Investigating the Impact of Mental Stress on Electrocardiological Signals through the Use of Virtual Reality

Abstract

This article presents a new 3D extreme game for virtual reality (VR), which is used to evaluate the impact of generated mental stress on the cardiological state of the playing individuals. The game was developed using Java 3D and Blender. Generated stress is investigated by recording electrocardiograms for 20 min and determining heart rate variability (HRV) parameters in the time and frequency domains and by non-linear visual and quantitative analysis methods, such as the Rescaled Range (R/S) method, Poincarè plot, Recurrence plot, Approximate (ApEn), and Sample Entropy (SampEn). The data of 19 volunteers were analyzed before and immediately after the game, and a comparative analysis was made of two types of VR: immersive and non-immersive. The results show that the application of immersive VR generates higher mental stress levels than non-immersive VR, but in both cases, HRV changes (decreases), but more significantly in immersive VR. The results of this research can provide useful information about the functioning of the autonomic nervous system, which regulates the reactions of the human body during mental stress, to help in the early detection of potential health problems.

1. Introduction

Stress is one of the leading health problems of modern society, which can occur as a reaction to various external or internal factors [1,2,3,4]. The definition that the World Health Organization gives for stress covers a wide range of factors that can cause stress, as well as various manifestations of stress on a person’s body and psyche [5]. Research on the impact of stress on human health is extensive and covers multiple aspects, including its causes, mechanisms, and consequences for human health [6,7,8,9]. Depending on the causes of stress, it can be of two main types: physical and mental. Physical stress is caused by physical factors, such as pain, trauma, illness, overexertion, and others. Mental stress is a complex phenomenon that includes emotional, cognitive, and behavioural components. Understanding the impact of stress and effective coping strategies are vital to improving people’s health [10,11,12]. Stress is measured in two ways: by taking biometric data and conducting psychological tests [13,14]. Biometric data measurement includes measuring the heart rate, skin conductance, and cortisol levels, and psychological testing includes surveys and questions to assess the participants’ subjective senses of stress.

Virtual reality (VR) can be a tool for generating and managing stress because it provides the opportunity to create controlled and realistic environments that can be used to induce stressful situations [15,16]. Virtual reality can be both immersive (where the user is completely immersed in the virtual environment) and non-immersive (where the environment is presented on a screen or through other devices) to conduct the necessary research [17,18,19,20,21,22]. In immersive VR, situations can be created that realistically simulate stressful scenarios, where users can be placed in high-stress situations, such as public exams, high-risk scenes, or emergencies [23,24]. In non-immersive VR, scientists can use virtual environments that simulate stressful situations without being fully immersed in them. Such research can also provide information about the impact of stress on human health. Therefore, VR provides opportunities to study subjects in controlled and realistic stressful situations, which can help researchers better understand the interaction between stress and human health. Research on the stress induced by participating in 3D extreme VR games is an important area that can show the relationship between the impact of VR gaming and the psychological responses of players [25]. These games are usually characterized by strong visual and sound effects and fast-paced and exciting scenarios that can provoke intense emotions and physiological reactions, resulting in changes in heart rate, blood pressure, etc. These studies can contribute to a better understanding of the interaction between humans and virtual reality technology and help to develop more balanced and safe experiences in the virtual world.

Reference [26] presents a study involving VR simulations used in nursing education to simulate stressful medical scenarios. As a result of the generated stress, the heart rate variability (HRV) of the study subjects changed, which was found by determining high-frequency (HF) and low-frequency/high-frequency (LF/HF) parameters. The results show that the VR environment is an effective tool for inducing stress in a controlled manner.

The authors of [27] investigated the influence of two types of serious cognitive training games on the autonomic nervous system (ANS) of adults, observing the test subjects’ responses through an HRV analysis. The results of the experiments show a difference in HRV between the two types of games, from which it can be concluded that the type of game significantly influences the ANS response. According to the authors, this research can be used in designing learning and mood management games for adults.

