Usability Analysis of a Virtual Reality Exposure Therapy Serious Game for Blood Phobia Treatment: Phobos

Abstract

Phobias are characterized as the excessive or irrational fear of an object or situation, and specific phobias affect about 10% of the world population. Blood-injection-injury phobia is a specific phobia that has a unique physical response to phobic stimuli, that is, a vasovagal syncope that causes the person to faint. Phobos is a serious game intended for blood phobia treatment that was created to be played in virtual reality with an HTC Vive that has photorealistic graphics to provide a greater immersion. We also developed a console application in C# for electrocardiography sensor connectivity and data acquisition, which gathers a 1 min baseline reading and then has continuous data acquisition during gameplay. Usability tests were conducted with self-reported questionnaires and with a case study population of 10 testers, which gave insight into the previous game experience of the tester for both digital games and virtual reality games, evaluating the discomfort for hardware on both the sensor and the virtual reality headset, as well as the game regarding usability, user experience, level of immersion, and the existence of motion sickness and its source. The results corroborate that the immersion of the game is good, which suggests that it will help with triggering the phobia.

1. Introduction

The advances in computer graphics, virtual reality (VR), and sensors have opened up new possibilities in game design. We now have simulators for learning to drive automobiles and airplanes, virtual reality games that confront players with the fear of heights, and massive multiplayer online games that bring fantasy worlds to life, facilitating social and gameplay interaction among many people. The next step in game development is the integration with biometric sensors, which can combine biofeedback with game mechanics to create new gameplay experiences and enhance immersion.

Phobias are excessive or irrational fears of objects, situations, locations, or animals. When left untreated or not confronted, these symptoms can worsen, resulting in a poor quality of life for those suffering from the phobia, as they may avoid the phobic stimuli or even the situations or locations where the stimuli might occur. Phobias can be divided into three categories: agoraphobia, characterized by the fear or anxiety of a location; social phobia or social anxiety disorder, characterized by the fear of social situations; and specific phobias [1]. In the case of specific phobias, irrational fear is caused by a specific object or situation [2], affecting about 10% of the world’s population.

For this work, the focus was on blood-injection-injury (BII) phobia, also known as hemophobia. It is a specific phobia characterized by a fear of blood, injury, and injections, which can interfere with a person’s ability to function in daily life in more severe cases [3]. This phobia is unique in its physiological response pattern, which includes vasovagal syncope: a rise in heart rate and blood pressure followed by a sudden, rapid drop, leading to fainting due to low oxygen supply to the brain [4].

Phobia treatment methods include cognitive therapy, where patients learn to identify anxiety-inducing mental images and replace them with realistic concepts; relaxation techniques to control anxiety levels; exposure therapy involving gradual exposure to the phobia, making the trigger less effective; hypnosis, where patients are suggested under a state of hypnosis that they no longer have the fear; herbal remedies like valerian root to reduce anxiety levels; and pharmacotherapy [2].

Specifically, for BII phobia, treatment includes cognitive therapy with graded desensitization, starting with low-trigger phobic stimuli and gradually increasing the stimuli as the brain becomes accustomed to it, culminating with the highest-trigger stimuli. It also includes in vivo and imaginal exposure, where patients are either exposed live to the phobic stimuli or recount phobic episodes, respectively, and a combination of exposure therapy with relaxation [4]. Typically, BII is treated with exposure therapy—exposing the person to their phobic stimulus—and by employing applied tension exercises when dizziness symptoms occur due to a drop in blood pressure. The applied tension exercise, consisting of tensing the major body muscles, should always be performed in conjunction with exposure to the phobic stimuli to prevent the second part of the vasovagal syncope by increasing blood pressure, thus preventing the body’s fainting response [5,6,7].

The objective of this work was to explore serious games as support tools for the treatment of BII phobia. To achieve this, we created a virtual reality exposure therapy (VRET) serious game called Phobos, intended for BII phobia treatment. The research questions we aimed to answer as a case study analysis were:

(1)

Is smooth locomotion a feasible method in a virtual reality serious game designed for blood phobia treatment, and does it lead to motion sickness among players?

(2)

What is the adequate positioning for a BITalino electrocardiography (ECG) sensor to minimize its impact on a player’s performance while playing a VR serious game designed for blood phobia treatment?

(3)

What is the perception impact of photorealistic graphics on the level of immersion experienced by players in a serious VR game?

This paper is structured as follows: Section 2, Research Methodology, describes the methodology undertaken in this research; Section 3, Related Work, covers the literature; Section 4, Phobos Serious Game Development, provides an in-depth description of the game development of “Phobos”; Section 5, Console Application, describes the ECG sensor data acquisition console application; Section 6, Usability Validation, contains the usability case study analysis with a self-reported questionnaire, sensor data, and preliminary validation; Section 7, Discussion, provides a comparison between the findings of the usability test with other studies; and Section 8, Conclusions and Future Work, discusses final considerations and future developments.

2. Research Methodology

The research methodology used for this work was the action research case study, which focuses on solving a problem, in this case, creating a BII phobia VRET serious game to help BII phobia patients with the mitigation of their phobia, and the methodology consists of an iterative process divided into four steps (Figure 1).

The planning phase involves collaborating with stakeholders to identify the problem, defining project objectives, scope, and goals, formulating research questions or hypotheses, and selecting data collection methods. During this phase, the user strategically plans the actions necessary to address the problem, supported by relevant research.

The action phase is the result of prior planning efforts. In this stage, researchers execute the planned actions and gather pertinent data related to the problem. Data acquisition methods may include observations, surveys, interviews, document analysis, or other appropriate techniques.

The analysis phase entails reflecting on and analyzing the data collected during the action phase while assessing its alignment with the anticipated outcomes. This phase is essential for evaluating the impact of interventions and determining whether adjustments or further actions are warranted. If necessary, actions are refined and redefined, leading to subsequent cycles through the planning, action, and analysis phases.

The conclusion phase involves documenting the entire research process, encompassing problem identification, planning, actions taken, data collection, analyses conducted, and resulting outcomes. This phase culminates in a comprehensive conclusion summarizing the work undertaken.

In the first phase, there was a focus on research on the current studies, projects, and ongoing studies on the subject, which is the foundation of the theoretical framework, thus allowing a founded plan of action and development.

In the second phase, the focus was the development of both the game and the sensor’s integration with the game “Phobos”, where the biofeedback would change the phobic stimuli intensity level after measuring the stress level in comparison to the baseline levels measured before the play-session or would give the psychologist some insights about the emotional and physical state of the player. This development was used as a proof-of-concept for biometric sensor integration in a virtual reality context to enhance the experience and gather further results.

This could be achieved by either doing the sensor integration directly on Unity or by developing an external application to run parallel to the game and then communicating the retrieved data to a database for posterior analysis. The considered database is a Microsoft Structured Query Language (SQL) server relational database, where there will be a session table that has a varbinary column to store the session store file, two date-time columns to store the start and end date-time of the session, and an integer or Globally Unique Identifier (GUID) column to store the foreign key of the player.

