NEUROWOODS: EMOTION-ADAPTIVE VIRTUAL ENVIRONMENTS FOR ENHANCING EMOTION REGULATION A Thesis Presented to the Faculty of the Graduate School of Cornell University In Partial Fulfillment of the Requirements for the Degree of Master of Science by VANESSA(SIHUI) YANG May 2025 © 2025 VANESSA YANG - 2 - ABSTRACT This thesis examines the effectiveness of real-time EEG-driven feedback in a virtual forest environment designed to enhance emotional regulation and cognitive engagement. The NeuroWoods system integrates alpha wave-based feedback mechanisms into an immersive VR experience, enabling dynamic environmental adjustments in response to users' neurophysiological states. By comparing EEG-contingent feedback, randomized feedback, and non-feedback conditions, the study evaluates the impact of biologically contingent interaction on emotional outcomes, cognitive focus, and behavioral engagement within the environment. EEG data, self-reported emotional states, and behavioral metrics were collected and analyzed to assess the system’s influence on users’ relaxation, focus, and sense of agency. Findings demonstrate that real-time neurofeedback fosters greater increases in alpha relative power, positive emotional shifts, and mindful interaction patterns compared to control conditions. This work contributes to the growing body of research on affective computing and EEG-VR integration, offering insights for the development of emotionally adaptive digital environments that support psychological resilience and cognitive health. - 3 - BIOGRAPHICAL SKETCH Vanessa was born in Zhejiang Province, China. She earned her bachelor’s degree in environmental design in Canada. This experience ignited her passion for enhancing human well-being. She is currently advancing her studies at Cornell University as a Master of Science candidate in Matter Design Computation. Her research primarily explores how neurotechnology, cognitive neuroscience, and environmental psychology can be applied within digital environments to improve emotional and cognitive health. Vanessa is committed to using accessible digital tools to foster well-being. - 4 - DEDICATION To my parents, professors and colleagues for supporting me through times of challenges. - 5 - ACKNOWLEDGMENTS I am deeply grateful to Professor Jenny Sabin for founding this impactful project, providing me with the freedom to explore and research at the intersection of technology and design within our department. Her constant encouragement has been crucial in reinforcing my dedication to this field. I would like to thank my thesis chair, Professor Farzin Lotfi-Jam, for his invaluable guidance and support throughout my thesis process. His mentorship helped me navigate challenges and maintain my passion for this work. Special thanks to Professor Saleh Kalantari, my advisor and the director of LabDAIL, for his sustained guidance and the opportunity to engage with advanced VR projects that have significantly shaped my academic and professional growth. I also appreciate Muhaimin Sarker for his technical assistance and Armin Mostafavi, our lab manager, for his ongoing support and encouragement. Thank you all for your contributions to my journey at Cornell. - 6 - TABLE OF CONTENTS CHAPTER 1 ................................................................................................................................... 9 1.1 Background and Problem Statement ................................................................................. - 9 - CHAPTER 2 ............................................................................................................................. - 12 - LITERATURE REVIEW ...................................................................................................... - 12 - 2.1 Overview of EEG and its Applications in Emotional and Cognitive Studies ............. - 12 - 2.2 Virtual Reality as a Therapeutic Tool for Emotional and Cognitive Enhancement .... - 16 - 2.3 Integration of EEG and VR for Cognitive and Emotional Enhancement .................... - 21 - 2.4 Summary and Research Gaps ...................................................................................... - 24 - 2.5 Conceptual Framework for NeuroWoods ....................................................................... - 25 - 2.5.1 Theoretical Foundations ........................................................................................... - 26 - 2.5.2 Integrated Mechanism of Action .............................................................................. - 27 - CHAPTER 3 ............................................................................................................................. - 29 - METHODOLOGY ................................................................................................................ - 29 - 3.1 Research Design .......................................................................................................... - 29 - 3.2 Participants .................................................................................................................. - 29 - 3.3 Apparatus and Materials .............................................................................................. - 30 - 3.4 Measures ...................................................................................................................... - 32 - 3.5 Experimental Procedure (Expanded) .............................................................................. - 33 - 3.5.1 Pre-Intervention Phase .............................................................................................. - 33 - 3.5.2 VR Orientation and Setup......................................................................................... - 34 - 3.5.3 Experimental Intervention ........................................................................................ - 35 - 3.5.4 Post-Intervention Phase ............................................................................................ - 37 - 3.5.5 Data Processing and Analysis................................................................................... - 39 - 3.6 Game Environment Documentation ................................................................................. - 42 - 3.6.1 Game Environment Screenshots and Visual Elements ............................................. - 43 - 3.6.2 Core Environmental Features ................................................................................... - 43 - 3.6.3 Environmental Response Parameters ....................................................................... - 46 - CHAPTER 4: DATA ANALYSIS AND RESULTS ............................................................... - 48 - - 7 - 4.1 Overview of Data Analysis ............................................................................................. - 48 - 4.2 Participant Demographics ............................................................................................... - 49 - 4.3 EEG Data Analysis.......................................................................................................... - 50 - 4.3.1 Data Preprocessing ................................................................................................... - 50 - 4.3.2 Alpha Relative Power Analysis ................................................................................ - 50 - 4.3.3 Alpha Wave Temporal Dynamics ............................................................................ - 52 - 4.4 Emotional State Analysis ................................................................................................ - 54 - 4.4.1 PANAS Scores ......................................................................................................... - 54 - 4.4.2 Factor Analysis of Emotional States ........................................................................ - 55 - 4.5 Behavioral Metrics Analysis ........................................................................................... - 56 - 4.5.1 Movement Patterns ................................................................................................... - 56 - 4.5.2 Environmental Interactions....................................................................................... - 57 - 4.5.3 Correlations Between Behavioral Metrics and EEG Measures ................................ - 58 - 4.6 Qualitative Analysis ........................................................................................................ - 60 - 4.6.1 Thematic Analysis of Interview Data ....................................................................... - 60 - 4.6.2 Condition Differences in Thematic Content ............................................................. - 60 - 4.6.3 Representative Quotations ........................................................................................ - 61 - 4.7 Integration of Quantitative and Qualitative Results ........................................................ - 62 - 4.8 Summary of Findings ...................................................................................................... - 63 - CHAPTER 5: DISCUSSION AND CONCLUSION ............................................................... - 65 - 5.1 Overview of Key Findings .............................................................................................. - 65 - 5.2 Integration with Existing Literature ................................................................................ - 65 - 5.2.1 Neurofeedback and Alpha Wave Training ............................................................... - 65 - 5.2.2 Virtual Reality and Emotional Regulation ............................................................... - 66 - 5.2.3 Integration of EEG and VR ...................................................................................... - 68 - 5.3 Theoretical Implications .................................................................................................. - 68 - 5.4 Practical Implications ...................................................................................................... - 70 - 5.5 Limitations and Future Directions................................................................................... - 72 - 5.6 Conclusion ....................................................................................................................... - 73 - CHAPTER 6: CONCLUSION AND RECOMMENDATIONS .............................................. - 75 - 6.1 Conclusion ....................................................................................................................... - 75 - - 8 - 6.2 Practical Recommendations ............................................................................................ - 77 - 6.2.1 Implementation in Urban Wellness Programs .......................................................... - 77 - 6.2.2 Clinical Applications ................................................................................................ - 78 - 6.2.3 Design Considerations for Future Systems .............................................................. - 79 - 6.2.4 Educational Applications .......................................................................................... - 79 - 6.3 Research Recommendations ........................................................................................... - 80 - 6.3.1 Longitudinal Studies ................................................................................................. - 80 - 6.3.2 Population-Specific Research ................................................................................... - 81 - 6.3.3 Technical Enhancements .......................................................................................... - 81 - 6.3.4 Comparative Effectiveness Research ....................................................................... - 82 - 6.4 Final Thoughts................................................................................................................. - 83 - REFERENCES ......................................................................................................................... - 85 - - 9 - CHAPTER 1 INTRODUCTION 1.1 Background and Problem Statement Consequently, over 58 percent of highly stressed workers live in urbanized areas (WHO). A study shows that dense urban conditions, such as noise, crowding and other stressors can elevate levels of cortisol, a stress hormone that contributes to chronic stress (Evans & Cohen, 2010). People living in urban areas have 21% higher risk of anxiety disorders and 39% higher risk of mood disorders compared to the populations living in the rural areas (Peen et al. 2010). Aside from that, access to natural environments comes with a 40 percent rise in reported stress level (White et al., 2019). Typically, mindfulness mediation approaches resolve these challenges through their effect in improving psychological wellbeing by means of stress reduction and the control of emotions (Valim et al., 2019; Wheeler et al., 2017). However, these practices present barriers to adoption; ten to twenty hours of practice and instruction that must occur over multiple sessions of 10 to 20 minutes, several times per week. The abstract nature of Meditation Guidance, without objective feedback, also leaves one uncertain of progress and may hence result in discouragement to continue. While prior studies have shown the benefits of EEG neurofeedback or VR separately in supporting emotion regulation, very few have explored how real-time alpha-based neurofeedback within immersive environments affects both emotional and cognitive responses. Moreover, the interaction - 10 - between feedback mechanisms—especially the distinction between biologically contingent vs. randomized feedback—remains underexplored. This study addresses that gap. The NeuroWoods system represents a novel approach to addressing these challenges using brain computer interface (BCI) technologies merged with immersive virtual reality (IVR) to facilitate emotional regulation and cognitive resilience. By using objective measurement of mental states via electroencephalography (EEG), visual abstraction of such concepts as relaxation and attentional focus and delivering these traits in a virtual, accessible format, this virtual environment uses EEG driven real time biofeedback mechanisms to address the limitations of traditional approaches. Feedback loops are implemented in real-time within an immersive forest environment to dynamically adapt to the users' EEG data selectively focusing on the alpha wave patterns of the relaxed alertness. When the users' alpha wave activity increases, the virtual environment responds in turn by changing some of the environmental elements such as ambient sounds, wildlife activity, lighting condition, and water features. Through these contingent responses, there is a bridge between internal mental states and external environmental changes including greater knowledge and control of emotional regulation processes. NeuroWoods may offer an easy way into mindfulness practices, providing a fun, objective measure of reducing stress and emotional resilience training, it may serve as an accessible entry point to mindfulness practice for those that don’t understand why you meditate, why you should meditate, or who are resistant to attempting beneficial meditation practices in the first place. 1.2 Research Questions - 11 - This study aims to investigate the effectiveness of the NeuroWoods system through three primary research questions: 1. How effective is real-time EEG neurofeedback in a virtual forest environment at enhancing emotional regulation compared to control conditions (random feedback and no feedback)? 