Chronic stress can have severe consequences for a person’s physical and mental health, including an increased risk of cardiovascular disease, diabetes, and depression and anxiety disorders [6,9]. Searching for appropriate markers to register stress (mental or physiological) is challenging for scientific researchers, as there is currently no universally recognized standard for its assessment. In recent years, data have been published showing that both mental and physiological stress can be studied by applying HRV, which measures changes in the time between successive heartbeats [28,29], and this can be used as a tool to determine the effect of stress on people’s cardiac health.

The present article aims to study the impact of stress on volunteers’ cardiological conditions by applying a complex and multidisciplinary approach. The proposed innovative approach involves creating an extreme 3D game with different VR scenarios (immersive and non-immersive) as a way to generate negative mental stress. Stress is measured by recording the participants’ electrocardiological (ECG) signals, based on which HRV is determined before and immediately after the game. The registered ECG signals are analyzed by creating software that applied two mathematical methods: linear and non-linear. Linear methods include an analysis in the time and frequency domains, and regarding the non-linear methods for visual and quantitative analyses of HRV, the following are applied: the Rescaled Range (R/S) method, Poincarè plot, recurrence plot, Approximate Entropy (ApEn), and Sample Entropy (SampEn).

2. Materials and Methods

2.1. Types of Virtual Reality

Virtual reality is a computer-generated simulation of three-dimensional environments, which aims to create a sense of immersion in that simulated environment by allowing users to move, look at, and interact with objects in a way that mimics the physical world. According to the level of immersion, virtual reality can be classified into the following two categories:

• Immersive VR: This form of virtual reality offers complete user immersion in the virtual world. This is achieved through special devices such as helmets that cover the user’s entire field of vision and often include sound sensors to enhance the sense of realism. As a result, the user can experience being in a virtual environment and interact with it in a way that resembles the real world. Immersive virtual reality provides an opportunity to simulate different scenarios and experience situations that may be difficult or dangerous to experience in the real world [28,29].
• Non-immersive VR: This type of virtual reality provides an interactive experience without special VR helmets or equipment for total immersion. This technology uses anaglyph glasses, which are popular for watching 3D movies and TV at home, offering a convenient and affordable way to experience the three-dimensional experience without the need for complex electronics or additional power supply. The screen that displays the 3D content sends two different pictures or video frames, one to the left eye and one to the right eye, and these images are polarized differently. Anaglyph glasses consist of two filters: one in red (often with colour filtration) and one in blue. This type of glasses allows the viewer to see different images with both eyes, which merge into a three-dimensional perception of the scene. The principle of operation of 3D glasses and helmets is based on human eyes seeing each image from a different angle. Thus, by using the filtering lenses, the images are split and intersected, creating the illusion of three-dimensionality [30,31].

2.2. The Relationship between the Autonomic Nervous System, Heart Rate Variability, and Stress

The autonomic nervous system, heart rate variability, and stress are interrelated and significantly impact humans’ cardiac health [32,33,34]. The ANS is part of the peripheral nervous system that controls the unconscious functions of the human body, including the heart rate, blood pressure, breathing, and more. HRV represents the fluctuations in the intervals between successive heartbeats and is an indicator of the balance and functioning of the ANS, which consists of two main parts: the sympathetic and parasympathetic nervous systems [35]. The sympathetic nervous system is vital for the body’s rapid adaptation to stressful situations and plays a crucial role in maintaining homeostasis under physical and emotional stress conditions. Its activation leads to complex physiological reactions that prepare the human body for quick and energetic actions. The parasympathetic nervous system (PNS) is the opposite of the sympathetic nervous system, and its leading role is to maintain and restore the body to a state of rest and to support the recovery process. This part of the ANS is associated with the “rest and digest” reaction. The activation of the SNS reduces HRV, which is reflected in an increase in the heart rate and a decrease in the recovery time between heartbeats. This behaviour is typical for a situation of stress or physical activity. Activation of the PNS increases HRV by reducing the heart rate and increasing the recovery time between successive heartbeats. This is characteristic of states of rest and relaxation. A high HRV indicates good balance and adaptability of the ANS. At the same time, a low HRV can be a sign of imbalance and is associated with an increased risk of cardiovascular disease and other health problems. The study of HRV provides valuable information about the state of the ANS and the general health of the individual, and this can be useful in a variety of clinical and scientific applications.