In the third phase, there was a division into two stages: the first stage was for testing—both usability testing and functional testing—while the second stage was for analysis of the acquired data. There was a case study population of 10 testers for the conducted usability tests, one of which suffers from BII phobia, although the phobia of this tester manifests with their own blood instead of other people’s blood.

In the fourth and final phase, there was a review and retrospection of the previous phases and a correlation with the obtained data.

3. Related Work

In this section, we present the literature review on serious games for the treatment of anxiety disorders, more specifically phobias, on game design, starting from conventional game design to serious games game design with the commonly used frameworks/methodologies, on VRET for phobia treatment, and on biometric sensors used on games. The sections end with the research methodology used in this work and a research opportunities subsection that states the gaps in the current literature and where our research could fill in those gaps.

3.1. Serious Games

Serious games are digital interactive applications designed to educate or train users, imparting knowledge or altering their beliefs and attitudes [8]. They draw upon concepts from commercial computer games to visualize technical systems and convey educational content [9]. Serious games, regarded as multimedia systems, incorporate elements such as 3D graphics, audio processing, human factors, and real-time network communication [10]. Their potential extends to impacting research, development, and application in computer graphics and related fields, with diverse applications encompassing health, culture, interaction techniques, and 3D player data visualization [11]. These games, created using game technology and design principles, primarily serve an educational purpose, offering an engaging context to motivate and educate players.

The “serious game experience” characterizes the overall engagement, enjoyment, and satisfaction players derive from educational serious games. Typically, these games aim to educate or train users in specific subjects or skills. Key concepts used to understand and model the serious game experience include flow theory and immersion. Flow theory posits that players fully engage and immerse themselves in the game, optimizing learning outcomes when in a state of “flow.” Immersion refers to the sensation of being physically or virtually immersed in the game experience. Additional dimensions contributing to the serious game experience encompass competence, challenge, positive affect, negative affect, tension, and positive value, collectively influencing the overall experience and the effectiveness of serious games for educational purposes [12].

Individuals suffering from specific phobias often exhibit reluctance to seek treatment [13]. Nevertheless, research indicates that specific phobias are highly treatable psychological disorders when individuals overcome their hesitance and seek help [14]. Exposure therapy techniques are commonly regarded as the primary treatment approach and are widely employed with phobic patients, demonstrating substantial efficacy compared to other methods [14,15]. Consequently, computer-based tools offer a promising avenue to enhance psychological treatments, reduce attrition rates, and amplify their scope [16]. Research has explored the potential effectiveness of computer games as adjuncts to psychotherapy, such as solution-focused therapy games designed to boost adolescents’ motivation for therapy [14,17,18]. Other examples include serious games rooted in cognitive–behavioral therapy, like Treasure Hunt [19] and gNats [14]. Notably, the National Institute for Clinical Excellence recommends computer-aided psychotherapy programs, such as Beating the Blues [20] and FearFighter [21], that target adults with mood and anxiety disorders [15].

In recent years, an emerging trend in psychotherapy leverages advanced technology, particularly VR applications, for assessing and treating mental health issues [22,23]. VR, characterized as an advanced human–computer interface, empowers individuals to interact with problematic situations related to their issues within a controlled environment, free from the sense of threat [23]. Moreover, VR has applications in physical rehabilitation, benefiting post-stroke patients and others [24]. VRET has emerged as a novel medium for exposure therapy that shows promising efficacy, particularly for specific phobias [25,26,27]. It offers numerous advantages over in vivo or imaginal exposure techniques.

3.2. Game Design

Game design is a multifaceted field that encompasses elements such as game balance, storytelling, non-linear progression, player motivations, input/output mechanisms, artificial intelligence, and level design [28]. Designers rely on their prior experience and player feedback when developing games, drawing from past designs. However, when exploring new design territories, designers may require practical and technological support to facilitate feedback collection from unfamiliar contexts [29]. Game design also serves as an educational tool, allowing students across various subjects to create games, acquire design knowledge, and explore creative possibilities within specific constraints [30]. Furthermore, the concept of “designing as game playing” extends to architectural design courses, where students create games inspired by renowned architects, enhancing analytical skills and deepening design expertise [31].

The MDA framework (Mechanics, Dynamics, and Aesthetics) is a commonly used framework in game design. Mechanics represent the technical components at the code level, dynamics capture player interactions with these mechanics, and aesthetics encompass players’ emotional responses to these dynamics [32]. However, the MDA framework has faced criticism due to the subjectivity of aesthetics and two main shortcomings: a potential overemphasis on game mechanics—neglecting other design aspects—and its unsuitability for all game types [33,34,35]. To address these concerns, the DDE framework (Design, Dynamics, and Experience), introduced by Walk et al. in 2017, builds upon the MDA framework, Schell’s Elemental Tetrad [36], and the DPE framework (Design, Play, Experience) [37], specifically tailored for serious games. DDE restructures the MDA framework, subdividing each category into sub-categories. Design, like the previous Mechanics category, involves the intellectual analysis of game design, an aspect entirely controlled by game designers. It is further divided into Blueprint (containing world description and style), Mechanics (encompassing game code, rules, objects, and architecture), and Interface (comprising functional elements, diegetic/non-diegetic elements, spatial/meta information, and content, including narratives, sound, graphics, and interaction design). Dynamics represent player interaction with the design aspect, mediated by the player-subject, with designers exerting indirect control through rules. Interactions encompass player–game, player–player, and game–game interactions, ultimately creating an Antagonist for the player-subject. The Experience aspect, shaped by this Antagonist, represents the overarching design goal and further divides into Senses (encompassing audiovisual elements), Cerebellum (pertaining to emotions), and Cerebrum (involving intellectual challenges) and culminating in Perception, which includes aspects like gameplay, fun, beauty, and story, reflecting the emotions and experiences felt by players throughout their engagement [35].

While these frameworks significantly influence the design of entertainment games, game design for serious games involves critical factors contributing to their success or failure, categorized into three principal axes: project organization, game mechanics, and graphics [38]. Designing serious games to popularize medical science requires integrating relevant content into engaging and interactive game segments, presenting it through narrative and design methodologies, and guiding users to modify their attitudes and behaviors [39]. The Activity Theory-based Model of Serious Games (ATMSG) offers a visual framework for delineating the gaming, learning, and instructional aspects of learning game mechanics and flow [40]. Additionally, design principles for serious games can be derived from the analysis of project portfolios and outcomes, enriching the understanding of serious games from a gaming perspective [41]. Research into serious games has yielded incongruent findings, posing challenges in designing effective educational games [42].

