fixations M = 0.81 fixations (SD = 0.25), the saccade amplitude M = 13.74 degrees (SD = 2.60) and fixation/saccade ratio M = 0.24 (SD = 0.09).

Figure 1

Vital parameters pooled with oculomotor data of healthy participants (n = 37, age M = 48 years, SD = 17). The oculomotor data (fixation duration, number of fixations, saccade amplitudes and fixation/saccade ratio) is based on moving window calculations, whereas the vital sign measurements (MAP, HF, RF and SpO2) were sampled at each data points separately.

Identification of the visual target that appeared at the beginning, middle, and end of each video was reflected in peaks of number and duration of fixations and fixation saccade ratios, whereas, reversely, the saccade amplitudes dipped at the start, in the middle, and at the end of each video.

Effect of time and ratings on vital signs

Data analysis using linear mixed effect models (LMEMs) to generalize the findings revealed a significant reduction of vital signs (HF, t (276) = −2.704, p = 0.007, RF t (276) = −2.020, p = 0.044 and MAP, t (276) = −1.981, p = 0.049) during the course of the stimulation (negative time effect) across all participants, except in SpO2 (t (277) = −1.377, p = 0.170) which did not significantly decrease during the stimulation. However, in HF, there was a significant difference in slope between participants (inter-individual difference) during the course of the study. For RF and MAP the time effect was significantly higher for participants with higher baseline vital sign measurements. Overall, there was no significant interaction between time and age (fixed interaction effect) except for MAP (t (276) = −2.149, p = 0.033) indicating an increasing reduction of vital signs during the course of the stimulation (negative time effect) with increasing age. There was no significant effect of video type on any of the vital signs measurements. Furthermore, none of the questionnaire scales demonstrated a significant influence on vital signs. Given the high relationship between time and video type (given by the design), the variable video type was not removed from the LMEMs.

Effect of target presentation and time in oculomotor data

Generalized findings using LMEM analysis as shown in Table 1 across all participants revealed that oculomotor data during stimulation stayed more or less constant (time effect), indicated by the small effect in number of fixations, fixation duration, saccade amplitude and fixation/saccade ratio. Therefore, the time effect in oculomotor data is negligible. However, there was a strong significant effect of target presence on the number of fixations, fixation duration, saccade amplitude and fixation/saccade ratio.

Table 1 Effect of time and target in oculomotor data by using LMEM analysis.

Given the large set of oculomotor data, it was difficult to interpret the p-values (p-value inflation). Therefore, the LMEM accuracy (effect size) was assessed by examining the marginal (\({R}_{{LMM}(m)}^{2}\)) and conditional (\({R}_{{LMM}(c)}^{2}\))R2 as proposed by Nakagawa et al. and Nuthmann et al.36,37,38. The goodness of fit of the full model revealed that the proportion of the variance explained (\({R}_{{LMM}(m)}^{2}\)) by the fixed components was in average M = 18.89%, Min = 9.04%, Max = 32.32% while the whole LMEM, including the random and fixed components (\({R}_{{LMM}(c)}^{2}\)) was in average M = 52.64%, Min = 36.77%, Max = 71.08%.

Analysis of immersion, presence, sickness and usability

All average ratings of the five scales of the questionnaire by using t-tests (Table 2) were significantly above (i.e. usability, immersion, presence, realism, involvement) or below (i.e. sickness, oculomotor problems, disorientation) the midpoint of the scale (i.e. 2.5). The usability rating for the whole system was high and close to the maximum score of the questionnaire scale (max = 5). Sickness, oculomotor problems, and disorientation ratings were relatively low and close to the minimum score of the scale (min = 1). Immersion, presence, realism and involvement ratings were average, ranging around the midpoint of the scale with relatively large standard deviations.

Table 2 Ratings of the questionnaire by using t-tests. The items had a five-point rating scale between one and five.


Principal Results

In this study, we investigated the feasibility and proof of concept of VR stimulation in healthy participants in the ICU. In line with our first hypothesis, we found that exposure to virtual nature environments produced a relaxing effect as measured by vital signs. Second, there was no indication of fatigue in visual exploration during VR stimulation and target stimulus presentation affected visual exploration behavior (visual search vs. visual processing). Third, the VR scenario was well tolerated by the participants and allowed them to fully immerse themselves within the video content. Finally, we showed that a commercially available VR setup can be adapted to comply with hospital hygiene and safety requirements for use in critically ill patients.

Vital signs

The first main finding was that all vital sing measurements (heart frequency, respiratory frequency and mean arterial blood pressure) except blood oxygen saturation significantly decreased during the course of the VR stimulation. This can be interpreted in away, that VR stimulation using virtual nature environments was successful in reducing measures of physiological stress. Furthermore, the relaxing effect was independent of age in heart frequency and respiratory frequency. However, mean arterial blood pressure showed an increase of the relaxing effect in older participants. Since blood oxygen saturation increase did not reach significance, we find reason to assume that the participants’ condition was not deteriorated during VR stimulation.