2.3. Methods for Heart Rate Variability Analysis

According to the standard introduced in 1996 by the European Society of Cardiology and the North American Society of Electrophysiology, the methods for a mathematical analysis of HRV fall into two main classes: linear (analysis in the time and frequency domains) and non-linear [32]. Linear methods are standardized, as the reference values of the parameters are known, while non-linear methods for HRV analysis are under research.

2.3.1. Time-Domain Analysis

A time-domain analysis of the intervals between heartbeats (RR intervals) refers to the study of the intervals between successive R peaks in the ECG. This analysis can provide information about the cardiac health of the study subject [32,36,37,38]. Some parameters used for time-domain statistical analysis are as follows:

• MeanRR: This is the average time between successive R peaks in the ECG. This parameter can help estimate the average heart rate and its changes over time.
• SDNN (Standard Deviation of NN intervals)—This parameter shows the variation in normal RR intervals (NN intervals) compared to their mean value. A more significant standard deviation may indicate higher heart rate variability.
• RMSSD (Root Mean Square of Successive Differences): This parameter represents the square root of the arithmetic mean of squares of the differences between successive normal RR intervals. Higher values of RMSSD usually indicate increased heart activity.
• pNN50%: This is the percentage of all normal RR intervals that differ by more than 50 ms.

The geometric analysis of the HRV consists of constructing histograms composed of normal RR intervals, and the following two parameters determine their triangular approximation [36]:

• TINN (triangular interpolation of NN intervals): This parameter shows the distribution of intervals that is approximated to a triangle.
• HRVti: This parameter measures the number of variations in the time series of the normal RR intervals.

2.3.2. Frequency Domain Analysis

The frequency domain analysis parameters are based on spectral analysis, which gives the distribution of the following three spectral components: Very Low Frequency (VLF), Low Frequency (LF) and High Frequency (HF). The present study uses only the LF and HF parameters and their ratio [32,36]. The absolute power values in the frequency do-main give an idea of the overall level of ANS activity, while the relative values (percentage of total power) help to understand the contribution of different ANS components. The LF/HF ratio is used as an indicator of the sympathetic–parasympathetic balance. A high LF/HF value may indicate dominant sympathetic activity, often associated with stress or physical activity, while a low value suggests dominant parasympathetic activity. This type of analysis provides detailed information about the dynamics of the autonomic nervous system and its response to various physiological and psychological conditions. It can be used for diagnosis and monitoring in medicine, psychology, and sports science.

2.3.3. Non-Linear Methods

Non-linear HRV analysis methods provide additional information about the complexity and dynamics of cardiac activity that cannot be captured by linear methods alone. These methods are not standardized, but according to the introduced standard [32], the study of the applicability of non-linear methods for the analysis of HRV is an important priority area that opens new perspectives for their future use.

• Rescaled Range (R/S) Method

The Rescaled Range (R/S) method uses concepts from fractal geometry to analyze processes that exhibit self-similarity at different time scales. The RR interval series determined from the ECG signals can be considered as time series that exhibit the self-similarity property at different scales and provide information about the complexity and structure of these time series by determining the value of the Hurst parameter. With this parameter, the degree of self-similarity in the studied signals can be determined [39,40,41]. For time series that exhibit long-run correlation, the Hurst parameter is in the interval (0, 1), with larger values indicating a stronger long-run correlation and smaller values indicating a weaker correlation. Lower values of the Hurst parameter may also indicate a loss of complexity in the RR time intervals showing the dynamics of cardiac activity, which may be related to stressful conditions.

The interpretation of the values of the Hurst parameter (H) in terms of long-term correlations and the self-similarity of the studied RR time series is as follows:

• If H = 0.5, there are no long-term correlations;
• If H < 0.5, this indicates anti-persistence and shows that the high values tend to be followed by low values and vice versa;
• If H > 0.5, this indicates persistence and shows a long-term relationship where high values tend to be followed by highs and low values by lows.