3.3. Virtual Reality Exposure Therapy (VRET)

With the advancements in virtual reality hardware and software, the development of controlled environments for more effective exposure in VRET has become more accessible [43]. This therapy, often conducted with a therapist monitoring the VR display to detect patient avoidance patterns [1], has proven as effective as classical methods for treating anxiety disorders, including phobias [44].

Ongoing studies in VRET include a study for acrophobia, the fear of heights, involving 49 participants in a tailor-made virtual environment, where skin conductance and heart rate were measured to assess fear responses [45]. Another study focused on a prototype for treating multiple phobias using a handheld mobile VR headset [46]. A specific case study on one-session treatment of BII phobia, where the researchers did a randomized controlled trial, suggested post-treatment improvements in their fears but highlighted that single-session VRET treatment is not enough, and it needs to be accompanied by in vivo exposure therapy [47]. Other studies have also explored VRET and Augmented Reality Exposure Therapy (ARET) serious games for their potential to address specific phobias, which affect about 10% of the global population [48,49].

Realism and immersion are critical in VRET, with elements like photorealism, authenticity, immersion, and scenario development playing significant roles. While VRET’s effectiveness for fears like heights and flying is supported by evidence, further research, including randomized clinical trials comparing VRET with standard exposure and assessing VRET as a stand-alone treatment, is needed for other phobias. Following these trends, we consider that the development of a VR game with interactable objects, movement, immersion, and a well-crafted storyboard, alongside biometric signal measurement, could enhance BII phobia treatment and offer a superior alternative to current VR experiences.

3.4. Biometric Sensors

Biometric sensors, devices designed to measure and capture physiological and behavioral data, are used in various applications, including security, identification, health monitoring, and personal technology. They detect and record unique biological or behavioral characteristics for purposes like authentication, tracking, and analysis. Common types of biometric sensors and their applications include fingerprint sensors to capture the unique ridge and valley patterns on a person’s fingertip for security systems; smartphones and access control systems; iris scanners to capture unique patterns in the person’s iris and for secure access control and identity verification; face recognition sensors for the same applications of the previous two sensors that track facial features and unique characteristics of individuals; voice recognition systems, tracking vocal characteristics such as pitch, tone, and speech patterns used in voice authentication and voice-activated systems; heart rate monitors to track heart rate and heart rate variability used in fitness trackers and health monitoring devices; ECG to record the electrical activity of the heart for medical diagnosis and monitoring; electroencephalogram (EEG) to measure the electrical activity in the brain and map brain activity for medical applications, research purposes, and brain–computer interfaces; gait analysis sensors to capture the movement characteristics of a person’s walk pattern; skin conductance sensors to measure the electrical conductance of the skin for emotion recognition systems and lie detection algorithms; and other biometric wearables that allow systems and researchers to use the data acquired and transform it to be usable in their systems [50,51,52,53].

Biometric sensors are integral to a wide range of applications, including access control, identity verification, healthcare, and personal technology. They play a critical role in ensuring security, improving healthcare outcomes, and enhancing user experiences.

Proof-of-concept games that use these sensors include Dekker and Champion’s “Half-Life 2” modified level [54], where sensors alter the avatar movement speed and activate hearing ability and other game mechanics [55], and an academic project using the games “Torque” and “Burnout”, where a game level was adapted for immersive play in a low-cost dome, with biofeedback controlling an LED light rim based on player stress levels [54].

A significant study by Martinho, M. et al. explored a multibiometric system using an ECG and a blood volume pulse sensor for multimodal biofeedback in human–computer interaction [56].

3.5. Research Opportunities

Our literature review identified the three main phobia groups—agoraphobia, social phobia, and specific phobias, which divide into many specific phobias, including blood-injection-injury phobia and arachnophobia, among others. We examined their triggers and mental processes, management strategies, and treatment methods. In particular, we detailed BII phobia’s unique characteristics and treatment methods, including exposure therapy combined with applied tension exercises. We explored game design with accepted frameworks from Mechanics, Dynamics, Aesthetics to Design, Play, Experience and finally, Design, Dynamics, Experience, specified the game design aspects for developing serious games, and described serious games and their societal applications. We identified the ways that VRET is a viable treatment option for phobias, with the customization that can be done and with the safe and controlled environments that can be created. Additionally, we examined biometric sensors, their societal uses, and gaming applications.

Our research revealed a gap in serious games for BII VRET treatment, as current studies primarily involve passive VR experiences. This led to our objective of creating a virtual environment for BII phobia treatment utilizing serious games as a supportive tool. Our study of biometric sensors identified a lack of research correlating patient heart rate with exposure to phobic stimuli, prompting the development of a sensor application for heart rate tracking during VRET sessions and raising questions about the impact of sensor placement on gameplay performance. Given that specific phobias affect a significant portion of the population, our work aims to provide valuable data through the development of a photorealistic VR serious game using Unity 2021.3.18f1, addressing psychological and usability considerations and potentially inspiring further research in this field.

4. Phobos Serious Game Development

In this section, we present the development of the game Phobos with the game requirements, where we describe the system, functional and non-functional requirements, the design process, programming, and virtual reality development.

4.1. Game

Phobos emerged from a compelling need to assist individuals battling various phobias, employing an uncommon treatment method known as VRET [43]. Utilizing the latest consumer-grade VR technology, we aimed to sculpt a hyper-realistic environment for players to experience incremental and secure exposure therapy tailored to their phobias. Through in-depth dialogues with our colleagues, BII phobia was selected as the focus due to its distinct physiological response. We harnessed the power of photorealistic VR to submerge players into the experience, diverting their spatial awareness and concentrating their attention on overcoming the phobic stimuli presented in the game.

The game is designed as a first-person VR serious game with role-playing elements. The player assumes the identity of a detective with the freedom to navigate an apartment in a virtual setting. They can collect clues, assimilate information about their phobia, learn about therapeutic methods, and face their fears through controlled exposure. The core gameplay mechanics involve gathering and scrutinizing objects or clues, absorbing their background, and solving intricate puzzles to unravel the enigma at hand. As players probe deeper and draw nearer to a resolution, their exposure intensifies. They are aware that unveiling the truth necessitates exposure, yet they retain command over it, enabling mental preparation for the personal challenge ahead.

4.2. Requirements

Functional requirements map the functionalities to be implemented by the developer so that the final user can complete their objectives. The functional requirements of the game are the following [57]:

• The game will be a 3D virtual reality single-player first-person game, where the player, as the user who will play the game, will be allowed to move freely through the game scenario with the rigid body of the VR set that the player controls.
• There will be collisions between the VR rigid body and game objects, which could be static, interactive, and movable objects.
• The sensor must register data with its ECG (electrocardiography) data acquisition for future dynamic control over the phobic stimuli.
• There will be spatial sound objects placed on specific objects to increase immersion, for instance, the sound of the air conditioners when the player gets near them.
• The game will have mechanics to unlock doors, to solve puzzles made of paintings, or to grab and inspect objects.
• The game must have blood objects for phobic stimuli, distributed gradually over the various rooms of the scene for gradual exposure.