Chirico, et al. and Joye and Bolderdijk et al. have shown that the exposure to VR nature environments can trigger specific psychophysiological patterns of parasympathetic activation, induce strong emotions and that the emotional reactions increases with the level of immersion. Emotions evoked by means of VR have been successfully implemented in treatments for anxiety and phobia stress inoculation39,40,41. Our finding of a relaxing effect could be due to the positive emotions of the immersive nature videos. This also matches the Stress Recovery Theory suggesting that exposure to virtual nature environments has a restorative influence that can reduce physiological stress responses and increase well-being32,42. This is further confirmed by recent studies that showed that stress recovery can be facilitated using immersive virtual natural scenes combined with sounds29,43.

Occulomotor data

The second main finding confirms that stimulation using natural videos was successful in capturing participants’ attention. The absence of a time effect in the oculomotor data indicates the absence of mental fatigue during VR stimulation which coincide with findings in literature44,45. In line with the Attention Restoration Theory, this might suggest that the virtual nature environments used in our study allowed participants to rest and restore attentional capacity34.

The fixation/saccade ratio represents the ratio between information processing and search activity, where a higher ratio reflects more processing and less search activity, while a low ratio reflects more search activity and less processing46. From this it is evident that participants were visually exploring the videos less when the target stimulus was present as compared to when the target stimulus was not present.

The goodness of fit of the LMEM models of the oculomotor data compared to a recent study reporting similar models36, indicating a very good fit of the models to the presented data, which further confirms the significance of the findings of oculomotor data.

Given that this study examined visual exploration using an HMD, it is difficult to relate our findings to those from previous studies that consistently used desktop screens. While the latter offers a field of view of 48 by 27 deg., the HMD in this study had a larger field of view of 95 by 106 deg. This is crucial since it was shown that the larger the visual field size, the lower the fixation durations and the larger the fixation amplitudes are47. Hence, it is hardly not surprising that the magnitude of fixation durations (170.2 ms) was almost half as high and saccade amplitude (13.7 deg.) twice as high when compared to the findings of a previous study during dynamic viewing of natural scenes (326 ms, 7.2 deg.)48. The larger visual field size in VR setups does not only more closely matches natural visual exploration behaviour, it also provides participants with a higher sense of immersion. Therefore, the use of HMD based VR stimulation, in line with relevant literature, is suitable for critically ill patients49.


The questionnaire responses, measured by a 5 – point scale (range one to five), clearly indicate that the HMD-based VR stimulation is well accepted and appreciated by the participants as shown in the high usability score (M = 4.5, SD = 0.6). Immersion (M = 3.60, SD = 1.2) and presence (M = 3.11, SD = 1.1) were also at a high level, and thus participants were not aware of their surrounding and could focus on looking at the VR scenario. We believe that participants were well isolated from the meaningless and unpatterned stimuli such as noise, light and bustle of the ICU environment. Symptoms of cybersickness (M = 1.16, SD = 0.4) were rated low on the questionnaire scale and are therefore negligible. The presented results are an important prerequisite for the future use of the implemented VR setup in critically ill patients.

Technical implementation

Overall, a beyond state of the art mobile HMD-based VR setup to stimulate participants in the ICU was successfully implemented. The setup was running solid and no data loss or any complication on participants occurred. Furthermore, the system fulfilled all standard medical hygiene and safety requirements for critically ill patients.

A number of studies have shown cognitive benefits of VR-based rehabilitation in patients with brain-injury and that VR was efficient and well tolerated in acute inpatient medical settings (e.g. pain management, eating disorders, cognitive and motor rehabilitation)50,51. To the best of our knowledge, however, no study has investigated the cognitive benefits of VR rehabilitation in critically ill patients during their stay in the ICU. We believe that VR stimulation will be beneficial for critically ill patients with neurological impairments and will become essential in the future since it has the potential for alternative and new trainings. The advantage of VR in the ICU is that it allows the patient to interact with the virtual environment via HMD directly at the bed with little or no support. In addition, VR offers the possibility to simulate many different real or imaginary scenes with an infinite number of targets and distractors while providing a safe and consistent environment52. Therefore, the stimulation can be easily adapted to critically ill patients. Typically, this opportunity is not possible for real-world exercises.


To rule out that the relaxing effect in vital signs was not simply a result from having the participant lying on a bed, we had the participants lying on the bed ten minutes prior to the start of the VR stimulation. In addition, the 2D videos were played in the VE without stereoscopic depth and active mode navigation. Nevertheless, the use of 3D VE and active mode navigation could even increase immersion and thus the relaxing effect.

In this study, we did not control for self-reported ratings of participants’ affect and emotional response to the VR stimulation. Even though induction of positive and reduction of stress-related and negative emotions plays an important role in the stress recovery effect of natural environments, this study focused on the effects of natural virtual environments on measures of physiological stress and visual exploration behaviour. Despite the very promising findings it remains unclear whether the findings can be generalized to critically ill patients.