The analysis of the RR interval series by determining the Hurst parameter (R/S method) will be applied in the present work to investigate the impact of stress on HRV, and it is expected to provide information on the heart rate response to stressful events.

• Poincaré Plot

The Poincaré plot (PP) is a method used to analyze heart rate variability, allowing for the visualization and quantification of the autonomic regulation of cardiac activity [42,43,44]. In the graphical representation of the RR time series, each point on the graph represents a relationship between the current value and the next value in the series. The shape of the graph constructed by the Poincaré plot can provide information about the characteristics of the time series and about the autonomic regulation of cardiac activity. The graph constructed with this method can be asymmetric or elliptical. If the shape is elliptical, it means the cardiac activity is normal, while if the ellipse is elongated along the line of identity y = x, it may indicate a predominance of sympathetic activity and reduced short-term variability. A round shape can be a sign of a healthy state with a good balance between sympathetic and parasympathetic activity, but if the points on the graph form irregular clusters, this may indicate the presence of arrhythmias or other pathological conditions. The quantitative parameters of the method are SD1, SD2, and SD1/SD2. The SD1 parameter represents the scatter of the points perpendicular to the line of identity and reflects short-term variability and parasympathetic activity. The SD2 parameter represents the scatter of points along the line of identity and reflects long-term variability and sympathetic activity. The relationship between SD1 and SD2 is a key indicator of the autonomic regulation of the heart rate and provides information about the balance between short-term and long-term HRV. The Poincaré plot can be used to study the effect of stress on the HRV.

• Recurrence Plot and Recurrence Quantification Analysis

The recurrence plot (RP) [45] and Recurrence Quantification Analysis (RQA) [46] are tools for analyzing heart rate variability and stress. They provide both visual and quantitative assessments of the dynamics of cardiac activity, which can help in the understanding of the autonomic regulation of the heart, the identification of pathological conditions, and the assessment of stress. A recurrence plot is a graphical method that visualizes moments of recurrence (recurrence) in a time series. It is a two-dimensional matrix in which the points represent times when the state of the system is close to a previous state. The main components of the graphs constructed with the method are diagonal, vertical, horizontal lines, and single points. Long diagonal lines indicate the periodicity and predictability of the studied signal, while short diagonal lines reflect short-term correlations and periods of predictability. Vertical and horizontal lines indicate stagnation or stable conditions in the system, and long lines of this type may indicate persistent or periodic states of stress or other physiological conditions. The single points without connecting lines indicate randomness or noise in the system. Some of the parameters for quantitative analysis are as follows:

• Recurrence Rate or Recurrence Count (REC%): The percentage of recurrent points in RP. A high REC may indicate high heart rate variability and good autonomic regulation.
• Determinism (DET%): The percentage of the recurring points that are part of the diagonal lines. A high DET indicates high predictability and regularity in the heart rhythm.
• Entropy (ENTR): A measure of the complexity of diagonal lines. A high ENTR value indicates complex and unpredictable dynamics of the studied time series.

• Entropy Methods

Entropy methods for the analysis of heart rate variability provide information on the complexity and unpredictability of the heart rhythm and can be useful for the assessment of the autonomic nervous system and for the diagnosis of various physiological and pathological conditions [47]. A typical representative of this group is ApEn, which measures the regularity and unpredictability of a time series. Smaller ApEn values indicate higher regularity and predictability, while larger values indicate higher complexity and unpredictability. Another representative of this group is SampEn, which is similar to ApEn. SampEn is a more accurate and robust time series complexity estimation method than ApEn, especially for short data. It provides a better estimate of heart rate dynamics, making it the preferred method for heart rate variability analysis. In this method, smaller values of SampEn also indicate higher regularity, while larger values indicate higher complexity and low predictability of fluctuations.