Non-functional requirements define how the system is and behaves, and Phobos’ are as follows [57]:

• The game must have medium to high stable fps, preferably above 30.
• The sensor must have a continuous connection to the computer.
• The area for the VR system must be unobstructed to reduce the possibility of accidents.
• The player must feel that the blood is realistic enough to trigger the phobia.
• The game must have immersion so that the player can be fully committed to the exposure therapy.

The minimum system requirements for Phobos are a Windows 10 OS, Intel Core i5-4590 equivalent or greater CPU, 8 GB of RAM, Nvidia 970 or AMD Radeon R9 or greater graphics card, 2 GB of storage space, Steam VR installed, and standing or room-scale VR. It also requires an HTC Vive Head Mounted Display (HMD) or HTC Vive Pro 2 HMD and a PLUX Biosignals BITalino Revolution Board Kit with the ECG sensor connected and the three-electrode cable.

For the development and testing of the game and sensor integration, the computer used had the following specifications: a Windows 11 OS, Intel Core i7-8750H CPU @ 2.20 GHz, Nvidia GeForce RTX 2080 8 GB GDDR6, and 32 GB RAM. For the HMD, the HTC Vive Pro 2 was chosen due to the graphic processing power with the 1080 × 1200 Dual AMOLED 3.6” screens for greater fidelity on graphics.

4.3. Design

The genesis of Phobos’ design was a collaborative effort initiated through strategic discussions with experts in psychology to discern the ideal game genres suitable for a VRET application targeting BII phobia. These discussions gave rise to a brainstorming phase, where developers synthesized a plethora of concepts into a cohesive narrative and design, selecting mechanics and dynamics that would culminate in a compelling gameplay experience.

A foundational principle in the creation of Phobos was the elimination of ‘safe zones’ to prevent players from employing avoidance strategies, which could be detrimental to therapeutic outcomes [1,2,3]. We endeavored to construct a game environment that would engage the player with a narrative that provokes curiosity and the motivation to progress, thus facilitating a measured and deliberate exposure to the phobic elements. Positioning the player as a detective in a meticulously designed penthouse, the game invites exploration of a multifaceted murder mystery, with clues ingeniously integrated with the phobic stimuli, blood (Figure 2), which are strategically placed and increase in intensity as the player navigates the house, ensuring a gradual and escalating exposure to the phobia.

In concert with the psychology experts’ insights, it was decided that Phobos would include an informative tutorial on the nuances of the phobia. The tutorial was crafted to be a multifaceted case folder that not only educates on the movements and interactions within the game but also imparts knowledge about the phobia’s physical and psychological aspects as provided by the psychology department [58].

The asset modeling for the game unfolded in two meticulous stages. The initial stage focused on crafting a house layout with precise scaling to provide a comfortable and believable space for players (depicted in Figure 3).

Following this, we sought out photorealistic shaders and materials to give the structure a lifelike appearance. The subsequent stage involved the selection of hyper-realistic assets from architectural visualization resources, bringing the environment to life with accurate lighting and post-processing effects that impart a tangible realism to each object within the game space (as illustrated in Figure 4).

4.4. Programming

Programming Phobos, conducted in the C# language within the Unity engine, required the development and attachment of scripts to various game objects. These scripts enable an assortment of operations, from the fundamental translation, rotation, and scaling of objects to more complex behaviors that alter the game’s dynamics. For instance, the game’s artwork is governed by a triad of scripts: one to manage VR interactions with the painting, another to execute a 90-degree rotation with a smooth transition upon interaction, and a third script that oversees the alignment of paintings, which, upon correct positioning, results in the revelation of a key object that is spawned from the middle painting.

The game’s case folder is a nexus of multiple scripts, monitoring interactions with clue objects, controlling auditory feedback, and managing the display of information within the folder. The auditory feedback is activated by the interaction of the player with the folder’s tab or page or by opening the folder, resulting in different sounds depending on the interaction. Additionally, door interactions are scripted to simulate a realistic lock mechanism, complete with sound cues and detection of key interactions to unlock doors.

4.5. Virtual Reality

To bring the best quality and performance to this work, amongst the multiple consumer VR sets available on the market, we have chosen the HTC Vive (Figure 5) and HTC Vive Pro 2, as they bring dual AMOLED 3.6” screens with a 1080 × 1200 pixels per eye resolution, 90 Hz refresh rate, and a field of view of 110 degrees for the HTC Vive and dual RGB low-persistence LCD screens with 2448 × 2448 pixels per eye resolution, 90/120 Hz refresh rate (limited to 90 if in use with the wireless adapter), 120 degrees of field of view, and high-resolution headphones for the HTC Vive Pro 2, which meet the requirements for photorealistic VR development.

The development of a VR game with the HTC Vive requires that we install both the Steam VR (1683151457) for setup and configuration and the Steam VR SDK for Unity, which brings a full set of tools and presets that we can use and extend for creating our custom actions and interactions.

Customized controller bindings were implemented to accommodate unique interactions, such as “Toggle Folder” actions on the grip button and trackpad-based navigation within the case folder with the methods “PreviousTab”, “NextTab”, “Previous Page”, and “NextPage” (see Figure 6 and Figure 7 for binding details).

The right controller was programmed for smooth locomotion, incorporating a script to interpret finger tracking on the trackpad and translate it into player movement, with raycasting applied to preemptively detect and prevent collisions. The smooth locomotion is coded in multiple steps. The first one is the speed variable with the direction where the HMD is facing in the world, applying a transform over the y-axis times the sensitivity variable, set to 2, and clamping the value between 0 and the set maxSpeed value, set to 5. Finally, we multiply the speed with deltaTime and a projection on the floor plane. We had to implement a ray cast to detect the collision with objects in the movement’s direction for the knee and head to prevent the collisions with said objects in 0.25 f distance.

For interaction with game objects, we had to include the SteamVR Interactable script in the game object, which allows for the pickup interaction that defines whether the hand is shown while holding the object; the throwable script, which enables the throwing action of the object; and the OnPickup and OnDetatchFromHand methods that were mapped to the “inspect” and “uninspect” custom scripts, which trigger the interaction with the objects, show information on one of the displays, and trigger the flags for puzzles and case folder information.

For the doors, paintings, and telephone interaction, since they are objects that cannot be picked up, we have created custom interaction scripts that are triggered when the player has a hand hovering over the object and checks if the grab button is pressed to trigger the corresponding mechanic of opening/closing the door, rotating the painting, or listening to the voicemail.

5. Console Application

This section delves into the intricate process of creating a console application tailored for managing data from an electrocardiography sensor. It covers the sophisticated data structures designed to encapsulate the ECG readings, the strategic placement of the sensor to ensure non-invasive yet precise data capture, and the comprehensive methods employed for the acquisition of ECG data.