Guidelines for future VR studies in the ICU

In order for patients to profit from the restorative immersive VR stimulation, the VR stimulation should be intuitive to use and the patients’ Richmond Agitation-Sedation Scale (RASS), a measure of sedation related to consciousness, should not be lower than minus two (light sedation, reaction to voice with eye contact)53. Furthermore, the setup should be audited by a notified hospital body regarding hygiene and medical eligibility requirements and declared secure to use in critically ill patients. This requires that all computer fans must be protected with a filter, that the system has a potential equalisation, that no current can be transmitted to the patient and that all parts in contact with the patient are disinfectable. In order that patients can fully immerse themselves, reduce stress and restore their visual attention, we hypothesize that natural VEs, VWs or stereoscopic videos must be fascinating, rich and harmonious as well as promote a sense of being away. To analyse the effect of VR stimulation on cognition in critically ill patients, cognitive tests should be administered prior to ICU admission, during the ICU stay and after discharge. Furthermore, neurological tests and medication, conscious state, delirium and duration of the stimulation should be recorded. Finally, future results should be compared to the current study with healthy participants.


VR stimulation had a relaxing effect as shown in vital markers of physical stress and was well tolerated by the participants. Furthermore, the VR stimulation was successful in engaging visual attention (exploration and processing) and allowed participants to fully immerse themselves in the virtual environment. This study indicates that VR stimulation has the potential to reduce psycho-physiological stress and to restore cognitive and attentional capacities in critically ill patients. Furthermore, the implemented HMD-based VR setup using visual and acoustic stimulation is feasible and has great potential for cognitive stimulation in critically ill patients.


Participant recruitment and Demographics

The study design was carried out in accordance with the current version of the Declaration of Helsinki and approved by the Ethics Committee of the Canton of Bern, Switzerland (KEK-Nr. 148/15). The study participants were recruited via email advertisement among the Seniors University of Bern and local research groups. All procedures related to the study were explained to the participants and a written informed consent was obtained prior to participation. The main exclusion criteria were age <18 years, visual- and auditory impairments and any long-term medication which could influence vital signs (e.g. blood pressure medication).


The study was conducted in a fully equipped 2 bed ICU cubicle. The VR setup to measure and stimulate the participants consisted of two main parts: Firstly, a custom-built desktop computer and a head-mounted display (Oculus Rift DK2 virtual glass) with a built-in SMI Oculus eye tracker system. Secondly, a vital signs monitoring system (Carescape Monitor B650, GE Healthcare, Little Chalfont, UK) was used to record vital parameters such as peripheral capillary oxygen saturation, mean arterial pressure, heart frequency and respiratory rate. This study was deliberately conducted under conditions as identical as possible to the situation critically ill patients experience in order to reduce the bias on results in future studies (Fig. 2).

Figure 2

Participant lying on the bed in the ICU during stimulation, including the whole setup.

The stimulation consisted of three 2D nature videos, each five minutes in length, played as side-by-side videos (SBS videos, scene filmed from only one point of view) in the HMD. We have chosen 2D videos instead of 3D videos since some scenes in 3D could potentially provoke fear (e.g. animal walks towards the user) or elicit vertigo (e.g. view from a mountain) in participants.

All three videos contained natural scenes that were chosen to be enjoyable and cheerful with a visual target (yellow smiley) that was shown at the beginning, the middle, and the end of each video sequence (videos available on our website: see URL). The 2D visual target with a visual angle of 4 degree was superimposed onto the centre of each video frame during the target presentation period. Participants were instructed to fixate the visual target when it was presented. The first video about landscape was filmed from a hot-air balloon perspective (i.e. mountains, meadow, jungle), the second video about water worlds were video scenes filmed either over (i.e. waves on a beach, waterfall in the jungle) or under sea level (coral reef and fishes) and the third video about nature were scenes about wild animals (i.e. panda bears, dolphins, birds). A combination of classic, pop, and orchestral music adapted to the video content was used for auditory stimulation. The music was played via noise-cancelling overhead headphones (Beats Studio by Dr. Dre, Beats Electronics, Santa Monica, USA) to cancel out noise in the ICU station.

The Oculus Rift DK2 (Oculus VR LLC, Irvine, USA) is a head mounted display (HMD) with a resolution of 1920 × 1080 pixels and a frame rate of 90 Hz. Measures taken to reduce cybersickness were a powerful computer (graphic card NVIDIA GTX980, Nvidia, Santa Clara, USA and a CPU Intel i7, Intel, Santa Clara, USA) to prevent low framerates and screen flickering54 and positioning participants in horizontal position. To play the videos, the VLC Media Player was used in combination with the Virtual Desktop application developed for the Oculus Rift. To measure gaze behaviour, the SMI Oculus eye tracking system (SensoMotoric Instruments, Teltow, Germany) fitted inside the housing of the Oculus HMD was used. Using image processing, the eye tracking system reconstructs a grayscale image of the eye using six infrared-emitting diodes around each eye. The SMI eye tracker had a sampling frequency of 60 Hz and a system accuracy of 0.5-1 degrees visual angle. To monitor the eyes and physical well-being of the participant, the grayscale images of the eyes were live-streamed to a tablet (iPad Pro, Apple, Cupertino, USA) and monitored by the study investigator. To record whenever a participant fell asleep during stimulation, the eye image stream was stored as video files at a frequency of 30 Hz. Patients which fell asleep were not awakened until the end of the video. Furthermore, all sleep phases were removed before analysing the data.