2. What are the measurable differences in emotional state changes (pre/post-intervention) between participants receiving genuine EEG feedback versus those in control conditions? 3. To what extent do behavioural metrics within the virtual environment (such as movement patterns and interaction engagement) correlate with improvements in emotional regulation as measured by EEG alpha wave patterns? 1.3 Significance of the Study More specifically, this study evaluates what effect the integration of multiple sensory modalities in a VR environment, such as spatial design, auditory feedback, and interactive affordances, does control over emotional states, on the one hand, and cognitive engagement, on the other. The system is based on absolute time adaptability according to the neurophysiological responses of users in order to provide a novel approach to stress reduction and emotional resilience training. More specifically, the authors have conducted experiments to assess the effect of EEG-based feedback on users' emotional regulation and cognitive control by using three different experimental conditions. These insights have therapeutical, cognitive training and future BCI-VR integration implications for designing advanced neuroadaptive environments to improve emotional and cognitive health, particularly for people living in urban environments with limited access to natural settings. - 12 - CHAPTER 2 LITERATURE REVIEW 2.1 Overview of EEG and its Applications in Emotional and Cognitive Studies Electroencephalography (EEG) is a non-invasive technique of recording electrical activity generated by neuronal oscillations in the brain. EEG collecting live brain functioning data by placing electrodes on the scalp is essential in clinical and research applications. A high temporal resolution allows for the analysis of NBP processes with millisecond resolution, and it can explore rapid neural processes and understand cognitive and emotional functions at similar temporal rates (Sack et al., 2020). 2.1.1 Fundamentals of EEG Technology EEG primarily records postsynaptic potentials of the cortical neurons. Neurons communicate by generating ionic current flows, and voltage fluctuations can be identified on the scalp surface. However, not all neural activity is the same when being brought to life by EEG signals. EEG electrical activity monitors a small portion of the electrical activity of cortical neurons very near the placed electrodes since deeper structures like the thalamus, hippocampus, and brain stem contributions are blocked by signal attenuation through brain tissue, cerebrospinal fluid, skull and scalp (Nunez & Srinivasan, 2006). According to the International 10-20 electrode system, conventional scalp EEG records by placing electrodes so that electrodes are placed consistently across subjects and studies. There is one input for each differential amplifier, and a standard reference electrode is tied to the other input. Usually, - 13 - these amplifiers will amplify the voltage difference between the active electrode and reference by a factor of 1,000 to 100,000 (60 to 100 dB of power gain). Almost all modern EEG systems are digital, with analogue signals converted into digital format by anti-aliasing filters. In clinical applications, sampling rates usually fall between 256–512Hz, whereas in research applications, the rate may reach up to 20kHz to record fast oscillatory activity with greater precision (Luck, 2014). Conventional EEG signal analysis is done in frequency bands correlated with mental states and cognitive processes. In a deep sleep, the most dominant are delta waves (0.5 to 4 Hz) associated with restorative processes, and abnormal delta activity during wakefulness indicates brain injury or developmental disorder. Theta waves (4 to 8 Hz), known to arise during drowsiness, light sleep and meditative states, are crucial for memory consolidation, emotional processing, and creative thinking (Klimesch, 1999). In relaxed wakefulness, eyes closed, especially alpha waves (8-13 Hz) dominate, are negatively related to cortical excitability and have been related to internal attention, inhibitory processes, and relaxation states (Bazanova & Vernon, 2014). Beta waves (13 to 30 Hz) are often associated with motor activity, awake dreaming and active thinking, among other things. One example is that beta waves increase when performing concentration and problem-solving tasks. Gamma waves (30–100 Hz) are assigned to higher-order cognitive processes, including perception, attention and consciousness; gamma oscillations are presumed modulated by the binding of sensory information across distributed neural networks (Başar et al., 2001). Knowledge of these fundamental aspects of EEG technology allows one to understand virtual EEG data in terms of emotional and cognitive processes, which is important for interpreting neurofeedback data and brain-computer interface research. - 14 - 2.1.2 EEG in Emotion and Cognitive Functioning EEG has been widely used for studying neural correlations of emotional processing and regulation. An area of application used very prominently is for analyzing event-related potentials (ERPs) since they are time-locked EEG responses to a specific stimulus (or combination of stimuli). For instance, the late positive potential (LPP) is an ERP component linked to emotion and emotion regulation. Research has shown that the LPP changes with children’s emotion regulation abilities even as early as the age of five, suggesting that it might be a neural marker of emotional processing development (Dennis & Solomon, 2010). Indeed, frontal EEG asymmetry, especially in the alpha band, has been proven to be a reliable index of emotional processing and affective traits. It is well known that Davidson's influential work showed that asymmetrical frontal activity correlates with emotional valence and motivational direction (Davidson, 1992; Davidson & Irwin, 1999). Approach-related emotions: positive effect tends to be more left frontally, and withdrawal-related emotions, negative affect, more right frontally. It has been observed in studies concerning mood disorders and, therefore, has been seen to relate to affective dysfunction (Coan & Allen, 2004). Relaxation and attentional focus with internal focus were assessed by the key indicators of the relative power of alpha waves over other frequency bands. According to Bazanova and Vernon (2014), the normalised measure of alpha predominance in overall brain activity is alpha_relative (alpha_absolute power normalised by the sum of the absolute powers of the various frequency bands). A high alpha_relative value indicates a mentally relaxed but attentive state where the attention is not externally directed, the very state focused on by many mindfulness and meditation practices. - 15 - EEG has tremendously advanced the study of cognitive processes such as attention, memory and executive functions in cognitive neuroscience. ERPs widely studied in attention and stimulus evaluation are the P300 component elicited around 300 milliseconds after stimulus presentation. The amplitude and latency are used as appliances of cognitive processing efficiency and have been used in brain-computer interface development (Farwell & Donchin, 1988). The basic steps of information processing in the brain would be thought of as EEG microstates— brief periods during which the scalp electrical potential field is stable—proposed as the “atoms” of the limbic system, including the cerebral cortex. Such microstates clearly show lawful complex evolution with age, thus indicating important cerebral maturation over different life stages (Koenig et al., 2002). Later, EEG integration with other neuroimaging techniques like functional magnetic resonance imaging (fMRI) further enabled us to understand cognitive functions using EEG's excellent temporal resolution and superior spatial resolution. Such multimodal approaches offer a more comprehensive view of the spatiotemporal dynamics of the neural mechanisms involved in cognition (Sack et al., 2020). 2.1.3 EEG in Neurofeedback Applications Biofeedback is a specialized application that teaches how to regulate the body's action through real-time displays of body activity, and neurofeedback is a type of biofeedback that uses accurate time displays of brain activity to teach self-regulation of the brain. EEG-based neurofeedback as a potent tool to increase cognitive performance and bolster emotional regulation has received much attention in clinical and non-clinical populations (Sitaram et al., 2017). - 16 - Multiple studies show that this can be accomplished with neurofeedback training. It has been demonstrated by Wang and Hsieh (2013) that neurofeedback training on specific EEG frequency improved attention and working memory performance in healthy adults. Viviani and Vallesi (2021) also conducted a systematic review confirming that EEG-neurofeedback can also reduce executive function deficits in healthy adults, keeping with the caveat that methodological rigours must be considered when designing the study. Rogala et al. (2016) conducted a comprehensive review of controlled neurofeedback studies with healthy adults, featuring notable successes as well as methodological pitfalls in the field. The authors stressed that the control condition is such an important issue that the three-group experimental design of the current NeuroWoods study directly addresses it. Neurofeedback mechanisms of successful neurofeedback are based on operant conditioning and utilised by providing reinforcement when sitting there, trying to transfer the brain equilibration towards the wanted neuro states. Feedback must also be contingent, immediate and interpretable for learning and control (Sitaram et al., 2017). The design of the NeuroWoods system, which immediately gives visual and environmental feedback based on real-time EEG measurements, is directly based on these principles. Specifically, there is some evidence that alpha-wave neurofeedback can improve relaxation and attentional control. Training for increased alpha power has been linked to decreased anxiety, improved cognitive performance and increased creative thinking (Gruzelier, 2014). This indicates the potential for alpha-based neurofeedback, as defined in the NeuroWoods system, to promote emotional and cognitive well-being. 2.2 Virtual Reality as a Therapeutic Tool for Emotional and Cognitive Enhancement - 17 - Virtual Reality (VR) has emerged as a promising therapeutic modality that provides a powerful, interactive, immersive environment for therapeutic purposes. VR’s evolution as a technology has transformed this set of tools from a trial to a commonly used tool in a host of psychological and neurological conditions. 2.2.1 Evolution of VR in Healthcare In the mid-20th Century, Morton Heilig introduced the 'Sensorama' in the 1960s, an early multi- sensory immersive experience (Mandal, 2013). Practical therapeutic applications, however, did not thrive until decades later as high computing power and display technologies matured. This was the point where head-mounted displays (HMDs) enabling truly immersive experiences entered the therapeutic VR arena in the early 1990s. VRET was first pioneered by Rothbaum et al. (1995), who successfully treated acrophobia (fear of heights) with VR environments. In this first, we achieved a breakthrough and proved that VR can be safely and efficiently used in clinical settings to simulate anxiety-provoking situations. VR applications did increase significantly in psychological treatments throughout the late 1990s and early 2000s in the treatment of several anxiety disorders, including posttraumatic stress disorder (PTSD), specific phobias, and social anxiety. Second, VF works, as accurately demonstrated by Difede and Hoffman (2002), in an impressive VR application of the exposure therapy for PTSD-related symptoms of people struck by the 9/11 attacks. During the subsequent decade, VR became a part of cognitive and physical rehabilitation. Rizzo et al. (2004) developed VR-based cognitive rehabilitation programs for people with traumatic brain injuries, using interactive environments to promote executive functioning and cognitive recovery. - 18 - Holden (2005) also found that VR performance was superior or at least equal concerning patient engagement and functional improvement outcomes compared to traditional therapies used in motor rehabilitation for stroke survivors. Recent technological advances, including wireless HMDs, high-fidelity graphics, and complex tracking systems, have greatly expanded the healthcare applications of VR. The technology available today allows for not only a more realistic environment but also integration with physiological monitoring and building closed-loop systems where the environment changes in real time to accommodate the physiological state of the user (Parsons & Rizzo, 2008; Wiederhold & Wiederhold, 2008). 2.2.2 VR in Emotional Regulation and Therapy In terms of emotional regulation training and therapy, virtual reality has particularly proven its worth as it enables emotional engagement in a controlled situation. VR exposure therapy was proven effective for anxiety disorders through a meta-analysis by Powers and Emmelkamp (2008), and the results from VR exposure are supposedly equivalent to or even better than traditional methods of in vivo exposure for anxiety disorders. Regardless of the anxiety condition, VRET has proven effective. Rothbaum et al. (2006) also confirmed VR efficacy for treating fear of flying and found benefits at a one-year follow-up; Garcia Palacios et al. (2002) showed that 83% of people with arachnophobia who undertook VRET also experienced clinically significant improvement after VRET. VR allows for the precise recreation of traumatic scenarios that might be difficult or even impossible to recreate in real life for post-traumatic stress disorder. According to Rizzo and - 19 - colleagues (2009), "Virtual Iraq/Afghanistan" is a VR system specifically for veterans with combat-related PTSD. The studies revealed a massive reduction in PTSD symptoms, and some participants no longer met diagnostic criteria after treatment. VR has also been shown to work as a treatment for depression and improve emotional health beyond anxiety disorders. Falconer et al. (2016) have shown improved positive effects and reduced self-criticism amongst depressed participants taking part in VR-based compassion-focused interventions. Thus, these findings are consistent with Freeman et al.'s (2017) research that brief VR sessions reduced anxiety and increased emotional well-being. Several key mechanisms support VR's therapeutic efficacy for emotional regulation. The first is that VR allows for emotional engagement while preserving safety and a feeling of control. Second, it gives you slow, systematic exposure to emotional stimuli under precise intensity. The third benefit is that VR can incorporate biofeedback elements, enabling users to visualise their physiological responses excitingly. Additionally, real-time adaptation of VR environments based on users' responses allows this VR environment to provide personalised therapeutic experiences (Colombo et al., 2021). Although VR could be employed for many other purposes, its capabilities make it especially well- suited for emotional regulation training based on the NeuroWoods system. NeuroWoods uses VR’s emotional engagement potential by creating an immersive natural environment that responds to users’ EEG-measured mental states and delivers concrete feedback on abstract mental processes. 