2.4. Statistical Analysis

A statistical analysis of RR intervals, through the use of a t-test, can be used to compare different groups and to assess statistical differences in the heart rate time series. The test result gives the statistical significance of the difference between the studied groups by determining the value of the parameter p. If the p-value is less than the set level of significance (p < 0.05), this indicates that the studied parameter has statistical significance, which gives a reason to reject the null hypothesis that there is no difference between the groups. If the p-value is greater than the significance level, then there is no statistically significant difference between the study groups.

2.5. Data

The data used for the research in this article was recorded with a Dynamic ECG Systems TLC9803 Holter monitor. The duration of the recordings is 20 min with approximately 1000 RR time intervals. Fully immersive virtual reality is realized using a high-level Photontree Pro 3D helmet, which creates a realistic experience without the need to install additional software and is compatible with the Java 3D programming language. Passive technology is implemented using anaglyph glasses with red and blue filters for the left and right eyes. The volunteers participating in the study included 19 men and women aged 20 to 35 years.

3. Results and Discussion

3.1. Three-Dimensional Extreme Game

A 3D extreme game was created using Java 3D and Blender, which was called “asteroid shower” with a duration of 5 to 20 min, and it was realized in virtual reality with immersion by using a virtual helmet and in non-immersive virtual reality by using 3D passive glasses (anaglyph). In the game for the creation of 3D objects based on a polygon mesh, a new algorithm for programming regular polygons, presented in reference [48], is used. Each 3D object is a combination of polygons located in different planes that form a “polygon mesh”. It is known that the starting points in 3D modelling are the geometric figures based on a regular polygon, such as pyramid, cylinder, prism, cone, etc. The author’s algorithm draws a regular polygon by using the dependence of parallel segments, and the segment generation process continues to infinity, thus obtaining a better result in the creation of 3D objects compared to the traditional (trigonometric) way of generating polygons by given numbers of vertices and radius. With the algorithm used, the geometric figures are created by setting the number of vertices and the length of the side, and in this way, better accuracy of the generated objects is achieved. The created polygons can be more easily processed using a 3D modelling programme, and in this game, Blender is used.

The following basic techniques were used in the creation of the game:

• Morphism—the animation of objects from the game, the explosion of asteroids, as well as their scattering are realized by using this technique, which allows the transition from one form of the object to another. A weapon was added to the app’s camera to shoot asteroids.
• Translation, rotation, and spline curves are three ways to drive the app’s camera. The most suitable is the use of a spline curve because it can be used to realize both rectilinear and curvilinear movements.
• Collision detection and sound effects are applied when the asteroids explode.
• The main features of the game are as follows:
• It simulates a stressful situation;
• The game can be played with varying degrees of extremity, resulting in asteroids falling from space to earth at varying speeds and rotations;
• The number of asteroids hit and the duration of the game are displayed.

Figure 1 shows the process of generating an asteroid, which is a game element, and Figure 2 shows two views of the game: a city and a broken asteroid.

3.2. Analysis of HRV

In order to examine the influence of stress generated by the created extreme 3D game, three types of electrocardiograms of the studied 19 volunteers are registered: before play, immediately after play with full-immersion VR using a helmet, as well as with VR without immersion by using anaglyph glasses. The HRV mathematical linear and non-linear analyses are based on RR time intervals determined by the registered ECG signals. The data are combined into three groups: Group 1—before play; Group 2—immediately after playing with full-immersion VR; and Group 3—immediately after playing with VR without immersion. The numerical results of the HRV analysis are presented as mean ± std and are shown in

Table 1. A comparative analysis of the HRV parameters of Group 1: before game; Group 2: after game (immersive VR); and Group 3: after game (non-immersive VR).