The respiratory rate and heart frequency were sampled using a five-lead electrocardiography configuration. Using pulse oximetry on a finger, the peripheral capillary oxygen saturation was measured, whereas the mean arterial pressure was measured by a cuff based blood pressure monitor. All vital parameters were stored every second minute and approximated to three measurements in each video (begin, middle, end). The vital parameters were sampled before and during the whole study including recovery phase.

The whole setup was checked by the University Hospital of Berne (Inselspital) and certified to as secure to use in the ICU for healthy participants as well for critically ill patients. All vital parameters were measured with existing and certified devices in the ICU. To prevent feedback into the power supply system and subsequent harm to other ICU devices and patients, a galvanic isolation (Thalheimer Transformers GmbH, Thalheim, Germany) was installed between all net-powered devices. An additional safety measure was the location of the computer 1.5 m away from the participant to guarantee that no current could be transmitted to participants. To prevent the system from dispersing dust and thus putting patients with lowered immune system functioning at risk, all computer fans were protected with a filter. Owing to the direct contact of the HMD and the headphones to the skin of the participants, disposable covers were used for the headphones, while a disinfectable cover made of polyurethane-coated polyester (Obatex bravo, OBA AG, Basel, Switzerland) was tailored for the HMD. The cover and the device were disinfected before and after each participant.

Overall the questionnaire (Table 3) was used to measure how well the VR setup was accepted by the participants and if the stimulation did provoke any discomfort, whereas the psychophysiological and oculomotor data measured the response of the parasympathetic system and the level of attention.

Table 3 Questionnaire to measure immersion, presence, usability and sickness in participants.

To measure cybersickness questions from the Simulator Sickness Questionnaire were used55. Cybersickness should not be confounded with motion sickness and does occur in all people. In the case of cybersickness, the user is immobile (e.g. standing) but experiences self-motion inside the VR scene. Self-motion is the feeling when the observers’ avatar is moving in the virtual world, while the actual body does not. The symptoms of motion and cybersickness are very similar in nature and include headache, pallor, sweating, and vertigo56. The two concepts about immersion (sense of being in the VR world) and presence (how a person reacts to immersion) was assessed by questions out of the Igroup Presence Questionnaire57. The higher the immersion and presence, the less a person can distinguish between the real and the virtual world. Thus, immersion and presence are related to technology and can also be described as how closely the experience reflects the real world. Furthermore, it has recently been shown that the higher the sense of presence is, the lower is the risk of experiencing cybersickness58. Finally, to assess participants opinion, attitude and perception of the VR system questions out of the System Usability Scale59 were used. The questionnaire items had a five-point rating scale between one and five.


First, the participants were instructed and prepared for the study (lying on the bed, fixation of the HMD, headphones, and vital sensors). The preparation phase was followed by a recovery phase of 10 minutes. Before and after the recovery phase, vital parameters were measured once. After the recovery phase, the vital parameters were monitored as two minute medians (median of the sampled vital parameters of the last two minutes) until the end of the study. The stimulation with videos started with a small break between videos after the recovery phase. Finally, after the stimulation participants were asked to answer the questionnaire questions.

Data pre-processing

To obtain data for analysis, the eye tracking raw data was quantified as a collection of fixation points (gaze points). Fixations were defined as eye movements with minimal duration of 100 ms and a maximal dispersion of two degrees61. To account for the small variability of tracking frequency, the gaze points were either up- or down-sampled to a frequency of 60 Hz. Next, image frames in which the participants had their eyes closed were excluded from the analysis. For this, a two-step noise cancelling, followed by canny edge detection (binary image) and a morphological operation was applied. Pupil detection was achieved with a Hough transformation from the pre-processed binary image. The noise cancelling included a Gaussian filter (\(\sigma =0.5\)) and a median filter of size 8 × 8 pixels. By morphological operation after edge detection, objects containing less than 50 pixels were removed. Eye tracking data was removed when participants closed their eyes for more than 20 frames (0.33 s). At the final stage after pre-processing a sliding window based on a length of 60 s (n−30s and n + 30 s) at a step size of one time unit (1/60 s) was used62 for feature extraction (fixation duration [ms/s], number of fixations per second, saccade amplitudes [deg.] and fixation/saccade ratios).