2.2.3 Nature-Based VR for Wellbeing - 20 - One particularly relevant application of VR for emotional well-being is the simulation of natural environments. Research has established that exposure to natural settings contributes to significant assessment of psychological benefits, including psychophysiological stress reduction, mood improvement, and cognitive function (Kaplan, 1995; Ulrich et al., 1991). The theoretical basis for these findings provides the foundation for "nature therapy" or 'ecotherapy' methods, which seek to use natural settings for psychological interventions. In her paper, Johnsen (2011) proposed a conceptual framework for how natural environments regulate emotion through their role in automatic attention restoration, stress relief, and adaptive emotional reactions. This framework offers an explanation as to why natural exposure—virtual or real—can be as useful in engaging emotional regulation processes. It has recently been confirmed that virtual nature experiences bring about similar psychological benefits as visiting the natural environment. In a systematic review, browning et al. (2020) conclude that VR nature experiences reduce physiological stress measures and self-reported mood. Likewise, White et al. (2018) have found that natural settings 360-degree videos reduce stress and anxiety in laboratory settings. Theodorou et al. (2023) also indicated that virtual nature environments can help in emotion regulation when combined with cognitive reappraisal techniques. The NeuroWoods design considers that virtual nature may be a suitable utility for practicing emotional regulation skills, which is one of the key considerations of the design. More specifically, it offers auspicious opportunities for emotional regulation training by the augmentation of nature-based VR with biofeedback mechanisms. Nature-based VR and HRV biofeedback together have been found by Rockstroh et al. (2019) to be better relaxing than either - 21 - component in isolation. The synergistic effect is in line with the NeuroWoods approach to combine nature-based VR and EEG neurofeedback to support the enhancement of emotional regulation. 2.3 Integration of EEG and VR for Cognitive and Emotional Enhancement EEG monitoring in the context of VR environments is an integrated novel approach with high potential for both therapeutic application and cognitive enhancement. The combination of these together creates closed loop systems which can adapt in real time to the user’s neural states for personalized, engaging and self-regulation friendly experiences. 2.3.1 Technical Foundations of EEG-VR Integration Several technical issues must be resolved for effective EEG-VR integration to be implemented. Due to potential artifacts from head motions, electromagnetic interference from the VR equipment, and practical difficulties of wearing EEG electrodes and VR headsets at the same time, VR is unwelcome means of acquisition in EEG recording. However, these challenges have been only partially addressed by wireless EEG systems, dry electrodes which do not require conductive gel, and sleek headsets intended for fitting with VR equipment. Consumer grade EEG headbands such as the Muse from InteraXon Inc. make for promising EEG-VR applications due to their reliability of measurements of key frequency bands at frontal and temporal parietal locations (Krigolson et al., 2017). - 22 - Figure 1 Muse 2 headset Reference: FPz However, EEG data, if processed in real time with minimal latency, has to be done for a meaningful integration with the VR environment, it should provide immediate feedback within the VR. Typically, this involves pre-processing steps to remove artifacts, extract relevant frequency band power (e.g. alpha relative), and translate these measurements in the related environmental responses in the virtual environment (Vourvopoulos & Bermúdez i Badia, 2016). 2.3.2 Empirical Evidence for EEG-VR Efficacy Combined EEG-VR interventions are demonstrated to enhance cognitive and emotional states across a large body of research. Choe et al. (2002) achieved improved attentional performance in adults when EEG based neurofeedback was provided through VR. In the case of the current NeuroWoods study with its three groups, greater actual EEG feedback led to significantly improved performance, a direct result for this design. - 23 - In fact, the research on electrophysiology has further validated that immersive VR can sensibly induce the respective mental states as the targets of neurofeedback training. According to Tarrant and Cope(2018), VR guided meditation resulted in a greater increase of alpha and theta EEG activity than normal guided meditation, supporting the assumption that VR may induce respiratory to neurofeedback interventions. Finally, the combined EEG-VR approach has been shown promising for the regulation of emotions in general. For healthy adults, EEG neurofeedback through a relaxing VR environment was found to effectively decrease anxiety in healthy adults, found Rodríguez et al., 2015. Similarly, Berger and Davelaar (2018) showed that alpha neurofeedback in VR was capable of increasing emotional regulation capacity and decreasing stress reactivity. Importantly, several studies did directly compare genuine EEG feedback to sham feedback conditions and have found significantly greater benefits for genuine feedback, in line with the NeuroWoods study methodology. Marzbani et al. (2016) reviewed neurofeedback applications in anxiety management in which they confirmed the importance of genuine contingent feedback in order to achieve therapeutic outcome. 2.3.3 Applications for Stress Reduction and Urban Wellbeing Particular promise is held by EEG-VR techniques in integrating them to solve stress related challenges in urban but also other environments. It has been previously found, as stated in Chapter 1, that urbanization is associated with increased risk for stress disorders and reduced exposure to natural environments capable of regulating emotions (Peen et al., 2010; White et al., 2019). - 24 - An accessible alternative for people living in the city might include virtual nature experiences with neurofeedback. Brief VR nature exposure significantly reduces stress in urban office workers, and such effects were equivalent to that of real short term nature exposure, according to Anderson et al. (2017). These interventions get even stronger when enhanced with biofeedback elements (Rockstroh, et al., 2019). The NeuroWoods approach contributes specifically to the requirement of brief, effective interventions that can be included into urban lifestyles. Many traditional mindfulness practices involve lengthy time commitments (15-20 minutes a day for 3 to 4 weeks or more) and thereby may be less attainable due to constraints placed on busy urban populations. In contrast, compared to EEG-VR interventions, these may further facilitate learning by increased engagement and concrete feedback allowing for fewer sessions yet achieving meaningful benefits (Campos & Lhullier, 2020). 2.4 Summary and Research Gaps Based on the theoretical foundations and empirical evidence supporting EEG based neurofeedback, therapeutic applications of VR, as well as their combination for cognitive and emotional enhancement, this literature review has been carried out. That leads to several key insights about the current NeuroWoods study. EEG based neurofeedback is also effective for training self-regulation of brain activity, alpha wave training in particular for relaxation and attentional control. Second, VR is an immersive and controlled environment which allows the user to become emotionally engaged while maintaining safety, which is an ideal environment for Emotional Regulation training. Third, virtual nature experiences provide psychological benefits as great as engaging with natural environments, - 25 - suggesting a possible but neglected ways to enhance wellbeing in urban populations. Fourth, by integrating EEG neurofeedback with VR, closed loop systems, which can adapt in real-time to changes in the user’s neuro state, are created, which may improve the efficacy of the training. However, several research gaps remain. However, both EEG neurofeedback and VR have strong evidence bases, and research on integration is just getting started but small scale evidence is beginning to emerge. Finally, further clarification is needed with respect to the specific mechanisms whereby EEG-VR interventions impact emotional regulation. In addition, the third point is that the optimal parameters of such interventions (such as session duration, feedback mechanisms and environmental design) are not well explored. By using a controlled three group design where EEG feedback is compared to random feedback, and no feedback we are addressing these gaps in the current NeuroWoods study. The rationale for this approach is to make it clear what is the influence of contingent EEG feedback, other than that which is due to VR nature exposure more generally. The study will engage in further elucidation of EEG VR Intervention Efficacy by examining both subjective emotional changes and objective behavioral metrics from within the virtual environment. Overall, this research strikes to investigate this in expanding very urbanized societies, maybe giving usable replacement choices for typical dullness practices intend to appropriately advance the emotional wellbeing of individuals who are searching for better emotional regulation aptitudes. 2.5 Conceptual Framework for NeuroWoods The NeuroWoods intervention is built upon an integrative theoretical framework that combines elements from neurofeedback theory, attention restoration theory, and the biophilia hypothesis. - 26 - This framework elucidates the mechanisms through which EEG-driven virtual nature experiences may enhance emotional regulation and cognitive functioning. 2.5.1 Theoretical Foundations Three primary theoretical pillars are combined to form the basis of understanding how EEG driven virtual environments could impact emotional regulation according to the NeuroWoods conceptual framework. The foundations of these three theories are the neurofeedback theory and the principles of operant conditioning. According to this theory, accurate neurofeedback can teach people to become able to regulate their brains (Sitaram et al., 2017). Under the theory, contingent, immediate and interpretable feedback promotes gradual mastery of particular neural process, in particular, when the feedback is intrinsically rewarding. The second theoretical foundation is Attention Restoration Theory (ART) (Kaplan 1995). According to ART, natural environments promote recovery from attentional fatigue because they convene involuntary attention through 'soft fascination', while maintaining free attention channels available for directed attention recovery. Natural setting, maximizing stimulation that provides no - 27 - excessive demands of attentional resources, provides perfect combination of stimulation supporting cognitive restoration, and thus, the conditions for relaxed and recovered mind. Wilson (1984) introduces the third pillar of Biophilia Hypothesis. The theory is that humans evolved to have an innate affinity with nature and natural processes. The hypothesis behind this is that exposure to natural or natural(e) environments elicit positive psychophysiological responses that contribute to wellbeing and stress reduction. An evolutionary perspective is helpful in understanding why nature based interventions may be the case particularly for emotional regulation and stress responses. 2.5.2 Integrated Mechanism of Action The theoretical basis of these NeuroWoods system is based on these theoretical foundations in a cyclic feedback mechanism that integrates more processes to support emotional regulation. EEG is used to continuously monitor the user’s alpha wave patterns, subsequently tracking alpha_relative power, as an act of index of relaxed attentional states. The adaptive responses in the virtual environment are based on this neurophysiological monitoring. The virtual environment is based on real time alpha measurement and adapts with blurred alterations of the visual and auditory elements towards a direct representation of the objective state of the user. These reactions by mobile devices are immediate feedback that help make users aware of their mental state and influence it. At the same time, ART was created by nature on the basis of soft fascination in the scientific sense of the word, which naturally helps restore attention in the virtual environment, lighting the cognitive load and conditions for relaxation simultaneously. - 28 - The increased awareness of the connection between the users mental state and changes in the environment is gained because of the operant conditioning mechanism that is described in the neurofeedback theory which describes the users’ ability to control their alpha wave activity over time. The naturalistic design of the environment adds to this process by triggering innate positive associations with natural settings and reducing stress responses in accord with the proposed biophilia hypothesis. Through multiple complementary pathways of direct neural feedback, attention restoration, and biophilic engagement, NeuroWoods provides this integrated framework that explains how emotional regulation is assisted by this technology. The role of the virtual setting as not only a medium but also as its own therapeutic environment results in synergy, which could be greater than either the medium or the environment could provide on their own. - 29 - CHAPTER 3 METHODOLOGY 3.1 Research Design In this study, a mixed methods experimental design was used to develop a neurobiological profile of the effectiveness of the NeuroWoods EEG driven adaptive virtual environment for emotional regulation. In a between subject design three experimental conditions used were (1) EEG driven condition with pseudo neurofeedback using real time neuro feedback, (2) Pseudo feedback condition with pseudo data used as fbk. and (3) Control condition no feedback environment was used. The design was such that we could evaluate the specific effects of contingent neurofeedback while holding constant general effects of exposure to virtual nature and perceived control. Qualitative and quantitative data (EEG measurements, behavioral metrics within the virtual environment, standardized emotional state assessment) were collected from the participant impressions and experiences. Taken together, this approach provided a comprehensive evaluation both with regard to objective physiological effects and subjective experiential outcomes related to the NeuroWoods intervention. 3.2 Participants Participants for this study were primarily recruited through posters displayed across the Cornell University campus and digital advertisements posted on university-affiliated websites and forums. Recruitment specifically targeted students and working professionals experiencing elevated levels of academic or occupational stress. Secondary recruitment strategies included email announcements sent via departmental listservs and relevant student organizations. All screening - 30 - processes, initial questionnaires, and consent forms were administered securely online. Eligible participants included healthy adults aged 18 to 45 with normal or corrected vision, without any history of seizures, severe motion sickness, or current psychiatric medication use. Individuals with more than three months of regular meditation experience were excluded from participation. A power analysis conducted with G*Power software used effect sizes from Marzbani et al. (2016) and Tarrant & Cope (2018) to decide participant sample sizes. For detecting medium effect sizes across primary outcome measures researchers needed a study sample of minimum 21 participants with seven participants in each condition using alpha 0.05 for a desired power of 0.80. The presented content was randomly distributed to each participant who received one of the three experimental conditions with equal participant numbers across all groups. 