5.1. Biometric Sensor

For the heart rate monitoring component, we opted for the BITalino Revolution Board Kit ECG sensor from PLUX Biosignals. This selection was backed by the sensor’s compliance with industry benchmarks [59] and its versatile multimodal capabilities. This sensor kit also includes an array of integrated sensors, such as those for electrodermal activity and electroencephalograms, paving the way for expansive future research [60].

We developed a new C# console application incorporating the BITalino base library for sensor interfacing alongside the Newtonsoft.Json library for data management. The application initializes by listing available devices and prompting for the MAC address of the BITalino sensor; this step is a one-time setup, as the address is then stored locally. Post-connection, the application commences data acquisition at a rate of 1000 Hz, populating a list with 1000 ECG frames and initializing a separate list for baseline heart rate readings—a one-minute test that players can opt to retake.

The ECG readings are derived using the following Equation (1):

$$\mathrm{E}\mathrm{C}\mathrm{G}\left(\mathrm{m}\mathrm{V}\right)=\frac{\left(\frac{\mathrm{A}\mathrm{D}\mathrm{C}}{2^{\mathrm{n}}}-\frac{1}{2}\right)\times \mathrm{V}\mathrm{C}\mathrm{C}}{{\mathrm{G}}_{\mathrm{E}\mathrm{C}\mathrm{G}}}\times 1000$$

(1)

where:

♢

ADC represents the raw digital value;

♢

n denotes the sampling resolution, typically 16-bit or 8-bit;

♢

VCC is the operating voltage, generally 3 V; and

♢

G_ECG signifies the sensor gain.

This formula facilitates the monitoring of each heartbeat, identifying instances where the readings exceed the 0.1 mV threshold. These occurrences are archived in both .json (for individual readings) and .txt (for the complete ECG dataset) formats, ensuring data are available for subsequent analysis.

5.2. Data Structure

We designed a class named ECGReading to store the ECG data snapshots. It comprises a ‘double’ for the reading value, a ‘DateTime’ object for the precise timestamp, a ‘long’ to count the heartbeats since the beginning of the recording, and a ‘bool’ to distinguish baseline tests from gameplay readings. Post-baseline, the heartbeat counter resets.

This class facilitates recording each gaming session’s ECG data in a .json file. These files are then intended to be organized within a relational database that includes a ‘Session’ table with temporal markers, binary columns for .txt and .json files, and a ‘Player’ table with pertinent information. A ‘SessionECGReading’ table, with a many-to-one relationship to the ‘Session’ table, will hold the extracted ECG objects from the session files, enabling detailed analyses.

5.3. Sensor Placement

Initial tests with a dual wet electrode ECG sensor on the player’s wrist yielded excessive noise during VR gameplay, leading us to discard this placement. Subsequent investigation directed us towards the use of a triple dry electrode kit positioned in accordance with the Einthoven triangle formation (as depicted in Figure 8 by Biosignalsplux [61]). This configuration places the electrodes on the right arm, left arm, and left leg, with the heart at the triangle’s center for optimal signal clarity.

To further mitigate muscular movement interference, electrodes were alternatively positioned on the right and left collarbones and the left hipbone, significantly reducing noise, and enhancing data quality during VR interactions.

5.4. Data Acquisition

Data acquisition was meticulously carried out during usability tests, with subjects donning the sensor in the Einthoven configuration. The process encompassed an initial baseline test for sensor calibration, followed by data collection throughout gameplay. Data were captured in .json format for heartbeat events and .txt format for the full ECG trace. Participants were fully briefed on the data collection scope and consented to the recording of biometric and screen data.

Looking forward, the aim is to transition from local storage to a web-based database, enabling secure uploads and facilitating remote analysis via a web application. This would support comprehensive data examination by psychologists overseeing phobic patients, as illustrated in the proposed architecture (Figure 9).

6. Usability Tests

In this section, we present the case study experimental results gathered. It starts with the usability questionnaire that was created, followed by the results of the usability tests that were conducted, followed by the data acquired by the sensor, and ending with a preliminary validation and suggestions that we were able to gather, both during the usability tests and from our peers from the Psychology Department of Minho University, Portugal.

6.1. Questionnaire

To gather results, we did usability tests followed by self-report questionnaires (Appendix A,

Table A1. Self-report questionnaire.

	Question	Possible Answer(s)
Q1	What is your gender?	Female; Male; Prefer not to disclose; Other.
Q2	Choose your age from the options given.	<15; 15–20; 21–30; 31–40; 41–50; 51–60; >60.
Q3	What are your academic qualifications?	Under four years of scholarity; Pre- school; Middle school; Highschool; Bachelor; Master; PhD; Prefer not to disclose
Q4	Do you usually play digital games?	Yes. No.
Q5	Have you ever played a virtual reality game before this one?	Yes. No.
Q6	Did you find the sensors’ placement invasive?	Likert scale 1–7: 1- Strongly Disagree 7—Strongly Agree
Q7	Did the sensors’ placement obstruct your movements?	Likert scale 1–7: 1- Strongly Disagree 7—Strongly Agree
Q8	Did you find the virtual reality gear uncomfortable?	Likert scale 1–7: 1- Strongly Disagree 7—Strongly Agree
Q9	If you felt any discomfort using the virtual reality gear, choose from the following answers those that apply.	Due to the weight. Due to the cables. Due to the lenses. Due to the controllers. Other
Q10	Did you find the controls of the game to be straightforward? (GUESS)	Likert scale 1–7: 1- Strongly Disagree 7—Strongly Agree
Q11	Did you always know how to achieve the objectives/goals of the game? (GUESS)	Likert scale 1–7: 1- Strongly Disagree 7—Strongly Agree
Q12	Did you find the blood exposure difference evident between the various rooms?	Likert scale 1–7: 1- Strongly Disagree 7—Strongly Agree
Q13	Did you feel the game gave you enough freedom to act how you want? (GUESS)	Likert scale 1–7: 1- Strongly Disagree 7—Strongly Agree
Q14	Did you feel that the game’s audio (e.g., sound effects, music) enhanced your game experience? (GUESS)	Likert scale 1–7: 1- Strongly Disagree 7—Strongly Agree
Q15	Did you think the game was visually appealing? (GUESS)	Likert scale 1–7: 1- Strongly Disagree 7—Strongly Agree
Q16	When you were playing the game, you lost track of the world around you? (UES)	Likert scale 1–7: 1- Strongly Disagree 7—Strongly Agree
Q17	Did you feel frustrated playing the game? (UES)	Likert scale 1–7: 1- Strongly Disagree 7—Strongly Agree
Q18	Did you feel like you were able to interact with the environment the way you wanted? (HRC)	Likert scale 1–7: 1- Strongly Disagree 7—Strongly Agree
Q19	Did you find that the environment was responsive to the actions that you initiated (or performed)? (UES)	Likert scale 1–7: 1- Strongly Disagree 7—Strongly Agree
Q20	Did you find the puzzle of the painting in the living room?	Yes. No.
Q21	Did you find the mobile phone on the ground in the living room?	Yes. No.
Q22	Did you feel motion sickness during the gameplay session?	Likert scale 1–7: 1- Strongly Disagree 7—Strongly Agree
Q23	If you felt motion sickness, choose the reason from the following answers.	Due to the graphics. Due to the movements’ velocity. Due to the movement mechanics. Other.
Q24	Would you play this game again once it’s fully developed?	Likert scale 1–7: 1- Strongly Disagree 7—Strongly Agree

), which were based on multiple relevant questionnaires on usability, engagement, experience, and satisfaction combined with questions that we considered relevant for this specific game [57].