Statistical analysis

To analyse oculomotor and vital data linear mixed effect models (LMEM), fit by maximum likelihood were used instead of classical linear models or repeated measure ANOVA, since data was clustered and contained numeric values. Conditional F-tests were used to test global fixed effects of categorical variables, conditional t-tests were used for fixed effect parameters and likelihood ratio tests were used for random effects.

First, to examine whether the VR stimulation had a relaxing effect (decreasing of vital signs during stimulation), a LMEM (model 1) with vital sign variables as dependent variables, time and video type as fixed effects, a random intercept for each participant as random effect was fitted and tested for a negative fixed effect of time. To examine whether there were inter-individual differences in the effect of time, a random slope of time per participant was added as random effect to the LMEM (model 2) and tested for variance of the random slopes. To further check whether the relaxing effect was higher for participants with higher baseline vital sign measurements, the correlation between the random intercepts and slopes of time was included in the LMEM (model 3). For each dependent variable, the optimal mixed-effect model was chosen based on the significance tests. In a final explorative step, fixed effects for participant-based ratings of the VR stimulation were added to the respective model in order to investigate the influence of participants experience with the VR environment on vital sign variables by using Holm adjusted p-values.

Secondly, we fitted LMEMs with oculomotor data as dependent variables, fixed effects and a random intercept for each participant as random effect. In order to investigate the relevant predictors, effect sizes were used instead of significance tests, as proposed by Nakagawa et al. and Nuthmann et al. since the sample size of the oculomotor data was extremely large. The effect size of a dependent variable was calculated by evaluating the goodness of fit of LMEMs by using the marginal (\({R}_{{LMM}(m)}^{2}\)) and conditional (\({R}_{{LMM}(c)}^{2}\)) R236,37,38. Altogether, we tested eight LMEMs: a zero model without any fixed effects, a full model including all fixed effects (time, target and video type), three individual models for each fixed effect, and three models to quantify the increase of the explained variance by time, target and video type. (Full model without the considered predictor variable). The target was a dummy variable with values zero (not target present) or one (target present). The questionnaire ratings of the VR stimulation were analysed using one-sided t-tests to check whether a significant majority was above the midpoint of the rating scale (2.5) or not.

In the result section only the main findings were discussed, whereas more details about the different LMEMs can be found in the supplementary section. The analysis of the eye tracking and vital data was conducted using MATLAB 2015a63 and R for statistics64.


Videos that were used in the VR stimulation are available online on our website

Data and Code availability

All relevant data and codes supporting the findings are presented in this paper and in the supplementary information. Further data and code of the study are available upon request.


  1. 1.

    Timmers, T. K., Verhofstad, M. H., Moons, K. G., van Beeck, E. F. & Leenen, L. P. Long-term quality of life after surgical intensive care admission. Arch Surg 146, 412–418, (2011).

    Article  PubMed  Google Scholar 

  2. 2.

    Hofhuis, J. G. M. et al. The Impact of Critical Illness on Perceived Health-Related Quality of Life During ICU Treatment, Hospital Stay, and After Hospital Discharge: A Long-term Follow-up Study. Chest 133, 377–385, (2008).

    Article  PubMed  Google Scholar 

  3. 3.

    Hopkins, R. O., Choong, K., Zebuhr, C. A. & Kudchadkar, S. R. Transforming PICU culture to facilitate early rehabilitation. Journal of pediatric intensive care 4, 204–211, (2015).

    Article  PubMed  PubMed Central  Google Scholar 

  4. 4.

    Wergin, R. & Modrykamien, A. Cognitive impairment in ICU survivors: assessment and therapy. Cleveland Clinic journal of medicine 79, 705–712, (2012).

    Article  PubMed  Google Scholar 

  5. 5.

    Wilcox, M. E. et al. Cognitive dysfunction in ICU patients: risk factors, predictors, and rehabilitation interventions. Critical care medicine 41, S81–S98, (2013).

    Article  PubMed  Google Scholar 

  6. 6.

    Brummel, N. et al. Feasibility and safety of early combined cognitive and physical therapy for critically ill medical and surgical patients: the Activity and Cognitive Therapy in ICU (ACT-ICU) trial. Intensive care medicine 40, 370–379, (2014).

    CAS  Article  PubMed  Google Scholar 

  7. 7.

    Pandharipande, P. P. et al. Long-term cognitive impairment after critical illness. New England Journal of Medicine 369, 1306–1316, (2013).

    CAS  Article  PubMed  PubMed Central  Google Scholar 

  8. 8.

    Brummel, N. E. et al. A combined early cognitive and physical rehabilitation program for people who are critically ill: the activity and cognitive therapy in the intensive care unit (ACT-ICU) trial. Physical therapy 92, 1580–1592, (2012).

    Article  PubMed  PubMed Central  Google Scholar 

  9. 9.

    Azouvi, P., Vallat-Azouvi, C. & Belmont, A. Cognitive deficits after traumatic coma. Progress in brain research 177, 89–110, (2009).