3.3 Apparatus and Materials 3.3.1 EEG Recording Equipment The Muse 2 headband (InteraXon Inc., a wireless consumer grade EEG device) with four channels laid at TP9, AF7, AF8, and TP10 and a reference at FPz was used to collect EEG data. The sampling rate is 256Hz and it comes equipped with in built notch filters at 50Hz and 60Hz to remove the power line interference. For EEG data, first, it was used to access raw EEG data to extract frequency band power information from first and second raw EEG data using Muse software development kit (SDK). Absolute alpha power (8-13Hz) was calculated relative to absolute power across all frequency bands (delta, theta, alpha, beta, gamma) at the electrode sites to get the alpha relative power. This normalized measure normalizes the alpha predominance for individual differences in skull - 31 - thickness and overall EEG amplitude, making this more reliable index of alpha predominance (Bazanova and Vernon 2014). 3.3.2 Virtual Reality Environment Specifically, the NeuroWoods virtual environment was created in the Unity game engine (version 2021.3) and presented in an Oculus Quest 2 headset. The forest depicted was in a computer- generated (CG) setting with a winding path, water features and an array of flora. The environmental audio is ambient forest sounds, for example birds, rustling leaves, and flowing water. Three experimental conditions were created to the environment and three versions were made. In Condition 1, the EEG adaptive version, real time environmental change was introduced according to participants' alpha relative power. The environment responded with: actors increasing the volume of ambient music, allowing more butterflies, birds, and warding off a growing presence of companions roaming the banks. With more activity, the sun became a brighter source of light, the colors permeated the scene with more vibrancy, and the water flow increased. They changed gradually with increasing measured alpha level along a continuum of increasing alpha levels, not as the binary events. We designed pseudo-adaptive version (Condition 2) which had identical environmental changes based on random fluctuations that modelled a typical alpha wave pattern without being essentially connected to the EEG data of the participant. - 32 - A control condition consisting of a static environment that did not change was also utilized to test the effect of VR nature exposure without the use of feedback elements, which was the non adaptive version (Condition 3). 3.4 Measures 3.4.1 Psychological Measures The Positive and Negative Affect Schedule (PANAS; Watson et al., 1988) was administered before and after the VR exposure to measure emotional states. This is a 20 item measure which has 10 positive, and 10 negative affect items rated on a 5 point Likert scale. In order to identify naturally occurring groupings of emotional states, factor analysis of pre-intervention responses was performed, using particular attention to items relating to alertness, relaxation, and anxiety. Furthermore, a custom post session questionnaire gave subjective experience of the virtual environment, perceived control on environmental elements and self reported engagement on a 7 point Likert scales. 3.4.2 Physiological and Behavioral Measures The primary EEG measures of outcome were mean alpha relative power and changes in alpha relative power during the session. Therefore, separate values were calculated for frontal (AF7, AF8) and temporal parietal (TP9, TP10) regions to distinguish between possible different patterns of activity. Total distance moved, movement speed, patterns of interactions with the environmental elements, head movement, and viewing patterns were automatically recorded as behavioral metrics within - 33 - the virtual environment. These data offered objective measures of engagement with virtual environment potential correlates of emotional and attentional states. 3.5 Experimental Procedure (Expanded) The experiment was conducted in a controlled laboratory environment with consistent lighting, temperature and acoustic condition in all sessions. The protocol was completed by each participant independent of one another in sessions of about 45 minutes. 3.5.1 Pre-Intervention Phase This was the beginning of the preintervention phase and prior to this informed consent and screening procedures were done. Participants who were able to attend the laboratory were asked to review and sign the consent forms and complete the screening to determine whether they could participate, and exclusion criteria were met. For the study, screeners for normal or correction to normal vision, no history of seizure disorder or severe motion sickness, and not taking any medication that would interfere with the EEG measurement, were included. After the consent process, the subjects were subject to some series of baseline assessments. Sequenced within them was an overall demographic questionnaire and their responses to age, gender, previous education level, and prior experiences in meditative, virtual reality, or neurofeedback technologies. In order to have a baseline emotional state before their VR experience, participants filled out the pre intervention assessment of the Positive and Negative Affect Schedule (PANAS). However, the positive and negative affect dimensions had it standardized so together they could be compared to post intervention scores, very strongly accounting for both of them. - 34 - After manufacturer guidelines for the Muse 2 headband, the EEG setup procedure was then carried out. The reference electrode was placed at FPz on the participant's forehead, the device was positioned carefully, across all the participants, on their forehead with this reference electrode. Through the system interface, impedance values were checked at all electrode sites (TP9, AF7, AF8, TP10) to ensure proper skin contact at the electrode sites. The Muse Monitor application was used to confirm data integrity and to assess signal quality before proceeding in subsequent steps. After which, resting state neural activity patterns were established for each subject through a 2- minute eyes closed baseline EEG recording in order to serve as the reference point to analyze the effects of intervention. 3.5.2 VR Orientation and Setup To allow the participants to successfully and comfortably participate in the VR environment while ensuring quality EEG signals, a VR orientation phase was created. Participants first underwent standardized instructions on preparing and setting up the Oculus Quest 2 headset as well as using the controllers. Instructions for correct headset positioning, interpupillary distance adjustment, and controller grip techniques to minimize motion artifacts of the EEG signal while allowing for natural interaction in the virtual environment were provided. According to the first instructions, the VR headset was carefully fitted over the Muse headband. This process was given special attention so that both devices will be safely placed in position without crossing in the way of one another. The headset straps caused minor adjustment as necessary until the headband of the EEG fit comfortably within the headset while remaining properly aligned for the optics of each participant. Before placing the VR headset, the confirmation - 35 - of EEG signal quality after the fitting process was made to ensure that electrode contact was not broken. All participants also first did a brief navigation training session in a neutral virtual environment that was already separate from the NeuroWoods forest before entering the experimental environment. Participants were made familiar with basic movement mechanics, turning and interaction with environmental elements as part of this training. In order to avoid priming effects prior to exposure to the experimental conditions, the training environment was made emotionally neutral. Once participants were comfortable with the VR interface, they received enough time until they demonstrated proficiency in making the necessary movements and interactions. 3.5.3 Experimental Intervention The experimental intervention phase used a between subject’s design with participants allocated; one of three distinct experimental conditions. First, real time via the Muse head band: the EEG adaptive environment provided feedback based on participant’s alpha relative power. The second condition was a pseudo adaptive environment where environmental changes were dependent on Figure 2 Muse 2 headset - 36 - prerecorded alpha patterns but was not related to the actual EEG measurements of the participant. Thus, the third condition was presented as a non-adaptive environment maintaining absence of environmental changes during the experience. Due to isolation of the specific effects of contingent neurofeedback from general effects of virtual nature exposure, this design enabled the specification of these factors. Expectancy effects were controlled by the same instructions for all participants, regardless of assigned condition. These standardized instructions were instructed to go back to the environment at their own pace, try to seek peace of mind, and finally, that the environment would respond as stimuli. Even with this final instruction, any observed differences could not be attributed to differences of expectations regarding the interactive nature of experience, because only Condition 1 had genuine EEG linked changes. For the NeuroWoods experience, participants were in the virtual environment for a set of 15 minutes, non-stop. For Condition 1, the environment provided feedback based on the user's EEG data. If the user's EEG data indicated a focused, relaxed mental state characterized by increased alpha wave activity, glowing orbs would appear within the environment. Users could manually collect these orbs, which served as positive reinforcement, and upon successful collection, directional arrows would emerge to guide them toward the bonfire destination. If EEG readings fell outside this relaxed threshold, the environment would instead provide natural ambient sounds to help users become more relaxed. In Condition 2, the same environmental changes took place but were formed from random fluctuations which had the same number of cycles as observed alpha wave patterns, instead of from the participant's actual neural activity. In Condition 3, all parameters in the environment were constant and no dynamic changes occurred in any elements. - 37 - Figure 3 Participant wearing the VR headset (KIject model) during the EEG-VR intervention. The approximately 15 minute session served as an opportunity to collect comprehensive data including continuous EEG on all channels, detailed recording of movement patterns like position, speed, and acceleration on the virtual environment, monitoring of head orientation and viewing patterns for attentional focus, and recording of all activity on environmental elements. The use of this multimodal data collection approach yielded rich physiological responses and behavioral patterns to access the neurological dynamics in terms of emotional states as well as engaging with the virtual environment. 3.5.4 Post-Intervention Phase - 38 - Immediately after the end of the VR experience, participants were followed through a post- intervention phase to collect their emotional states and subjective experiences during this period that the effects of the intervention continued to be salient. The first part of this research involved participants completing the PANAS post intervention assessment using the same instrument they would have to use in the pre intervention phase, permitting direct comparison of an onset of emotional states before and after the NeuroWoods experience. Next comes a custom experience questionnaire devised for this study to measure participants' subjective experience of the virtual environment, and their perceived control over the environmental elements, their levels of self reported engagement, and immersion with the virtual environment. The questionnaire contained items that colleges were asked to rate on 7 point Likert scales regarding dimensions such as relaxation, attentional focus, connection with nature, and awareness of mental states during the experience. Figure 4 Screenshot of post intervention questionnaire - 39 - The quantitative assessments were followed by a semi structured interview with each participant aimed at generating rich qualitative data pertaining to their experience. We interviewed participant in the intervention to explore participant’s subjective experience of relaxation and focus, participant’s perception of the relationship between participant’s mental state and any observations changes occurred during the environment in the NeuroWoods, and how participant overall feel using the NeuroWoods environment. The interviews contained interesting contextual information and insight into the phenomenological aspects of the intervention which might not be isolated through quantitative measures alone. Transcripts of all interview responses were audio recorded with participant consent and carried out a thematic analysis of the whole. The debriefing session was a final phase which ended the study with the participants being well informed about the actual contingencies of the condition they were assigned to and the rationale behind the whole study. Due to his/her belief that the environment would respond to the mental state and yet his her experiences with either pseudo feedback or no feedback, this debriefing was especially important to those participating in Conditions 2 and 3. This experimental control had to happen, as the debriefing explained, and then passed on the opportunity for the participants to ask questions around the study design and objectives. Participants were thanked for their participation and paid according to the protocol approved. 