One of the studied questionnaires had a similar virtual reality project for a heart rate-controlled interaction game (HRC) [62], another was the User Engagement Scale [63], and the other relevant questionnaire was the Game User Experience Satisfaction Scale [64]. Both UES and GUESS use Likert scales, which measures the satisfaction levels of the users.

6.2. Results

We conducted the usability tests (

Table 1. Usability questionnaire answers part I.

	Q1	Q2	Q3	Q4	Q5	Q6	Q7	Q8	Q9	Q10	Q11	Q12	Q13	Q14
A1	Male	31	High school	1	0	1	1	1		7	7	7	7	7
A2	Male	21	Masters	1	1	2	1	1		6	6	7	7	7
A3	Male	31	Bachelor	1	1	1	1	2	Due to the lenses	6	1	3	7	5
A4	Female	21	High school	1	0	1	1	1		6	5	7	7	1
A5	Male	21	Masters	1	1	3	1	2		5	3	5	6	5
A6	Female	31	Bachelor	0	1	1	1	5	Due to the weight, Heat	4	5	6	6	5
A7	Male	31	Masters	1	1	2	2	3	Due to the lenses	4	6	6	5	5
A8	Female	31	Bachelor	0	0	1	1	2	Due to the lenses	6	6	6	7	6
A9	Female	41	Bachelor	0	1	1	1	2	Due to the weight	5	6	6	6	6
A10	Male	31	Bachelor	1	1	1	2	3	Due to the cables	6	6	7	5	7
σ		6.32		0.48	0.48	0.70	0.42	1.23		0.97	1.79	1.25	0.82	1.78
μ		29		0.7	0.7	1.4	1.2	2.2		5.5	5.1	6	6.3	5.4
Median		31		1	1	1	1	2		6	6	6	6.5	5.5

and

Table 2. Usability questionnaire answers part II.

	Q15	Q16	Q17	Q18	Q19	Q20	Q21	Q22	Q23	Q24
A1	7	7	1	7	7	0	1	1		7
A2	6	5	3	6	6	1	1	1		6
A3	5	6	5	6	6	1	1	7	Due to the movements’ velocity	7
A4	7	6	1	7	7	1	0	7	Due to the movements’ velocity	7
A5	5	6	1	5	5	0	1	1		6
A6	5	7	1	3	3	1	1	6	Due to the movement mechanics	7
A7	6	5	2	2	3	1	1	5	Due to the movements’ velocity	3
A8	6	7	2	7	7	1	1	7	Due to the movements’ velocity, Due to the movement mechanics	4
A9	6	3	3	5	6	1	0	4	Lack of practice with these games	6
A10	6	7	2	5	4	1	1	3	Due to the movement mechanics	7
σ	0.74	1.29	1.29	1.70	1.58	0.42	0.42	2.57		1.41
μ	5.9	5.9	2.1	5.3	5.4	0.8	0.8	4.2		6
Median	6	6	2	5.5	6	1	1	4.5		6.5

) with a population of 10 testers whose ages ranged from 21 to 50 years old; 60% were 31 to 40 years old, and 60% were male. While half of the testers had a bachelor’s degree, 30% had already finished a master’s degree, and 20% had not finished their bachelor’s degree. The selection of these testers was based on their prior experience with playing digital games, which rendered their input valuable. Ethical considerations were also taken into account, ensuring voluntary participation. Each participant was informed in advance that they were free to choose to participate and could leave the study at any time without facing any repercussions.

For previous experience with digital and VR games (

Table 3. Participants’ demographic characteristics.

Demographic Variables		Frequencies (%)	Q4 (Mean/std. dev.)	Q5 (Mean/std. dev.)
Genre	Male	6 (60%)	1.00/0.000	0.83/0.408
	Female	4 (40%)	0.25/0.500	0.50/0.577
Age	[21,22,23,24,25,26,27,28,29,30]	2 (20%)	1.00/0.000	0.50/0.707
	[31,32,33,34,35,36,37,38,39,40]	7 (70%)	0.71/0.488	0.71/0.488
	[41,42,43,44,45,46,47,48,49,50]	1 (10%)	0.00/-	1.00/-
Academic Qual.	High school	2 (20%)	1.00/0.000	0.00/0.000
	Bachelor’s	5 (50%)	0.40/0.548	0.80/0.447
	Master’s	3 (30%)	0.70/0.483	0.70/0.483
Q4 (mean/std. dev.)				0.70/0.483
Q5 (mean/std. dev.)				0.70/0.483

), 70% usually play digital games, and 70% had played VR games before the test. We can also note on this matter that every male participant had prior usual experience in playing digital games, and 83% had played VR games before; as for female participants, only 25% had prior gameplay experience, and 50% had already experienced VR games. The experience across the age groups was as follows: the participants 21–30 years old had the habit of playing digital games, but only one had prior VR experience; for the participants 31–40 years old, 71% had the habit of playing digital games, and 71% had prior VR experience; and the participant in the 41–50 age range did not usually play digital games and had played a VR game before the test. The experience with respect to academic qualifications was as follows: the participants with high school degrees had usual digital gameplay experience, and it was their first time experiencing a VR game; for the participants with Bachelor’s degrees, 40% had usual digital gameplay experience, and 80% had previously tried VR games; and the participants with Master’s degrees had both usual digital gameplay and VR experience.

From the feedback gathered about the sensor, since we used a Likert scale of 1–7 where lower is better for Q6, Q7, and Q8 (

Table 4. Q6, Q7, and Q8 statistics. These questions provide information about the sensor’s placement, comfortability, and obstruction of movement.

		Q6	Q7	Q8
N	Valid	10	10	10
	Missing	0	0	0
Mean		1.40	1.20	2.20
Median		1.00	1.00	2.00
Mode		1	1	2
Std. Deviation		0.699	0.422	1.229

), we can acknowledge that the testers did not find the placement invasive and that it did not obstruct their movement, as the median and mode was 1, strongly disagree, to Q6 and Q7 with a close to 1 mean of 1.40 and 1.20, respectively.