    Article  PubMed  Google Scholar 

  10. 10.

    Karnatovskaia, L. V., Johnson, M. M., Benzo, R. P. & Gajic, O. The spectrum of psychocognitive morbidity in the critically ill: a review of the literature and call for improvement. Journal of critical care 30, 130–137, (2015).

    Article  PubMed  Google Scholar 

  11. 11.

    Balas, M. C. et al. Effectiveness and safety of the awakening and breathing coordination, delirium monitoring/management, and early exercise/mobility (ABCDE) bundle. Critical care medicine 42, 1024, (2014).

    Article  PubMed  PubMed Central  Google Scholar 

  12. 12.

    Jackson, J. et al. Cognitive and physical rehabilitation of ICU survivors: results of the RETURN randomized, controlled pilot investigation. Critical care medicine 40, 1088, (2012).

    Article  PubMed  PubMed Central  Google Scholar 

  13. 13.

    Gordon, S. M., Jackson, J. C., Ely, E. W., Burger, C. & Hopkins, R. O. Clinical identification of cognitive impairment in ICU survivors: insights for intensivists. Intensive care medicine 30, 1997–2008, (2004).

    Article  PubMed  Google Scholar 

  14. 14.

    Lindenmuth, J. E., Breu, C. S. & Malooley, J. A. Sensory overload. AJN The American Journal of Nursing 80, 1456, (1980).

    CAS  PubMed  Google Scholar 

  15. 15.

    Stuart, G. W. Principles and Practice of Psychiatric Nursing-E-Book. (Elsevier Health Sciences, 2014).

  16. 16.

    Foreman, M. D., Fulmer, T. T. & Milisen, K. Critical care nursing of older adults: best practices. (Springer Publishing Company, 2009).

  17. 17.

    Getchel, N. Advanced Concepts in Clinical Nursing. AJN The American Journal of Nursing 72, 362–363, (1972).

    Google Scholar 

  18. 18.

    Hu, R.-F., Hegadoren, K. M., Wang, X.-Y. & Jiang, X.-Y. An investigation of light and sound levels on intensive care units in China. Australian Critical Care 29, 62–67, (2016).

    Article  PubMed  Google Scholar 

  19. 19.

    Tainter, C. R. et al. Noise levels in surgical ICUs are consistently above recommended standards. Critical care medicine 44, 147–152, (2016).

    Article  PubMed  Google Scholar 

  20. 20.

    Leach, J. Psychological factors in exceptional, extreme and torturous environments. Extreme physiology & medicine 5, 1, (2016).

    Article  Google Scholar 

  21. 21.

    Turon, M., Fernandez-Gonzalo, S., Gomez-Simon, V., Blanch, L. & Jodar, M. Cognitive stimulation in ICU patients: should we pay more attention? Critical Care 17, 1, (2013).

    Article  Google Scholar 

  22. 22.

    Chen, Y. & Peng, N. A design of virtual reality based coma stimulation program system. 2006 7th International Conference on Computer-Aided Industrial Design and Conceptual Design, 1–6, (2006).

  23. 23.

    Morie, J. F. & Chance, E. Extending the Reach of Health Care for Obesity and Diabetes Using Virtual Worlds. Journal of Diabetes Science and Technology 5, 272–276, (2011).

    Article  PubMed  PubMed Central  Google Scholar 

  24. 24.

    Triberti, S. & Chirico, A. Healthy Avatars, Healthy People: Care Engagement Through. Transformative Healthcare Practice through Patient Engagement, 247, (2016).

  25. 25.

    Székely, G. & Satava, R. M. Virtual reality in medicine. BMJ: British Medical Journal 319, 1305, (1999).

    Article  PubMed  PubMed Central  Google Scholar 

  26. 26.

    Riener, R. & Harders, M. Virtual Reality in Medicine (Springer London, 2012).

  27. 27.

    Jäger, M., Gruber, N., Müri, R., Mosimann, U. P. & Nef, T. Manipulations to reduce simulator-related transient adverse health effects during simulated driving. Medical & Biological Engineering & Computing 52, 601–610, (2014).

    Article  Google Scholar 

  28. 28.

    Pickenhain, L. A neuroscientist’s view on theories of complex movement behaviour. Advances in psychology 50, 463–487, (1988).

    Article  Google Scholar 

  29. 29.

    Valtchanov, D., Barton, K. R. & Ellard, C. Restorative effects of virtual nature settings. Cyberpsychology, Behavior, and Social Networking 13, 503–512, (2010).

    Article  Google Scholar 

  30. 30.

    Gamble, K. R., Howard, J. H. Jr & Howard, D. V. Not just scenery: viewing nature pictures improves executive attention in older adults. Experimental aging research 40, 513–530, (2014).

    Article  PubMed  PubMed Central  Google Scholar 

  31. 31.

    Berman, M. G., Jonides, J. & Kaplan, S. The cognitive benefits of interacting with nature. Psychological science 19, 1207–1212, (2008).