3.5.5 Data Processing and Analysis A comprehensive analytical method was then followed to process and interpret the multimodal dataset based on data collection. In the first step, the raw EEG data processing pipeline included preprocessing required to obtain signal quality and to remove artifacts Factor that would interfere with analysis. Further preprocessing consisted of automated artifact rejection to remove segments - 40 - with too much noise or movement artifact. To remove frequency components outside the range of interest for these processes, a band-pass filter (1-40 Hz) was applied. In order to reduce the effects of eye blinks, muscle activity, and other such non-neural artifacts, Independent Component Analysis (ICA) was then used to identify and remove components from the EEG data, leaving cleaned EEG data for subsequent analysis. The frequency analysis phase employed power spectral density analysis in order to extract absolute power values of the major frequency bands that were important to emotional and cognitive states. The frequency range included 0.5-4 Hz, 4-8 Hz, 8-13 Hz, 13-30 Hz and 30-40 Hz respectively. The calculation of alpha relative power (the ratio of absolute alpha power to total power across all frequency bands) was considered special. The resulting normalized measure was a more reliable index of alpha predominance that controls differences in individual skull thickness and overall EEG amplitude. Analyses were performed separately for frontal (AF7, AF8) and temporal-parietal (TP9, TP10) regions in order to evaluate whether alpha activity is characterized by regional differences and their correlation with emotional regulation processes. - 41 - Figure 5 Graphical representation of changes in alpha wave activity across the EEG adaptive, pseudo-adaptive, and non-adaptive conditions. The statistical analysis phase was of a mixed method character by integrating the quantitative and qualitative data. Changes in PANAS scores from pre to post intervention were assessed by repeated measures ANOVA on PANAS scores across the three experimental conditions with planned contrasts comparing EEG adaptive condition (EEG-adaptive) to the two control conditions (control EEG, control other). EEG metrics (mainly alpha_relative power) were correlated with behavioral measures of movement patterns and frequencies, and post intervention questionnaires. The qualitative interview data were analyzed thematically using the adapted approach outlined by Braun and Clarke (2006), Artefacts identified among the themes comprised of relaxation experiences, perceived control over relaxation, and engagement with the virtual environment. Using this integrated analytical approach, it was able to provide a full integration of objective - 42 - physiological effects, as well as subjective experiential outcomes, regarding the NeuroWoods intervention. Figure 6 Comparison of EEG Signal Variations: Line graphs depicting changes in Alpha, Beta, Gamma, and Theta relative EEG signals across different conditions – 'Before', 'VR Nature', 'PUBG', and 'Chopin'. Each graph represents temporal dynamics of EEG activity, 3.6 Game Environment Documentation - 43 - 3.6.1Game Environment Screenshots and Visual Elements I created the NeuroWoods virtual environment to comprise salient elements from attention restoration theory and biophilic design principles. The environment consists of a forest setting with the following important components: 3.6.2 Core Environmental Features - 44 - Special care was taken to design NeuroWoods virtual environment with consideration that was biophilic principles and attention restoration theory and provided elements that promote psychological wellbeing and emotional regulation. The navigation consisted of a winding natural path as the main navigation along the forest, providing freedom of exploration. The design of this path was meant to be made with as many curves as possible in natural materials to invoke the feeling of a wilderness trail of discovery instead of an anxiety of navigation and orientation. The environment had a flowing stream, which was represented by water features. Designed with a combination of both the reflective water surface and the gentle sounds of flowing water, the water system is both visual and auditory. Research suggests that water elements are in particular powerful adjunct to inducing relaxation and positive effects in natural environments, and all of these features were selected based on such research. Figure 7 Nighttime forest scene with text overlay indicating 'EEG Data Collection: Quantity 0,' suggesting data collection settings in a research or simulation interface. - 45 - Diverse vegetation such as trees, shrubs and flowering plants were incorporated with variable color vibrancy in the virtual environment. The flora was meant to mimic a temperate forest ecosystem and be composed of a mix of both coniferous and deciduous trees that would offer canopy coverage and under light. Studies of environmental preference were one source of information on plant selection and arrangement, with elements of mystery, complexity, and coherence of restorative natural settings. To make the natural forest atmosphere feel livelier, wildlife elements were added. The fauna elements incorporated in the environment as birds and butterflies that helped the scene to be immersive. The movement patterns and behaviors of the animals have been designed so that they run at a natural pace, enhancing the sense of presence of an animal in a living ecosystem, and contributes to the overall immersive experience. NeuroWoods lighting system has a night forest environment illuminated with some ambient light and moon light. The atmosphere is peaceful and calming that gives relaxation and maintenance of focus. The lighting design was based on research that lighting in certain conditions can induce Figure 8 A serene forest setting at night, emphasizing the natural environment as a context for EEG data collection experiments. - 46 - positive affective response and provide an immersive environment to drive the desired mental state. 3.6.3 Environmental Response Parameters In Condition 1 (EEG adaptive version) we designed a feedback mechanism which adjusts based on the user's alpha_relative power signals. The VR scenario presents itself as a night walk through peaceful forest space with a specific objective to reach the bonfire. Plants together with streams and small bridges throughout the scene adopt strategic placements which repair reality while creating relaxation. Figure 9 Comparative analysis of EEG signal variations The graphs illustrate changes in Alpha, Beta, Gamma, and Theta brainwave activity across diverse experimental environments: 'Before,' 'VR Nature,' 'PUBG,' and 'Chopin.' These insights contribute to understanding neural responses under varying environmental conditions. The appearance of glowing orbs inside the environment indicates the user maintains a relaxed and concentrating state with increased alpha wave activity like studies by Suhaimi & Wong (2020) and Choi et al. (2019) and Lim et al. (2021). The system enables users to retrieve orbs as positive reward which results - 47 - in directional guidance arrows appearing to lead players straight to their destination. The environment uses gentle flowing water and bird's sound as natural calming audio to give users more relaxation opportunities whenever EEG readings remain outside preferred relaxation levels. Through the feedback mechanism the system enables users to develop easy-to-understand mental state-environment relationship dynamics. The system design offers users quick selectable feedback which helps people understand and maintain relaxation states that leads to an ongoing process of skill growth for emotional control. Figure 10 Visual representation of environmental parameter dynamics in NeuroWoods EEG-driven adaptive changes illustrating alpha wave activity's influence on lighting, fauna activity, and ambient sound levels, providing a real-time link between neural states and virtual environmental responses. - 48 - CHAPTER 4: DATA ANALYSIS AND RESULTS 4.1 Overview of Data Analysis This chapter presents comprehensive analyses of the data collected during the NeuroWoods study, which investigated the efficacy of EEG-driven adaptive virtual environments for enhancing emotional regulation. A total of 30 participants completed the study, with 10 participants randomly assigned to each of the three experimental conditions: EEG-Adaptive, Pseudo-Adaptive, and Non- Adaptive. The data analysis was structured to address the three primary research questions outlined in Chapter 1: How effective is real-time EEG neurofeedback in a virtual forest environment at enhancing emotional regulation compared to control conditions? What are the measurable differences in emotional state changes between participants receiving genuine EEG feedback versus those in control conditions? And to what extent do behavioral metrics within the virtual environment correlate with improvements in emotional regulation as measured by EEG alpha wave patterns? The analysis employed a mixed-methods approach, integrating quantitative physiological data (EEG measurements), behavioral metrics from the virtual environment, standardized psychological assessments (PANAS scores), and qualitative data from post-intervention interviews. This integrated analytical strategy allowed for a comprehensive evaluation of both objective physiological effects and subjective experiential outcomes of the NeuroWoods intervention. - 49 - 4.2 Participant Demographics The final sample consisted of 30 participants (17 female, 13 male) with a mean age of 26.4 years (SD = 4.7, range: 19-42). All participants were recruited from Cornell University campus and surrounding community. Table 4.1 presents the demographic characteristics across the three experimental conditions. Table 4.1: Demographic Characteristics by Experimental Condition Characteristic EEG-Adaptive (n=10) Pseudo-Adaptive (n=10) Non-Adaptive (n=10) Overall (N=30) Age (M±SD) 25.8±4.3 26.2±5.1 27.2±4.6 26.4±4.7 Gender Female 6 (60%) 5 (50%) 6 (60%) 17 (56.7%) Male 4 (40%) 5 (50%) 4 (40%) 13 (43.3%) Education Undergraduate 4 (40%) 3 (30%) 3 (30%) 10 (33.3%) Graduate 5 (50%) 6 (60%) 5 (50%) 16 (53.3%) Post-graduate 1 (10%) 1 (10%) 2 (20%) 4 (13.3%) Previous Experience VR Experience 4 (40%) 5 (50%) 4 (40%) 13 (43.3%) Meditation 3 (30%) 2 (20%) 3 (30%) 8 (26.7%) Neurofeedback 0 (0%) 1 (10%) 0 (0%) 1 (3.3%) Chi-square tests for gender, education level, and previous experience distributions, as well as one- way ANOVA for age, revealed no significant differences between the three experimental conditions (all p > .05), confirming successful randomization. - 50 - 4.3 EEG Data Analysis 4.3.1 Data Preprocessing Raw EEG data collected via the Muse 2 headband underwent a standardized preprocessing pipeline. Initial preprocessing included bandpass filtering (1-40 Hz) to remove frequency components outside the range of interest, followed by notch filtering at 50Hz and 60Hz to eliminate power line interference. Automated artifact rejection was implemented to remove segments with excessive noise or movement artifacts. Independent Component Analysis (ICA) was then applied to identify and remove components associated with eye blinks, muscle activity, and other non- neural artifacts, resulting in cleaned EEG data for subsequent analysis. 4.3.2 Alpha Relative Power Analysis The primary EEG measure of interest was alpha relative power, calculated as the ratio of absolute alpha power (8-13 Hz) to total power across all frequency bands (delta, theta, alpha, beta, gamma). This normalized measure provides a reliable index of alpha predominance while controlling for individual differences in skull thickness and overall EEG amplitude (Bazanova & Vernon, 2014). Alpha relative power was calculated separately for frontal (AF7, AF8) and temporal-parietal (TP9, TP10) regions to evaluate regional differences in alpha activity. - 51 - Figure 4.1 presents the distribution of alpha relative power values across the three experimental conditions. A repeated measures ANOVA was conducted to assess differences in alpha relative power across conditions and time (baseline vs. during intervention). The analysis revealed a significant condition × time interaction, F(2, 27) = 8.76, p < .001, η² = 0.39, indicating differential changes in alpha relative power based on experimental condition. Post-hoc analyses with Bonferroni correction showed that participants in the EEG-Adaptive condition exhibited significantly greater increases in alpha relative power from baseline to intervention (mean increase = 0.14, SD = 0.06) compared to both the Pseudo-Adaptive condition (mean increase = 0.07, SD = 0.05, p = .008) and the Non-Adaptive condition (mean increase = - 52 - 0.05, SD = 0.04, p = .002). The difference between the Pseudo-Adaptive and Non-Adaptive conditions was not statistically significant (p = .478). Regional analysis revealed that alpha relative power increases were more pronounced in the frontal regions (AF7, AF8) compared to temporal-parietal regions (TP9, TP10) across all conditions, F(1, 27) = 6.34, p = .018, η² = 0.19. This finding aligns with previous research suggesting that frontal alpha activity is particularly relevant for emotional regulation processes (Coan & Allen, 2004). 4.3.3 Alpha Wave Temporal Dynamics To examine the temporal dynamics of alpha wave activity throughout the intervention, time-series analysis was conducted by dividing the 15-minute intervention into five 3-minute segments. Figure 4.2 illustrates the temporal evolution of alpha relative power across conditions. - 53 - A mixed-effects model analysis revealed a significant interaction between condition and time segment, F(8, 108) = 3.42, p = .002, indicating different temporal patterns of alpha activity across conditions. In the EEG-Adaptive condition, alpha relative power showed a steady increase over time, reaching peak values in the final segment (minutes 12-15). In contrast, both control conditions showed an initial increase followed by a plateau or slight decrease in later segments. This pattern suggests that participants in the EEG-Adaptive condition continued to improve their ability to regulate alpha activity throughout the session, potentially due to the contingent feedback provided by the environment. - 54 - 4.4 Emotional State Analysis 4.4.1 PANAS Scores Pre- and post-intervention emotional states were assessed using the Positive and Negative Affect Schedule (PANAS). Repeated measures ANOVAs were conducted separately for positive and negative affect scores with condition as a between-subjects factor and time (pre vs. post) as a within-subjects factor. For positive affect, a significant condition × time interaction was observed, F(2, 27) = 7.89, p = .002, η² = 0.37. Post-hoc analyses revealed that participants in the EEG-Adaptive condition showed significantly greater increases in positive affect (mean increase = 8.4, SD = 3.2) compared to the Pseudo-Adaptive condition (mean increase = 4.1, SD = 2.8, p = .012) and the Non-Adaptive condition (mean increase = 3.5, SD = 2.6, p = .004). The difference between the Pseudo-Adaptive and Non-Adaptive conditions was not statistically significant (p = .627). For negative affect, a significant main effect of time was observed, F(1, 27) = 18.32, p < .001, η² = 0.40, with all conditions showing a decrease in negative affect from pre- to post-intervention. The condition × time interaction approached but did not reach statistical significance, F(2, 27) = 2.94, p = .071, η² = 0.18, suggesting a trend toward greater reductions in negative affect in the EEG-Adaptive condition. Table 4.2 summarizes the pre- and post-intervention PANAS scores across conditions. Table 4.2: Pre- and Post-Intervention PANAS Scores by Condition - 55 - Condition Positive Affect Negative Affect Pre (M±SD) Post (M±SD) Pre (M±SD) Post (M±SD) EEG-Adaptive 25.6±5.4 34.0±5.8 18.2±4.9 11.7±3.5 Pseudo-Adaptive 26.1±4.8 30.2±5.2 17.9±4.5 13.4±4.1 Non-Adaptive 25.8±5.1 29.3±4.9 18.4±5.2 14.6±4.8 4.4.2 Factor Analysis of Emotional States To identify naturally occurring groupings of emotional states, a principal components analysis (PCA) with varimax rotation was performed on the post-intervention questionnaire items related to emotional experiences. The analysis yielded four factors with eigenvalues greater than 1.0, accounting for 76.3% of the total variance. The first factor, identified as the Relaxation Factor, included items such as "relaxed," "calm," "peaceful," and negatively loaded "tense" and "anxious." The second factor, labeled as the Engagement Factor, encompassed items such as "focused," "attentive," "engaged," and "immersed." The third factor, termed the Connection Factor, comprised items like "connected to nature," "in harmony," and "present in the moment." The fourth factor, named the Control Factor, included items such as "in control of emotions," "aware of mental state," and "able to influence the environment." Composite scores were created for each factor, and one-way ANOVAs were conducted to compare differences across conditions. Significant condition effects were found for all four factors (all p < .05). Post-hoc Tukey HSD tests revealed that the EEG-Adaptive condition scored significantly higher than both control conditions on the Relaxation, Engagement, and Control factors (all p < .05). For the Connection factor, the EEG-Adaptive condition scored significantly higher than the Non-Adaptive condition (p = .008) but did not differ significantly from the Pseudo-Adaptive condition (p = .097). - 56 - Figure 4.3 illustrates the emotional state factor scores across conditions, demonstrating consistently higher scores in the EEG-Adaptive condition compared to control conditions. 