Parameter	Before Game Group 1 (n = 19) [Mean ± Std]	Fully Immersive VR Group 2 (n = 19) [Mean ± Std]	Non-Immersive VR Group 3 (n = 19) [Mean ± Std]	Pgr1,Gr2 Value	PGr1,Gr3 Value
Statistical analysis
MeanRR [ms]	725.98 ± 85.11	551.02 ± 61.33	605.33 ± 52.03	<0.0001	<0.0001
SDNN [ms]	87.27 ± 31.26	61.11 ± 19.03	77.01 ± 20.19	0.0036	0.2373
RMSSD [ms]	29.36 ± 10.02	16.67 ± 7.51	21.12 ± 8.99	<0.0001	0.0114
pNN50 [%]	21.51 ± 3.50	11.01 ± 6.83	19.09 ± 2.15	<0.0001	0.0145
Geometrical analysis
TINN [ms]	460.05 ± 101.71	230.21 ± 75.34	412.10 ± 90.06	<0.0001	0.1327
HRVti [-]	15.97 ± 6.71	10.02 ± 3.23	13.21 ± 3.08	0.0013	0.1119
Frequency analysis
LF [n.u]	58.06 ± 3.19	79.11 ± 11.05	63.76 ± 4.12	<0.0001	<0.0001
HF [n.u]	31.41 ± 1.89	27.01 ± 9.05	29.11 ± 2.03	<0.0001	0.0009
LF/HF [-]	1.80 ± 0.19	2.91 ± 0.34	2.16 ± 0.17	<0.0001	<0.0001
Non-linear analysis
Hurst [-] (R/S)	0.956 ± 0.05	0.753 ± 0.10	0.811 ± 0.09	<0.0001	<0.0001
SD1 [ms] (PP)	29.795 ± 7.34	11.228 ± 4.12	18.228 ± 5.34	<0.0001	<0.0001
SD2 [ms] (PP)	61.987 ± 9.13	30.912 ± 10.39	40.781 ± 9.210	<0.0001	<0.0001
DET [%] (RP)	81.013 ± 4.19	53.711 ± 3.92	65.019 ± 10.421	<0.0001	<0.0001
REC [%] (RP)	26.011 ± 3.65	11.721 ± 0.231	17.032 ± 9.023	<0.0001	0.0003
ENTR [-] (RP)	3.311 ± 0.121	4.657 ± 0.891	4.120 ± 0.095	<0.0001	<0.0001
ApEn [-]	0.913 ± 0.128	0.541 ± 0.012	0.783 ± 0.093	<0.0001	0.0194
SampEn [-]	0.896 ± 0.216	0.497 ± 0.109	0.801 ± 0.186	<0.0001	0.1550

, and the graphics results are shown in Figure 3, Figure 4, Figure 5 and Figure 6.

3.2.1. Linear Analysis of HRV

Based on the obtained experimental results shown in Table 1 and Figure 3, the following findings can be made regarding HRV analysis in the time and frequency domains:

• The values of statistical parameters in the time domain, such as MeanRR, SDNN, RMSSD, and pNN50%, decreased in both groups, reporting the stress (Group 2 and Group 3), indicating a dominance of the sympathetic nervous system and reduced parasympathetic activity. Group 1, showing the pregame data, had higher values for these parameters, and therefore, their HRV was higher, which is an indicator of a good balance between sympathetic and parasympathetic activity. From the obtained results, it follows that the statistical parameters are sensitive to the changes caused by the generated stress with the proposed 3D extreme game, and the game can be used to monitor and evaluate the impact of stress on the HRV and ANS of the human organism. Similar results were reported in [36].
• Geometric parameters, such as TINN and HRVti, also provide information about the autonomic nervous system and its response to stress. These parameters are based on the representation of intervals between heartbeats (RR intervals). Higher values of these two parameters at rest indicate greater heart rate variability and better autonomic nervous system function. A decrease in their values is an indicator of the presence of stress.
• The histograms shown in Figure 3 can be useful visual tools for studying HRV when assessing the body’s response to stressful conditions. The histograms show the distribution of the RR intervals, with the duration of the individual cardiac intervals on the horizontal axis and their number on the vertical axis. Figure 3A shows an individual’s pregame (resting) histogram, which consists of multiple RR intervals of varying length. A wide histogram with a symmetrical distribution of the individual’s pregame intervals suggests high heart rate variability, which is associated with a good balance between the sympathetic and parasympathetic nervous systems. The histograms shown in Figure 3B,C were obtained after play with immersive VR and with non-immersive VR. These histograms are narrow with a highly concentrated distribution around a certain value of the RR intervals (0.5–0.55 s), which are shifted to the left compared to the pregame histogram. This behaviour is due to the fact that due to the generated stress, the RR intervals have become more uniform and monotonous with small fluctuations, which indicate a low HRV and are signs of the dominance of sympathetic activity.
• A frequency analysis of HRV provides information about different components of the autonomic nervous system by analyzing fluctuations in RR intervals at different frequency ranges. This approach was applied to evaluate the effect of stress, resulting from the 3D game, on cardiac activity through the parameters LF, HF, and LF/HF, which reflect both sympathetic activity and parasympathetic activity. A frequency analysis of cardiac recordings taken immediately after the game showed that in both Group 2 and Group 3, individuals had higher LF (n.u.) values, while the HF (n.u.) values were lower compared to those of the group before the game. The increased value of LF/HF in these two groups indicates a dominance of sympathetic activity over parasympathetic activity. Similar results were reported in [36]. This behaviour is again due to the stress generated during the extreme 3D game, resulting in an increased heart rate and decreased HRV.