    Article  PubMed  Google Scholar 

  32. 32.

    Ulrich, R. S. et al. Stress recovery during exposure to natural and urban environments. Journal of environmental psychology 11, 201–230, (1991).

    Article  Google Scholar 

  33. 33.

    Berto, R. The role of nature in coping with psycho-physiological stress: a literature review on restorativeness. Behavioral sciences 4, 394–409, (2014).

    Article  PubMed  PubMed Central  Google Scholar 

  34. 34.

    Ohly, H. et al. Attention Restoration Theory: A systematic review of the attention restoration potential of exposure to natural environments. Journal of Toxicology and Environmental Health, Part B 19, 305–343, (2016).

    CAS  Article  Google Scholar 

  35. 35.

    McEwen, B. S. Physiology and Neurobiology of Stress and Adaptation: Central Role of the Brain. Physiological Reviews 87, 873–904, (2007).

    Article  PubMed  Google Scholar 

  36. 36.

    Nuthmann, A. Fixation durations in scene viewing: Modeling the effects of local image features, oculomotor parameters, and task. Psychonomic Bulletin & Review 1–23, (2016).

  37. 37.

    Nakagawa, S. & Schielzeth, H. A general and simple method for obtaining R2 from generalized linear mixed-effects models. Methods in Ecology and Evolution 4, 133–142, (2013).

    Article  Google Scholar 

  38. 38.

    Johnson, P. C. Extension of Nakagawa & Schielzeth’s R2GLMM to random slopes models. Methods in Ecology and Evolution 5, 944–946, (2014).

    Article  PubMed  PubMed Central  Google Scholar 

  39. 39.

    Chirico, A. et al. Effectiveness of immersive videos in inducing awe: An Experimental Study. Scientific Reports 7, (2017).

  40. 40.

    Joye, Y. & Bolderdijk, J. W. An exploratory study into the effects of extraordinary nature on emotions, mood, and prosociality. Frontiers in Psychology 5, 1577, (2014).

    PubMed  Google Scholar 

  41. 41.

    Stellar, J. E. et al. Positive affect and markers of inflammation: Discrete positive emotions predict lower levels of inflammatory cytokines. Emotion 15, 129, (2015).

    Article  PubMed  Google Scholar 

  42. 42.

    Alvarsson, J. J., Wiens, S. & Nilsson, M. E. Stress recovery during exposure to nature sound and environmental noise. International journal of environmental research and public health 7, 1036–1046, (2010).

    Article  PubMed  PubMed Central  Google Scholar 

  43. 43.

    Annerstedt, M. et al. Inducing physiological stress recovery with sounds of nature in a virtual reality forest—Results from a pilot study. Physiology & behavior 118, 240–250, (2013).

    CAS  Article  Google Scholar 

  44. 44.

    Di Stasi, L. L. et al. Microsaccade and drift dynamics reflect mental fatigue. European Journal of Neuroscience 38, 2389–2398, (2013).

    Article  PubMed  Google Scholar 

  45. 45.

    Boksem, M. A., Meijman, T. F. & Lorist, M. M. Effects of mental fatigue on attention: an ERP study. Cognitive brain research 25, 107–116, (2005).

    Article  PubMed  Google Scholar 

  46. 46.

    Goldberg, J. H. & Kotval, X. P. Computer interface evaluation using eye movements: methods and constructs. International Journal of Industrial Ergonomics 24, 631–645, (1999).

    Article  Google Scholar 

  47. 47.

    Osman, Z. M., Dupire, J., Topol, A. & Cubaud, P. Non intrusive measures for determining the minimum field of view for user search task in 3D virtual environments. The Seventh International Conference on Advances in Computer-Human Interactions pp. 208–213 (2014).

  48. 48.

    Dorr, M., Martinetz, T., Gegenfurtner, K. R. & Barth, E. Variability of eye movements when viewing dynamic natural scenes. Journal of vision 10, 28–28, (2010).

    Article  PubMed  Google Scholar 

  49. 49.

    Hoffman, H. G. et al. Feasibility of articulated arm mounted Oculus Rift Virtual Reality goggles for adjunctive pain control during occupational therapy in pediatric burn patients. Cyberpsychology, Behavior, and Social Networking 17, 397–401, (2014).

    Article  Google Scholar 

  50. 50.

    Larson, E. B., Feigon, M., Gagliardo, P. & Dvorkin, A. Y. Virtual reality and cognitive rehabilitation: A review of current outcome research. NeuroRehabilitation 34, 759–772, (2014).

    PubMed  Google Scholar 

  51. 51.

    Dascal, J. et al. Virtual reality and medical inpatients: A systematic review of randomized, controlled trials. Innovations in clinical neuroscience 14, 14 (2017).

    PubMed  PubMed Central  Google Scholar 

  52. 52.

    Rose, F. D., Brooks, B. M. & Rizzo, A. A. Virtual reality in brain damage rehabilitation: review. CyberPsychology & Behavior 8, 241–262, (2005).