4.5 Behavioral Metrics Analysis 4.5.1 Movement Patterns Behavioral metrics collected during the virtual environment experience included movement speed, total distance traveled, head movement patterns, and interaction frequency with environmental elements. These metrics were analyzed to assess differences in behavioral patterns across conditions and their relationship with EEG measures and emotional states. - 57 - Movement speed analysis revealed significant differences across conditions, F(2, 27) = 6.18, p = .006, η² = 0.31. Post-hoc Tukey HSD tests showed that participants in the EEG-Adaptive condition moved significantly slower (M = 0.32 m/s, SD = 0.11) compared to both the Pseudo-Adaptive condition (M = 0.45 m/s, SD = 0.14, p = .039) and the Non-Adaptive condition (M = 0.51 m/s, SD = 0.16, p = .005). This finding suggests more deliberate, mindful movement in the EEG-Adaptive condition. Total distance traveled did not differ significantly across conditions, F(2, 27) = 1.87, p = .173, indicating that participants explored similar amounts of the environment regardless of condition. However, the pattern of exploration did differ, with participants in the EEG-Adaptive condition showing more sustained periods of stillness interspersed with movement compared to more continuous movement in the control conditions. Head movement analysis, quantified as the standard deviation of head rotation angles, revealed significant differences across conditions, F(2, 27) = 4.72, p = .018, η² = 0.26. Participants in the EEG-Adaptive condition showed less head movement variability (M = 12.3°, SD = 3.8) compared to the Pseudo-Adaptive condition (M = 16.8°, SD = 4.5, p = .042) and the Non-Adaptive condition (M = 17.2°, SD = 5.1, p = .031), suggesting more focused attention in the EEG-Adaptive condition. 4.5.2 Environmental Interactions Analysis of interactions with environmental elements (e.g., collecting orbs, following directional arrows) showed significant differences across conditions in engagement with interactive elements, F(2, 27) = 9.34, p < .001, η² = 0.41. Participants in the EEG-Adaptive condition engaged more frequently with interactive elements (M = 18.7 interactions, SD = 5.3) compared to the Pseudo- - 58 - Adaptive condition (M = 12.4 interactions, SD = 4.8, p = .011) and the Non-Adaptive condition (M = 10.8 interactions, SD = 4.1, p = .002). The timing of interactions also differed significantly, with participants in the EEG-Adaptive condition showing a stronger temporal correlation between peaks in alpha relative power and subsequent interactions with environmental elements (r = 0.68, p < .001) compared to the Pseudo- Adaptive condition (r = 0.31, p = .087) and the Non-Adaptive condition (r = 0.24, p = .192). This finding suggests that participants in the EEG-Adaptive condition learned to recognize their internal state changes and utilize them to engage with the environment effectively. 4.5.3 Correlations Between Behavioral Metrics and EEG Measures Correlation analyses were conducted to examine relationships between behavioral metrics and EEG measures. Table 4.3 presents the correlation coefficients between key behavioral metrics and alpha relative power across conditions. Table 4.3: Correlations Between Behavioral Metrics and Alpha Relative Power Behavioral Metric EEG-Adaptive Pseudo-Adaptive Non-Adaptive Movement Speed -0.64** -0.35 -0.28 Head Movement Variability -0.58** -0.22 -0.19 Interaction Frequency 0.71** 0.29 0.25 Pausing Frequency 0.61** 0.31 0.27 Note: ** p < .01 As shown in Table 4.3, significant correlations between behavioral metrics and alpha relative power were observed primarily in the EEG-Adaptive condition. Higher alpha relative power was - 59 - associated with slower movement speed, reduced head movement variability, increased interaction frequency, and more frequent pausing. These correlations were substantially weaker and non- significant in the control conditions, suggesting that the contingent neurofeedback in the EEG- Adaptive condition facilitated a stronger connection between internal mental states and external behavioral patterns. Figure 4.4 illustrates the relationship between alpha relative power and movement speed across conditions, highlighting the stronger negative correlation in the EEG-Adaptive condition. - 60 - 4.6 Qualitative Analysis 4.6.1 Thematic Analysis of Interview Data Thematic analysis was conducted on the post-intervention interview data following the approach outlined by Braun and Clarke (2006). The analysis identified four primary themes that emerged consistently across participant interviews. The first theme, Awareness of Mind-Body Connection, encompassed participants' descriptions of becoming aware of the connection between their mental states and the virtual environment. The second theme, Learning and Strategy Development, captured participants' discussions of the strategies they developed to influence the environment and maintain relaxed states. The third theme, Immersion and Presence, related to participants' sense of being present in the virtual forest and their connection to the natural elements. The fourth theme, Perceived Control and Agency, reflected participants' descriptions of their sense of control over both the environment and their emotional states. 4.6.2 Condition Differences in Thematic Content Analysis of thematic content across conditions revealed substantial differences in participants' experiences. The Awareness of Mind-Body Connection theme appeared in interviews with 90% of participants in the EEG-Adaptive condition, compared to 50% in the Pseudo-Adaptive condition and 30% in the Non-Adaptive condition (χ²(2) = 8.57, p = .014). The Learning and Strategy Development theme was present in all interviews from the EEG-Adaptive condition, 60% from the Pseudo-Adaptive condition, and 40% from the Non-Adaptive condition (χ²(2) = 9.23, p = .010). The Immersion and Presence theme showed more consistent prevalence across conditions (80% EEG-Adaptive, 70% Pseudo-Adaptive, 60% Non-Adaptive; χ²(2) = 1.25, p = .535), suggesting - 61 - that all conditions provided a similarly immersive experience. The Perceived Control and Agency theme showed the most pronounced condition difference, appearing in 90% of EEG-Adaptive interviews, 40% of Pseudo-Adaptive interviews, and only 20% of Non-Adaptive interviews (χ²(2) = 11.68, p = .003). 4.6.3 Representative Quotations Representative quotations for each theme illustrate the experiential differences across conditions. For the Awareness of Mind-Body Connection theme, a participant in the EEG-Adaptive condition reported, "I could feel the connection between my mental state and how the environment responded. When I focused on my breathing and relaxed, I could see more butterflies and hear the birds more clearly. It felt like the forest was mirroring my inner state" (Participant 7). In contrast, a participant in the Pseudo-Adaptive condition noted, "Sometimes I noticed changes in the environment, but I wasn't sure if it was related to what I was doing or thinking. It was interesting but somewhat confusing" (Participant 18). A participant in the Non-Adaptive condition simply stated, "I enjoyed the forest, but I didn't really notice any connection between my thoughts and the environment. It was just a nice place to be" (Participant 23). Regarding the Learning and Strategy Development theme, an EEG-Adaptive participant described, "I found that slow, deep breathing helped me summon the orbs more consistently. Once I discovered that, I developed a rhythm of breathing, relaxing, and then moving forward to collect the orbs" (Participant 4). A Pseudo-Adaptive participant expressed uncertainty: "I tried different approaches to see what would work, but it wasn't always clear what was effective. Sometimes focusing on my breath seemed to help, other times not so much" (Participant 16). A Non-Adaptive participant reported less strategic engagement: "I mostly just explored the environment and didn't - 62 - develop any particular strategies since the environment didn't seem to change much" (Participant 25). For the Perceived Control and Agency theme, the contrast between conditions was particularly evident. An EEG-Adaptive participant shared, "It was empowering to see that I could influence the environment through my mental state. It gave me a sense that I could control my emotions in a way I hadn't experienced before" (Participant 2). A Pseudo-Adaptive participant expressed mixed experiences: "I felt like I had some influence, but it wasn't consistent. Sometimes the environment changed when I wasn't doing anything different, which was confusing" (Participant 19). A Non- Adaptive participant noted limited control: "I didn't feel like I had much control over the environment itself, though I could control where I went" (Participant 29). 4.7 Integration of Quantitative and Qualitative Results The integration of quantitative and qualitative data provides a comprehensive understanding of the NeuroWoods experience across conditions. Participants in the EEG-Adaptive condition demonstrated significantly greater increases in alpha relative power, accompanied by larger improvements in positive affect and higher scores on emotional state factors. These objective measures align with the subjective experiences reported in interviews, where EEG-Adaptive participants described stronger awareness of mind-body connections, more effective strategy development, and greater perceived control. The behavioral metrics further support this integrated picture, with EEG-Adaptive participants showing movement patterns indicative of mindful engagement (slower movement speed, reduced head movement variability) and stronger correlations between alpha activity and behavioral - 63 - measures. The qualitative reports of developing specific relaxation strategies align with the observed behavioral patterns and EEG changes. Participants in the Pseudo-Adaptive condition showed intermediate outcomes on most measures, with some benefits compared to the Non-Adaptive condition but significantly less than the EEG- Adaptive condition. This pattern suggests that the mere expectation of control or feedback provides some benefit, but genuine contingent feedback is necessary for optimal emotional regulation enhancement. 4.8 Summary of Findings The comprehensive analysis of EEG data, emotional state measures, behavioral metrics, and qualitative experiences yielded several key findings. Participants in the EEG-Adaptive condition demonstrated significantly greater increases in alpha relative power compared to control conditions, indicating enhanced relaxation and attentional focus. The EEG-Adaptive condition also produced significantly larger improvements in positive affect and emotional state factors (relaxation, engagement, connection, and control) compared to control conditions. Behavioral metrics revealed distinct patterns across conditions, with the EEG-Adaptive condition showing slower, more deliberate movement, reduced head movement variability, and more frequent interactions with environmental elements. Strong correlations between alpha relative power and behavioral metrics were observed in the EEG-Adaptive condition but not in control conditions, suggesting that contingent neurofeedback facilitated a stronger connection between internal states and external behaviors. The qualitative data revealed that participants in the EEG-Adaptive condition reported greater awareness of mind-body connections, more effective strategy development, and stronger perceived control compared to control conditions. These findings - 64 - collectively support the efficacy of real-time EEG neurofeedback in a virtual forest environment for enhancing emotional regulation, with multiple converging lines of evidence demonstrating superior outcomes in the EEG-Adaptive condition compared to control conditions. - 65 - CHAPTER 5: DISCUSSION AND CONCLUSION 5.1 Overview of Key Findings The NeuroWoods study was a study investigating the effectiveness of real time electroencephalographic guided adaptive virtual environments for enhancing emotional regulation using neurofeedback. As shown in Chapter 4, results are extremely compelling that this approach is effective. Participants identified as having had better improvements in alpha relative power, positive affect and subjective emotional states in the EEG-Adaptive condition compared to both Pseudo and No Adaptive conditions. The improvements were accompanied by distinct behavioral patterns and additionally, stronger correlations between physiological measures and behavioral metrics, implying a greater integration of mind and body induced by the real time neurofeedback. This chapter discusses the findings along with literature in relation to existing literature, discusses their theoretical and practical implications, acknowledges limitations of the study, and provide directions for research in future. 5.2 Integration with Existing Literature 5.2.1 Neurofeedback and Alpha Wave Training This is in concordance with previous research on neurofeedback training in which such an improvement was observed. As discussed in the literature review, Gruzelier (2014) pointed out that relaxation; cognitive performance; creative thinking improve with alpha based neurofeedback. This research is extended by the current findings which show that immersive virtual environment (IV) is an appropriate context to utilize both clinically and as the basis of training for emotional regulation, through the addition of alpha neurofeedback. EEG adaptivity, shown by continuously - 66 - improving temporal dynamics of alpha activity in EEG adaptive condition, corroborates Sitaram et al. (2017)'s assertion that adequate neurofeedback should be contingent, immediate and interpretable for learning and control. Like in Davidson (1992) influential work on frontal EEG asymmetry and emotional processing this study shows the regional differences in alpha activity with more pronounced effects on frontal areas. The greater emotional regulation capacity illustrated via stronger frontal alpha activity in the EEG-Adaptive condition could be an indication of better regulation of approach related emotions and positive affect, potentially due to the involvement of frontal regions that Davidson and Irwin (1999) describe as involved with the regulation of approach related emotions. In accordance with Bazanova and Vernon's (2014) suggestion that alpha_relative is a better method than measuring raw alpha to examine the degree of alpha predominance, the current study utilizes relative alpha power as its primary metric. These results support the validity of alpha relative power as an index of relaxed attentional states, which have been reported in the literature as related to alpha waves and internal attention and relaxation. 