3.2.2. Non-Linear Analysis of HRV

Based on the obtained results shown in Table 1 and Figure 4, Figure 5 and Figure 6, the following conclusions can be drawn regarding the non-linear analysis of the HRV of the three studied groups:

• The determination of the Hurst exponent using the R/S method plays an important role in the study of signals (RR time series) with fractal characteristics, i.e., the value of this parameter is between 0.5 and 1.0. The value of this parameter decreases with stress (Group 2 and Group 3) generated during the game, which is due to the reduction in the fractal complexity of the signal. Figure 4 shows the graphic results when determining the Hurst exponent values for the three studied groups. Similar results were reported in [49] when evaluating the HRV of patients with stress-induced cardiomyopathy, and it was found that the Hurst parameter value in the control group was greater compared to that in the cardiomyopathy group. This is due to the fact that in healthy individuals, as well as in those in a state of rest, the value of this parameter is higher than those with cardiovascular disease, as well as those in a stressful situation. This shows a higher degree of persistence as well as a higher HRV. High values of the Hurst exponent are usually associated with long-term correlation and stability in the heart rate, while low values show greater chaos and less predictability.
• A visual analysis of RR time intervals with a Poincarè plot (Figure 5) can reveal important features of HRV and be used for autonomic nervous system monitoring. At rest (Figure 5A), the points in the graph are more concentrated around the line of identity (the main diagonal), and the graph has a comet-like appearance with a pointed bottom, indicating that HRV is higher and has a greater balance in the autonomic nervous system. Under stress (Group 1 and Group 2), there is less scatter of the points in the graph (Figure 5B,C), which are more compact (compressed) and have lower values for the RR intervals, which is more noticeable in the stress generated during full-immersion gameplay. The obtained results show that under stress, there is lower heart rate variability, and in this condition, the sympathetic nervous system is activated. Similar results were reported in a publication [50]. A quantitative evaluation with Poincarè plot is performed with the parameters SD1 and SD2, and in a state of rest, they have higher values. Under stress, the values of these two parameters decrease, which is due to the increased influence of the sympathetic nervous system, as a result of which HRV decreases.
• A recurrence plot is a visual tool for a time series analysis that shows how system states repeat over time. The method constructs a two-dimensional matrix in which each point indicates whether two states of the system are close to each other. Figure 6 shows the graphs of an individual at rest (pre-game) immediately after a game with VR with and without immersion. At rest, the RP has a more ordered and structured matrix, indicating that the system states repeat regularly and with high correlation. This is an indication of stability and regularity in the data, resulting in higher HRV. Under stress, the RP exhibits a more chaotic and unstructured matrix. The points are scattered and do not form distinct clusters, indicating a reduced correlation and increased randomness in the data. Similar results were reported in [51]. The quantitative evaluation with the method was performed with the parameters REC%, DET%, and ENTR. The REC% parameter values measured using the RP may vary depending on the data being analyzed. In a resting state, the system is more stable and returns more often to previous states, and these result in higher REC% values due to the balance between the sympathetic and parasympathetic nervous systems. Under stress, the system is more dynamic and chaotic, resulting in lower REC% values and lower HRV. The DET% parameter values are also lower under stress, indicating that the system is more unpredictable and chaotic. The ENTR parameter also provides an estimate of the degree of chaoticity and variability in the system, providing information on how the dynamics change in response to different physiological states, such as rest and stress. At rest, the value of this parameter is smaller, and conversely, at stress, it is higher, which indicates increased chaos and complexity in the dynamics of the time series.
• ApEn and SampEn are used to study the complexity and predictability of the RR time series, which are identical but use different algorithms for their calculation. Higher values of ApEn/SampEn generally indicate greater unpredictability and chaos, while lower values indicate more regularity and predictability of the investigated signals. At rest, the RR signals are more stable and regular, resulting in lower ApEn/SampEn values, while under stress, the signals become more dynamic and chaotic, and the regulation of physiological processes is disturbed. Similar results were reported in [51].