    Article  Google Scholar 

  53. 53.

    Ely, E. W. et al. Monitoring sedation status over time in ICU patients: reliability and validity of the Richmond Agitation-Sedation Scale (RASS). Jama 289, 2983–2991, (2003).

    Article  PubMed  Google Scholar 

  54. 54.

    Lee, E. C., Heo, H. & Park, K. R. The comparative measurements of eyestrain caused by 2D and 3D displays. Consumer Electronics, IEEE Transactions on 56, 1677–1683, (2010).

    Article  Google Scholar 

  55. 55.

    Kennedy, R. S., Lane, N. E., Berbaum, K. S. & Lilienthal, M. G. Simulator sickness questionnaire: An enhanced method for quantifying simulator sickness. The international journal of aviation psychology 3, 203–220, (1993).

    Article  Google Scholar 

  56. 56.

    LaViola, J. J. Jr. A discussion of cybersickness in virtual environments. ACM SIGCHI Bulletin 32, 47–56, (2000).

    Article  Google Scholar 

  57. 57.

    Schubert, T., Friedmann, F. & Regenbrecht, H. In Visual representations and interpretations 269–278 (Springer, 1999).

  58. 58.

    Busscher, B., De Vliegher, D., Ling, Y. & Brinkman, W.-P. Physiological measures and self-report to evaluate neutral virtual reality worlds. Journal of Cybertherapy and Rehabilitation 4 (1) 2011, 1–18 (Preliminary version) (2011).

  59. 59.

    Brooke, J. SUS-A quick and dirty usability scale. Usability evaluation in industry 189, 4–7 (1996).

    Google Scholar 

  60. 60.

    Witmer, B. G. & Singer, M. J. Measuring presence in virtual environments: A presence questionnaire. Presence: Teleoperators and virtual environments 7, 225–240, (1998).

    Article  Google Scholar 

  61. 61.

    Salvucci, D. D. & Liu, A. The time course of a lane change: Driver control and eye-movement behavior. Transportation Research Part F: Traffic Psychology and Behaviour 5, 123–132, (2002).

    Article  Google Scholar 

  62. 62.

    Bulling, A., Ward, J. A., Gellersen, H. & Troster, G. Eye Movement Analysis for Activity Recognition Using Electrooculography. IEEE Transactions on Pattern Analysis and Machine Intelligence 33, 741–753, (2011).

    Article  PubMed  Google Scholar 

  63. 63.

    MATLAB v. (R2015a) (The MathWorks Inc., Natick, Massachusetts, 2015).

  64. 64.

    A Language and Environment for Statistical Computing v. 3.3.1 (R Foundation for Statistical Computing, Vienna, Austria, 2015).

Download references


The authors would like to thank the clinical nursing and doctor team at the Department of Intensive Care Medicine at University Hospital Bern (Inselspital) for their engagement during the project and for helping test all participants. In addition, we would like to thank all participants who participated in the studies. We would also like to thank all co-workers at the Gerontechnology and Rehabilitation Research Group at University of Bern for advices and support during the study. The study was founded by the Department of Intensive Care Medicine at University Hospital Bern (Inselspital) and the Gerontechnology and Rehabilitation Research Group at ARTORG Center for biomedical engineering at University of Bern.

Author information




S.G., M.M.J., T.N. S.M.J., R.M., P.U. designed the study. S.G. developed the setup and measured together with M.M.J. participants. S.G., P.W. and A.C. analysed the data. S.G., A.C., T.N. wrote the manuscript. All authors approved the final manuscript.

Corresponding author

Correspondence to Tobias Nef.

Ethics declarations

Competing Interests

The authors declare that they have no competing interests.

Additional information

Publisher's note: Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.

Electronic supplementary material

Rights and permissions

Open Access This article is licensed under a Creative Commons Attribution 4.0 International License, which permits use, sharing, adaptation, distribution and reproduction in any medium or format, as long as you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons license, and indicate if changes were made. The images or other third party material in this article are included in the article’s Creative Commons license, unless indicated otherwise in a credit line to the material. If material is not included in the article’s Creative Commons license and your intended use is not permitted by statutory regulation or exceeds the permitted use, you will need to obtain permission directly from the copyright holder. To view a copy of this license, visit

Reprints and Permissions

About this article

Verify currency and authenticity via CrossMark

Cite this article

Gerber, S.M., Jeitziner, MM., Wyss, P. et al. Visuo-acoustic stimulation that helps you to relax: A virtual reality setup for patients in the intensive care unit. Sci Rep 7, 13228 (2017).

Download citation

Further reading


By submitting a comment you agree to abide by our Terms and Community Guidelines. If you find something abusive or that does not comply with our terms or guidelines please flag it as inappropriate.


Nature Briefing

Sign up for the Nature Briefing newsletter — what matters in science, free to your inbox daily.

Get the most important science stories of the day, free in your inbox. Sign up for Nature Briefing