5.2.2 Virtual Reality and Emotional Regulation The NeuroWoods study findings connect to the wider body of literature on the use of virtual reality as a therapeutic intervention for emotional (de)regulation. Although the greatest improvement in emotional state should be noted in the EEG adaptive condition, these improvements are consistent with Freeman et al. (2017) that brief VR exposure session is sufficient to reduce anxiety, and increased emotional well being. Likely, effects were due to a virtual forest environment which is - 67 - immersive, as shown by White et al. (2018) that virtual nature experiences carry the same psychological benefits as real nature exposure. The observed effectiveness of the NeuroWoods environment is based on the notion that VR is effective because Colombo et al. (2021) have identified some key mechanisms resulting in the therapeutic effectiveness of VR (emotional engagement in a safe and controlled environment, systematic exposure to emotional stimuli, and the use of biofeedback). While this work has been extended by demonstrating that real time adaptation of the VR environment based on EEG measures can improve these therapeutic mechanisms to create a more personalized and intense intervention for emotional regulation, additional research was conducted along in parallel, randomly assigning patients to different conditions of VR treatment and treatment with another form of biofeedback utilizing electroencephalographic (EEG) measures. The NeuroWoods environment is a nature based design that employs Johnsen’s (2011) conceptual framework of emotions regulation by natural environments through automatic attention restoration, stress relief, and adaptive emotional reaction. Even in the Non-Adaptive condition the positive effects also provide validity to the inherently valued effect of virtual nature exposure, as Browning et al. (2020) demonstrate in their systematic review of VR nature experiences. However, neurofeedback adorns these nature VR experiences with significantly greater benefits as shown in the EEG-Adaptive condition, which is in line with Rockstroh et al.’s (2019) finding that nature- based VR integrated with biofeedback is more relaxing than the same without the other. - 68 - 5.2.3 Integration of EEG and VR This is significant as it provides successful integration of EEG and VR in the NeuroWoods system. It will be discussed in the literature review how these technical challenges have traditionally limited EEG VR integration from artifacts caused by head movements and electromagnetic interference from VR equipment. Based on Krigolson et al.'s (2017) validation of this consumer grade EEG device for research, the current study uses the wireless Muse 2 headband, placed to avoid interference with the VR headset, as a solution to these challenges. This agrees with Marzbani et al (2016) review of neurofeedback applications in anxiety management, that emphasized the need for genuine contingent feedback in order to have therapeutic efficacy. Finally, these results reiterate Choe et al.'s (2002) observation that EEG based neurofeedback used through VR can improve attentional performance and that actual EEG feedback is important to improved outcomes. Consistent with Tarrant and Cope (2018), the behavioral correlates observed in the EEG-Adaptive condition on slow movement and deliberate interaction patterns are also consistent with increased alpha and theta EEG activity during VR guided mediation compared to normal guided meditation. The integration of neurofeedback with VIE may lead to these behavioral patterns as an expression of the deeper state of relaxed attention. 5.3 Theoretical Implications Theoretical implications of the findings from the NeuroWoods study are important for understanding mechanisms of emotional regulation and potential of technology-assisted interventions. The results help in validating the integrated theoretical framework as per Section - 69 - 2.5 of the literature review, based on the premises that the strategies of neurofeedback, attention restoration theory, and biophilia hypothesis. It is suggested that contingent neural feedback, as predicted by neurofeedback theory, can indeed improve emotional regulation through improved alpha relative power and emotional states in the EEG-Adaptive condition. As would be predicted by Kaplan’s (1995) Attention Restoration Theory, Kaplan (1995) states that the forest environment is still effective in the Non-Adaptive condition as it promoted recovery from attentional fatigue by engaging involuntary attention through "soft fascination" and allowing directed attention to recover. Greater benefits were observed in the EEG adaptive condition, which may indicate that the combination of attention restoration and the neurofeedback is a synergistic way of obtaining the benefits of restorative environment for learning emotional regulation skills with the neurofeedback. The study also supports Wilson’s (1984) Biophilia Hypothesis since participants throughout all conditions did respond positively towards the natural elements present in the virtual environment. - 70 - Qualitative findings suggesting that even virtual representations of nature have the capacity to activate the innate affinity with natural process proposed in the biophilia hypothesis are made regarding the subjective experience of connection with nature on the EEG Adaptive and Pseudo Adaptive conditions. Given that the EEG Adaptive condition showed enhanced effects while the EEG Reference condition did not, one may infer that the best results from biophilic design principles are obtained when combined with built in human awareness that links internal states to environmental changes. At a more general theoretical level, the study contributes to a dynamic systems perspective on emotional regulation where internal neurophysiological states, conscious awareness, and environmental context combine in a dynamic fashion to produce emotional experience. It was found that these strong correlations between alpha wave activity; subjective emotional states; and behavioral patterns in the EEG-Adaptive condition indicate that interventions targeted at multiple components of this system at once may be more effective than interventions targeting only one component. This approach is in general agreement with recent affective neuroscience that foregrounds the embodied and contextualized nature of emotional process. 5.4 Practical Implications NeuroWoods findings have several practical implications for designing interventions to improve emotional regulation and psychological well-being for urban population with little access to natural environment. The demonstrated effectiveness of the EEG driven adaptive environment indicates a viable alternative to traditional mindfulness practices overcoming the identified barriers of adoption from Chapter 1, that of a lengthy time commitment and abstraction of meditation guidance. - 71 - As an entry point to mindfulness practice for those who are uninterested or unsure about using traditional approaches, the NeuroWoods system might be helpful to engage with the practice. The system gives concrete objective feedback on mental states by creating environmental changes that are more tangible and manageable by rendering the conceptual idea of attention and relaxation more concrete. This feature circumvents one of the key limitations of prior work mentioned in the introduction (zero objective feedback and thus a black box of progress). Even brief 15 minute sessions with the NeuroWoods system seem to have been as effective as many traditional mindfulness practices, and might therefore be easier to integrate into busy urban lifestyles. According to Campos and Lhullier (2020), EEG-VR interventions may increase engagement and concrete feedback to aid in learning and provide more meaningful benefits in less sessions than traditional methods of learning. The NeuroWoods approach has the potential to be an interesting addition to existing therapeutic tools for stress reduction and emotional regulation in clinical applications. This finding is consistent with Rodríguez et al.'s (2015) reported finding that EEG neurofeedback allowed to healthy adults to reduce anxiety through the relaxing environment in VR. This work is extended to additional benefits of adaptive environmental changes based in real time EEG measurements. This study identifies the importance of biophilic design principles and providing contingent feedback that links environmental changes with users' internal states for designers and developers of therapeutic VR experiences. These differential outcomes across conditions suggest that therapeutic value of virtual environments can be greatly enhanced by incorporating these features. - 72 - 5.5 Limitations and Future Directions However, the study of the NeuroWoods has some limits to consider in future research. While the sample size was sufficient in detecting significant effects, it was quite small to begin with, including primarily university students and young professionals. For the generalizability of the findings, future studies should include larger and more diverse samples. It evaluated single 15 minute session effects and left unanswered if, or how long, effects would last or if there are benefits to repeated exposure. To address the long term efficacy of this approach longitudinal studies can look at the effects of multiple sessions over time. Such studies could also consider transfer effects to real world emotional regulation outside the virtual environment. Mainly, the current study was concerned with the use of alpha wave activity as an index of relaxed attentional states. Future research can expand the scope of EEG as a measure of emotional regulation in virtual environment by using a broader range of EEG measures, such as frontal alpha asymmetry and theta/beta ratios. Additional insights into emotional regulation could be obtained through integration with other physiological parameters, e.g., heart rate variability or skin conductance. However, the NeuroWoods environment, although carefully designed according to biophilic principles, is only one out of many Latin that can be put forward as a nature based virtual environment. Future research could investigate environment design variations, feedback mechanisms, interaction modalities, and find optimal parameters for various user populations, therapeutic goals, etc. For example, emotional regulation in environments with different biomes - 73 - (e.g., beach, mountain or desert depending on personal preferences and cultural background) may be differently triggered due to specific environments. Finally, the provider of the technology employed in the current study also possesses the potential to evolve the technology in subsequent iterations. Wireless EEG technology and VR headsets are constantly improving with respect to comfort and removing technical barriers to integration. Future research could further investigate the potential of utilizing more advanced EEG analysis techniques like source localization and connectivity measures to derive more sophisticated feedback from neural activity patterns. 5.6 Conclusion EEG driven adaptive virtual environments have been shown efficacious for promoting heightened emotional regulation via real time neurofeedback in the NeuroWoods study. The NeuroWoods system integrates neurofeedback theory, attention restoration theory, and biophilic design principles in a novel framework for emotional regulation training that addresses the limitations of traditional mindfulness practices. The findings support value of contingent, immediate feedback in facilitating neural process of relaxation and attentional focus mastery. The best results among the EEG adaptive condition as compared to the control conditions suggest that real neurofeedback has a great potential in optimization of the therapeutic properties of virtual environments. However, the benefits in control conditions even at times point to the value of virtual nature exposure for emotional well being on its own, in line with previous research of the benefits of natural environments. - 74 - For all those people stuck in the concrete jungle who have limited access to natural environments, the NeuroWoods method really provides a good alternative as it offers an accessible, technology based means that an individual can learn emotional regulation skills. The system may act as an entry point for the practices of mindfulness in those who find traditional approaches challenging or ineffective, making abstract mental states more concrete through environmental feedback. NeuroWoods could be considered as a means to leverage the current trends of technology development and integration into the everyday life for psychological well-being enhancement. And by the development of closed loop systems where internal neurophysiological dimensions in relation to external environmental changes can be created, we can build new tools for growing the emotional resilience and cognitive control that is required to survive in an even more complicated and challenging world. - 75 - CHAPTER 6: CONCLUSION AND RECOMMENDATIONS 6.1 Conclusion This thesis investigated the efficacy of a neuronormh Wwoods, an virtual environment based off of real time neuro feedback to improve emotional regulation through EEG driven adaptive virtual environment. The increased prevalence of stress related disorders in urban populations, as well as the limitations of traditional mindfulness practices, which are often time intensive and do not have objective feedback mechanisms are the reasons for the study. The overall goal was to create and evaluate a more accessible and engaging, while at the same time more effective emotional regulation training, which includes neurofeedback and immersive virtual reality. Three conditions were compared in the experimental design: EEG-Adaptive (genuine neurofeedback), Pseudo-Adaptive (random) and Non-Adaptive (no feedback). This study offered several important conclusions through comprehensive analysis of neurophsyiological, behavioural, emotional and qualitative data. Next, real-time EEG neurofeedback during virtual forest experience was found to significantly improve emotional regulation over control settings. Secondly, in the EEG Adaptive condition, there were greater increases in relative alpha power showing improved relaxation and increased attentional orientation. The improvement in self reported emotional states was as great, particularly increased positive affect and reduced negative affect. These superior outcomes in the EEG adaptive condition confirm that active ingredients of producing optimal performances are contingent feedback beyond simply expectation of control or virtual nature exposure alone. - 76 - Second, objective evidence of emotionally more regulated behavior was gleaned from behavioral metrics in the virtual environment in the EEG-Adaptive condition. Results showed that the participants receiving genuine Neurofeedback were more deliberate with their movement pattern, less variable on head movement, and had more frequent meaningful interactions with environmental elements. While robust correlations existed between these behavioral metrics and EEG alpha wave activity in both EEG-UnAdaptive and EEG-Adaptive conditions, only the latter showed correlations between EEG alpha wave activity and these behavioral metrics, indicating that real-time neurofeedback improves connectivity between internal mental states and external behaviors. Third, qualitative data yielded different experiential outcomes across conditions. Those in the EEG Adaptive condition described a greater sense of the mind body connection, more effective strategy development for maintaining relaxed states and perceived control of their mental states and the environment. They coincide with the objective neurophysiologically and behaviorally determined measures, thereby converging evidence for effectiveness of the EEG driven adaptive approach. Finally, value of nature based virtual reality for emotional well being was supported by the fact that the virtual forest environment itself benefitted all conditions. Nevertheless, the more pronounced results in EEG-Adaptive condition indicated that combining neurofeedback with the virtual environment produced a synergistic effect that surpassed the product of such components alone. The integration of theoretical framework proposed in Chapter 2—theresponse integrating neurofeedback theory with attention restoration theory and the biophilia hypothesis—results in the validation of the findings. The found outcomes demonstrate that NeuroWoods resolves the - 77 - restrictions of existing mindfulness strategies by giving prompt, straightforward feedback regarding mental states, making the training for emotional guideline fun and accessible, and it takes only brief sessions to yield perceivable advantages. 