3.3. Statistical Analysis Results

The evaluation of the importance of differences in HRV in different conditions (rest and stress) through the studied groups helps to evaluate whether changes in HRV are the results of random fluctuations or have statistical significance, which is important in the interpretation of RR time intervals. The resulting p parameter values obtained via the t-test indicate that there is a statistical significance when comparing Group 1 and Group 2 for all studied parameters, while some of the parameters have no statistical significance when comparing Group 1 and Group 3. This different behaviour is due to the fact that when playing using a virtual helmet (full-immersion virtual reality), higher stress is generated, while in the game using anaglyph glasses (virtual reality without immersion), less stress is generated, and the difference in the values of the studied parameters between Group 1 and Group 3 are smaller.

4. Conclusions, Limitations, and Future Work

Despite the importance of the relationship between mental stress and HRV, to date, there are a relatively limited number of studies that have examined this relationship in depth. In areas such as psychology, cardiology, sports medicine, and others, in recent years, interest in HRV as an indicator of the impact of stress on the autonomic nervous system has been growing due to the relevance of this topic. The application of the approach proposed in this article to investigate the impact of stress generated by an extreme 3D game on HRV can be used to analyze the cardiac health of people who cannot undergo physical exercise due to health reasons.

The obtained experimental results show that there is a relationship between HRV and stress, which can be used to determine the activity of the autonomic nervous system, reflecting the balance between the functions of the sympathetic (stress-related) and parasympathetic (relaxation-related) nervous systems. Lower HRV is generally associated with increased stress and sympathetic activation, while higher HRV indicates relaxation and parasympathetic dominance. The results show that VR can significantly affect HRV by inducing stress depending on the type of VR, namely whether it is immersive or non-immersive. Stressful VR scenarios can decrease HRV, reflecting the body’s natural stress response, while calming VR experiences can increase HRV, promoting relaxation and stress relief. Therefore, the interaction between HRV, stress, and VR highlights the potential of VR as a research tool and as a tool for therapeutic intervention in case stress is positive. The ability of VR to modulate HRV makes it a valuable tool for studying stress responses and developing techniques to manage stress and improve mental health.

One of the limitations of this study is that only 19 volunteers were included in it, but nevertheless, the obtained results of the statistical analysis show that a large part of the studied parameters had statistical significance between the studied groups. This suggests that, despite the limitation, the results can be considered reliable and provide useful information, although further research with a larger sample could confirm and extend this information.

Future work on this topic will continue in the following directions:

• Creating a 3D VR game to generate positive stress that can be applied for therapeutic purposes. This game will enable users to face challenges and stressful situations in a way that leads to positive emotions, confidence, and reduction in anxiety and stress.
• Creating a 3D VR game for smartphones by applying CardBoard technology. Using CardBoard technology is a cost-effective way to deliver a VR experience, allowing users to use their smartphones. This technology is supported by both Android and iOS devices.
• Creating 3D games for students to help them train in the subject of discipline stereometry, which is one of the most problematic disciplines for students who lack spatial imagination. These games will allow for a visual representation of the learning material being studied, and students will be able to generate, observe, and manipulate geometric figures themselves.