The VR HMD, however, had mixed feedback. Although the median was only 2 for disagree with finding the headset uncomfortable (Q8), we had one person that had 4, neutral opinion and two testers with 3, lightly disagree. With Q9, 60% of the participants answered; furthermore, we were able to segregate the cause of discomfort of the VR HMD device (

Table 5. Q9 frequencies, reporting the reasons for the discomfort felt by the testers while using the virtual reality equipment.

		Responses		Percent of Cases
		N	Percent
Q9 a	Due to the lenses	3	42.9%	50.0%
	Due to the weight	2	28.6%	33.3%
	Due to the cables	1	14.3%	16.7%
	Heat	1	14.3%	16.7%
Total		7	100.0%	116.7%

a. Dichotomy group tabulated at value 1.

). We learned that the discomfort caused by the lenses affected three testers, the weight affected two testers, and the HMD cables affected one tester; additionally, one tester said the problem was due to the heat of the HMD while it was running. While the lenses might have a solution by further calibration to the player’s head, the cables can be substituted by a wireless add-on, and the heat could be mitigated by having the HMD within a temperature-controlled room, the weight cannot be mitigated, which makes for 28.6%.

The next queries, Q10 to Q16, inform us about the game’s usability and experience, where a higher rating on the Likert scale equals positive feedback (Figure 10). With a population of 10 testers, the representation of these data scales from 0–100%, where each color has an associated number for the number of frequencies for that value. We can observe that overall, 80% of the participants at least lightly agree with the questions, and over 50% agree or strongly agree with the statements via segmentation.

The testers mostly did not feel frustrated during the gameplay session, per Q17 (Figure 11). Where a lower score on the Likert scale is better, we observed that 40% strongly disagreed, 30% disagreed, 20% lightly disagreed, and only 10% lightly agreed with the statement; after being asked for the reason, they related their evaluation to not having more interactable objects, as stated before.

This relates to the following questions, Q18 and Q19 (Figure 12), where a higher score on the Likert scale is better. There were mixed reactions for Q18, with the interaction with the environment having a median of 5.5, 20% strongly disagree or disagree with the statement, and 30% only lightly agree, which relates to the feedback from Q13 of having more objects to interact with, and most found that the environment was responsive when interacted with (Q19), with a median of 6 for agree, which is evident with 60% having positive feedback.

We had mixed responses on motion sickness: 50% felt motion sickness, where 30% of these strongly agreed with feeling motion sickness, and the other 20% were equally distributed between agree and lightly agree with the statement, which is a higher number than we would like for Q22. The median was 4.5 with a standard deviation of 2.573 (

Table 6. Q22 statistics, describing the feeling of motion sickness by the tester.

Q22
N	Valid	10
	Missing	0
Mean		4.20
Median		4.50
Mode		1 a)
Std. Deviation		2.573

a). Multiple modes exist. The smallest value is shown.

); 10% were neutral, 10% lightly disagreed with feeling motion sickness, and 30% were comfortable and strongly disagreed. When we analyzed the answers for Q23 (

Table 7. Q23 frequencies, enumerating the reasons why the testers felt motion sickness.

		Responses		Percent of Cases
		N	Percent
Q23 a)	Due to the movements’ velocity	4	50.0%	57.1%
	Due to the movement mechanics	3	37.5%	42.9%
	Lack of practice with these games	1	12.5%	14.3%
Total		8	100.0%	114.3%

a). Dichotomy group tabulated at value 1.

), we observed that most responses were due to the movements’ velocity, totaling 57.1% of cases, 42.9% were due to the movement mechanics, and 14.3% stated that it was due to lack of practice/experience with VR games. Both the movements’ velocity and mechanics can be addressed by changing the way the player moves; if we change from a smooth locomotion movement to a teleport movement, although this can break the realism of the game, this stops the motion sickness caused by in-game movement while the person’s body is standing still.

Finally, in the analysis of Q24 (

Table 8. Q24 frequencies, describing the likeliness of the tester to play the game once it has become a final product.

Q24
	N	%
Lightly Disagree	1	10.0%
Neutral	1	10.0%
Agree	3	30.0%
Strongly Agree	5	50.0%

), 80% of the testers would like to play the game again once it is fully developed, with 50% strongly agreeing and 30% agreeing, and only 10% were neutral, which brings us a sense of accomplishment and reason to continue developing the game. The last 10% that lightly disagreed with the statement said that since he does not suffer from the phobia, he does not feel the urge to play this game, for which the sole focus is the treatment of BII phobia.

The feedback gathered from these case study usability tests allowed us to have a more concrete view of the usability and the user experience of the game, which will allow us to conduct the necessary adjustments for the tests with the phobic patients so that they do not feel any unnecessary external factors, i.e., the motion sickness coming from the movement of the player in the virtual world or the speed at which the player moves, and also improve some gameplay aspects, such as the ability to interact with more objects than those which are already interactable, and highlight or emphasize the location of certain key objects in the virtual space so the player can be more perceptive of the blood stimuli next to these objects. These case study usability tests served as a basis to prepare future tests with BII phobia subjects, which will be conducted paired with the tests to validate the effectiveness of the game in triggering the phobia. Overall, we observed that the game was in a good testing phase that allowed the players to have the freedom to explore the mechanics while being connected to the ECG sensor, which gave us insights into the impact that this sensor has on the gameplay of the VR game.

6.3. Sensor Data

With the analysis of both .txt files and .json files of the data from each test, we gathered that every tester had a higher baseline heart rate than during the play session, with their values ranging from 1 to 4 beats per minute higher; everyone averaged 60 beats per minute during the play sessions (

Table 9. Heart rate values during the tests (BPM).

	T1	T2	T3	T4	T5	T6	T7	T8	T9	T10
Baseline (BPM)	62	61	61	63	62	64	63	61	64	62
Avg (BPM)	60.536	60.092	60.312	60.256	60.064	60.234	60.368	60.486	60.300	60.149

), which is the average heartbeat rate of a resting adult [65]. This discrepancy is due to the fact that the baseline measure is done as soon as the tester is equipped with the VR HMD and sensor when the body is still adjusting to the new sensory feedback from having the equipment attached and starting to experience VR for the first time in most of the cases. After that adjustment, during the play session, the testers started to feel less anxious about the novel experience, and the heartbeat rate stabilized to 60 heartbeats per minute on average.

We also observed that there were no sudden relevant increases in heart rate followed by decreases, so no observable syncope on any of the test subjects, which tells us that the testers felt comfortable during the play sessions and did not have any scare or anxiety during their playthrough.

One thing to note on these results from the sensor data—since the conducted tests were for usability testing, with only one tester suffering from the phobia, we were only expecting to see a stable heartbeat variance throughout the playtests and were not expecting to see differences in the heart rate measurements.

6.4. Preliminary Validation Preparation

For blood phobia triggering preliminary validation preparation, we performed one of the tests with a tester who suffers from the phobia and asked the psychology department from Minho University, Portugal, for their input on the blood stimuli across the game. The BII phobia tester’s test outcome was satisfactory; this tester’s phobia manifests when they see their own blood coming out of their body, which was not the objective of this game. One suggestion from the tester was to make a cutscene of sorts to make the in-game VR glove start to get covered in blood after the player grabs the broken wine glass, which will possibly suggest to the brain that their hand was cut by it. The other suggestion was that they were expecting more blood than they found, suggesting that every blood stimulus must be bigger and have more splashes to feel more realistic and trigger the phobia.

The feedback from the psychology department, to whom we sent a video containing every blood stimulus in the game, was that since they do not suffer from the phobia and do not have sensitivity towards blood, the blood stimuli did not trigger any reaction.

Further research with a larger number of subjects that suffer from BII phobia will be conducted.

7. Discussion

The favorable response to ‘Phobos’ in usability testing not only supports existing literature on VR’s efficacy in phobia treatment but also introduces new dimensions to consider. Our study’s emphasis on user engagement and interactive design elements resonates with recent trends in gamified therapy, as highlighted by Freitas, J. et al. [1] and Krzystanek, M. et al. [43] in their reviews of phobia treatments, suggesting that the gamification of therapeutic processes can significantly enhance patient motivation and participation for patients who are unable to participate or reticent about participating in in vivo exposure therapy.

The integration of biometric sensors, while challenging, has unveiled a critical area of research. This aspect diverges from the traditional focus of VR literature, which primarily addresses visual and auditory immersion. Our study’s approach to incorporating physiological feedback in real time adds a novel dimension to VR therapy, aligning with the vision described by Houzangbe, S. et al. [62] on sensor integration in VR applications.

The increased occurrence of motion sickness in our study presents a significant deviation from our expectations and the current trends in VR technology. This contrasts with the advancements suggested in the literature, indicating a gap between theoretical advancements in VR hardware and user experiences in practical applications. The high incidence of motion sickness, as reported by participants, underscores the necessity for further refinement in VR design, particularly in the context of therapeutic applications.

This finding challenges the assumption that modern VR systems universally reduce discomfort. It aligns with the concerns raised by Albakri, G. et al. [48] and Gromer, D. et al. [45] about the persistent issue of motion sickness in certain VR environments. Our study, therefore, contributes to a growing body of evidence that suggests a need for a deeper understanding of the factors contributing to motion sickness in VR, especially when used for therapeutic purposes.

Given this unexpected outcome, future iterations of ‘Phobos’ must prioritize addressing this issue. Potential avenues for exploration could include adjusting the game’s motion dynamics, exploring different VR hardware, or integrating advanced user customization features to mitigate these effects. Additionally, further research should also explore the relationship between motion sickness and therapeutic efficacy, as this represents a critical factor in the viability of VR-based treatments.

In addressing the limitations of our study, the need for a more diverse participant pool becomes evident. While our findings are promising, they are based on a limited demographic, which may not fully represent the broader population suffering from BII phobia. Future studies should aim to include a wider range of ages, ethnicities, and backgrounds to ascertain the effectiveness of ‘Phobos’ across diverse groups, as inclusivity in treatment efficacy is a key area of interest in contemporary medical research.

Finally, while our study provides immediate insights into the benefits of VR-based therapy, the long-term impacts remain unexplored. Sustained research efforts, as suggested by our literature review, are necessary to establish the durability of therapeutic outcomes achieved through VR. Longitudinal studies could provide invaluable data on the lasting effects of VR interventions, potentially leading to a paradigm shift in how phobias are treated.

8. Conclusions and Future Work

In this article, we elaborated on phobias and their treatment methods, serious games, game design for serious games, virtual reality exposure therapy, and biometric sensors used in games. We have detailed the development of the game Phobos. The Phobos serious game development section had a brief introduction to the game, detailed the game design process, programming and virtual reality development process and decisions, and a detailed explanation of the console application created for the ECG sensor while also detailing the positioning of the sensor, the data structure, and data acquisition process. In the subsequent chapter, we elaborated on the usability questionnaire and the case study results gathered throughout the testing process.

We were able to create a stable, testable version of “Phobos”, the serious game designed for BII phobia treatment by VRET, where we had to upgrade the game engine and redevelop many of the game’s systems, taking into account the psychology department design suggestions and, thus, addressing the main objective of exploring serious games as a support tool for the treatment of BII phobia.

Answering our first research question, “Is smooth locomotion a feasible method in a virtual reality serious game designed for blood phobia treatment and does it lead to motion sickness among players?”, through the case study usability tests that were conducted regarding motion sickness, we concluded that having smooth locomotion as a locomotion system is not a feasible locomotion system for a phobia VRET serious game, as it provokes motion sickness, which is unadvised since it is an external factor that will cause avoidance of playing the serious game. In our study addressing the research question, “What is the adequate positioning for a BITalino ECG sensor to minimize its impact on a player’s performance while playing a virtual reality serious game designed for blood phobia treatment?”, we developed a console application to interface with a BITalino sensor-kit equipped with an ECG sensor. This setup enabled us to collect biometric data, specifically heart rate measurements, by utilizing the Einthoven triangle formation for electrode placement. Our findings indicate that this specific positioning of the ECG sensor, which employs dry electrodes, is non-invasive and does not hinder the player’s performance during gameplay. The case study usability tests further substantiated that players were largely unaware of the sensor during the game, demonstrating that this electrode configuration effectively integrates into the gaming experience without adversely affecting player engagement or performance. Addressing the research question, “What is the perception impact of photorealistic graphics on the level of immersion experienced by players in a serious VR game?”, our findings from questions 15–16 suggest that the high-quality, realistic graphics significantly enhanced player immersion. This enhancement was evident as players reported losing awareness of their real-world surroundings while engaged in the game. This level of immersion suggests that photorealistic graphics play a crucial role in creating a deeply engaging and immersive virtual reality experience.

As a limitation of the developed study, although our main contributions from this work are the development process and usability testing of a serious game intended for the treatment of hemophobia as well as sensor placement invasiveness, this study was done with a small-sized sample of people who do not suffer from BII phobia as preparation for a test phase on a larger scale with subjects who do suffer from the phobia. Additionally, the different exposures to phobic stimuli and the approach need to be tested and validated by a greater number of phobic subjects.

In future work, we will consider increasing the number of testers with hemophobia as well as further developing the game by integrating AI capabilities for game-responsive adaptation. The possibility of implementing Neural Radiance Field (NeRF) 3D models to replace the apartment models, thus improving the immersion and realism of the game, is being taken into consideration. Another interesting application would be the development of an extended reality (XR) version of the game, where the phobic patients would see the real world as is with the camera of the XR device and then use blood stimuli models to simulate bleeding on the hands or on some specific objects. This XR version would eliminate the motion sickness issue because the patients’ movements would be their own and not a simulated version of their movement in a virtual world; it would also enable testing on a broader range of subjects since the development could encompass a larger range of devices.