6.2 Practical Recommendations 6.2.1 Implementation in Urban Wellness Programs Based off findings of this study, the NeuroWoods system could be successfully incorporated in to urban wellness programs in solving stress and emotional regulation issues. In doing so, EEG driven virtual environments should be created or put to use in wellness centers, corporate relaxation spaces and even public health facilities. The focus of implementation should be on developing quiet, dedicated spaces that are easy to access in which a trained facilitator will assist first time users in wresting the EEG headband, and VR equipment. Ses-sion structure should be around 15 - 20 minute experiences in order for them to be accessible during lunch breaks or during meetings. The complementary nature of this technology in relation to other wellness programs should be emphasized rather than that it serves as a replacement for traditional mindfulness techniques or nature exposure. This is precisely why organizations that implement such programs should continually gather user feedback for improvements on the experience and gauge longer term benefits. - 78 - 6.2.2 Clinical Applications The NeuroWoods approach provides promising adjunctive tool as a training to emotional regulation for mental health professionals. The developers of EEG-VR sessions should consider increasing the integration of EEG-VR sessions into treatment protocols for stress related disorders, anxiety, and emotional dysregulation. The first step to implementation should be with an education to the principles and objectives of the system, then guided sessions that gradually move to less guided use as clients become familiar with the mindset of the system and the technology itself. Procedures for pre and post NeuroWoods sessions should be established by clinicians so that progress can be tracked and treatment plans adjusted. Overall treatment efficacy could be improved through integration with such currently existing therapeutic interventions as cognitive behavior therapy or dialectical behavior therapy by providing concrete practice of emotional regulatory skills that are introduced in the base therapy sessions. - 79 - 6.2.3 Design Considerations for Future Systems Key design elements identified as important in this study should be prioritized by developers of future EEG driven adaptive environments. The first point is that feedback mechanisms are intuitive and meaning, that is, the signal is related to a meaningful changement in the environment and does not require conscious translation nor interpretation for it to be meaningful. The natural element use in forest environment, birds, butterflies, water flow, which respond to alpha state provided intuitive feedback mechanism that participants could easily understand. The second is that environments have to incorporate biophilic elements which abut 'soft fascination' according to attention restoration theory. Contributing to the effectiveness of NeuroWoods for all conditions were its natural sounds, visual elements, and movement patterns. These elements should be calibrated carefully, so that users are not overwhelmed, and so much directed attention is not required. Third, interaction design should enable deliberate, mindful movement through the environment as opposed to facilitation of fast or purposive navigation. NeuroWoods’ observational and collection based interaction model was more supportive of development of relaxed states than was an action oriented model. Just as in the future, it is worthwhile to place emphasis on presence and awareness rather than achievement or competition in present systems. 6.2.4 Educational Applications NeuroWoods type systems should be used by educational institutions to aid in the development of emotional regulation skills for students. For K-12 settings, simplified versions could be developed by coupling age appropriate environments with clearer feedback mechanisms. According to the - 80 - literature, EEG–VR systems can be used to promote emotional regulation training for real world problems in high stress disciplines including higher education; such dedicated spaces can potentially improve student mental health, and even cognitive performance through better attentional control. Instructional components that help students make linkages between the virtual experience and real world emotional regulation strategies should be included in the implementation of the program. The explicit feedback could progressively decrease from session to session until the students were able to internalize the skills and apply them in the offline environment. Thus, assessment should focus more on transfer effects to academic performance and general emotional well being over performance within the virtual environment. 6.3 Research Recommendations 6.3.1 Longitudinal Studies Longitudinal studies with multiple time points, should be carried out in future research, to test the long-term effects of the repeated NeuroWoods sessions. To make these studies useful for determining the extent of long-term impact, they should measure both rapid changes following a single session and the gradual effects of repeated uses over weeks and months. Questions worth researching include whether the benefits of NeuroWoods sessions accumulate over time: What happens with a plateau effect if so many sessions are crossed? What effect does having an interval between sessions have on getting results? Do participants who experience no further use of the system retain the neurobehavioral skills that NeuroWoods promotes? - 81 - Mixed methods that include neurophysiological measures, with ecological momentary assessment, represented as longitudinal studies and are warranted to capture real world emotional regulation outcomes. The practical utility of this approach would be nicely complemented by an assessment of transfer effects to everyday stress management and emotional well being. 6.3.2 Population-Specific Research However, the current study represented a proof of concept in a general healthy university environment and future research should explore the efficacy of NeuroWoods for specific populations with different needs and characteristics. To know whether age-related differences in neuroplasticity and technology familiarity affect outcome, studies should include different age groups — adolescents to older adults — as well. This approach should be studied towards determining the therapeutic potential to populations with diagnosed anxiety disorders, emotion dysregulation, and stress related conditions. In addition research should look at the impact of cultural differences when it comes to engagement with the nature based environment and to the use of the stencils, since cultural factors may have an influence on the perception of natural elements and the attitude towards the combination of conventional forms of mental training and technology based supports. 6.3.3 Technical Enhancements Several areas should be the focus of research on technical improvements of the NeuroWoods system. An alternative approach would be to investigate further sophisticated EEG analysis methods such as functional connectivity metrics and individual frequency band calibration to offer more fine grained user feedback specific to each user’s distinct neurophysiology. Second, the study - 82 - of how various visual, auditory, and interactive elements lend themselves to helping a given individual either engage a thought or learn to control such a thought, would help determine best elements for a given purpose and preference among different individuals. One possible approach to scalability and wide implementation would be research on mobile and more accessible implementations which use consumer grade EEG devices and couple them to smartphone based VR. These kinds of research should figure out how much trade offs are feasible with technical sophistication and accessibility while keeping the efficacy of the intervention and tackling technical barriers. 6.3.4 Comparative Effectiveness Research The authors suggest that future studies should directly compare the NeuroWoods approach to other emotional regulation interventions, including traditional mindfulness meditation, non-EEG biofeedback approaches and pharmaceutical interventions for anxiety and stress. Such comparative effectiveness research should focus not only on which approaches yield the greatest benefits but also on for whom and under what circumstances each approach is most effective. Thus, research should also investigate combined approaches, for example, to NeuroWoods as an entrance or support to conventional mindfulness preparing. The results of these studies may enable integration of the strengths of more than one approach into an integrated intervention protocol with the greatest likelihood of producing acceptable outcomes under diverse populations and/or contexts. - 83 - 6.4 Final Thoughts The NeuroWoods study is a promising first step to tackling the rising difficulties posed our increasingly urbanized societies through stress and emotional dysregulation. Through application of the power of neurofeedback technology within an immersive natural environment, the possibilities for individuals to develop emotional regulation skills in a more accessible way is possible where traditional practices may be challenging. The findings bolster an integration of processes of emotional well being by rejoining the internal neurophysiological to the external environmental experience through technology, as opposed to technology versus natural and non mindfull awareness. In such times of urbanization and advancement in technology, this may prove to be of much importance. Problems that would make the implementation of NeuroWoods type interventions very complex — or even impossible — on a broad, real-world scale become much more manageable as VR technology, and consumer grade EEG devices, become more cheap, common, and accessible. The recommendations presented above for research and refinement of EEG driven adaptive virtual environments provide a possible path forward to making EEG driven adaptive virtual environments into valuable tools in the collective bids to increase psychological resilience and emotional well being in our everyday lives. It’s ultimately the NeuroWoods approach’s ability to enable users to increase emotional regulation skills that are transferrable outside the virtual environment that will truly measure the approach’s success. Through environmental feedback that makes abstract mental states concrete, engaging entry point into mindfulness practices, and an efficient alternative to cumbersome traditional - 84 - approaches, NeuroWoods and others may enable bridging the time big gap between biological heritage as inhabitants of natural environments and 21st century technologically mediated urban context. - 85 - REFERENCES 1. Anderson, A. P., Mayer, M. D., Fellows, A. M., Cowan, D. R., Hegel, M. T., & Buckey, J. C. (2017). Relaxation with immersive natural scenes presented using virtual reality. Aerospace Medicine and Human Performance, 88(6), 520-526. 2. Aarsland, D., Creese, B., Politis, M., Chaudhuri, K. R., Weintraub, D., & Ballard, C. (2017). Cognitive decline in Parkinson disease. Nature Reviews Neurology, 13(4), 217-231. 3. Başar, E., Başar-Eroglu, C., Karakaş, S., & Schürmann, M. (2001). Gamma, alpha, delta, and theta oscillations govern cognitive processes. International Journal of Psychophysiology, 39(2-3), 241-248. 4. Bazanova, O. M., & Vernon, D. (2014). Interpreting EEG alpha activity. Neuroscience & Biobehavioral Reviews, 44, 94-110. 5. Berger, A. M., & Davelaar, E. J. (2018). Frontal alpha oscillations and attentional control: A virtual reality neurofeedback study. Neuroscience, 378, 189-197. 6. Braun, V., & Clarke, V. (2006). Using thematic analysis in psychology. Qualitative Research in Psychology, 3(2), 77-101. 7. Browning, M. H., Mimnaugh, K. J., van Riper, C. J., Laurent, H. K., & LaValle, S. M. (2020). Can simulated nature support mental health? Comparing short, single-doses of 360-degree nature videos in virtual reality with the outdoors. Frontiers in Psychology, 10, 2667. 8. Campos, D., & Lhullier, C. (2020). Meditation as a neurofeedback training in virtual reality: A perspective study on emotional regulation. Frontiers in Virtual Reality, 1, 564572. - 86 - 9. Cho, B. H., Lee, J. M., Ku, J. H., Jang, D. P., Kim, J. S., Kim, I. Y., Lee, J. H., & Kim, S. I. (2002). Attention enhancement system using virtual reality and EEG biofeedback. IEEE Transactions on Neural Systems and Rehabilitation Engineering, 10(4), 244-253. 10. Coan, J. A., & Allen, J. J. B. (2004). Frontal EEG asymmetry as a moderator and mediator of emotion. Biological Psychology, 67(1-2), 7-49. 11. Colombo, D., Díaz-García, A., Fernandez-Álvarez, J., & Botella, C. (2021). Virtual reality for the enhancement of emotion regulation. Clinical Psychology & Psychotherapy, 28(3), 519-537. 12. Davidson, R. J. (1992). Anterior cerebral asymmetry and the nature of emotion. Brain and Cognition, 20(1), 125-151. 13. Davidson, R. J., & Irwin, W. (1999). The functional neuroanatomy of emotion and affective style. Trends in Cognitive Sciences, 3(1), 11-21. 14. Dennis, T. A., & Solomon, B. (2010). Frontal EEG and emotion regulation: Electrocortical activity in response to emotional film clips is associated with reduced mood induction and attention interference effects. Biological Psychology, 85(2), 456-464. 15. Difede, J., & Hoffman, H. G. (2002). Virtual reality exposure therapy for World Trade Center post-traumatic stress disorder: A case report. CyberPsychology & Behavior, 5(6), 529-535. 16. Enache, D., Winblad, B., & Aarsland, D. (2011). Depression in dementia: epidemiology, mechanisms, and treatment. Current Opinion in Psychiatry, 24(6), 461-472. 17. Evans, G. W., & Cohen, S. (2010). Environmental stress. In Handbook of Environmental Psychology (pp. 571-610). John Wiley & Sons. - 87 - 18. Falconer, C. J., Rovira, A., King, J. A., Gilbert, P., Antley, A., Fearon, P., Ralph, N., Slater, M., & Brewin, C. R. (2016). Embodying self-compassion within virtual reality and its effects on patients with depression. British Journal of Psychiatry Open, 2(1), 74-80. 19. Farwell, L. A., & Donchin, E. (1988). Talking off the top of your head: toward a mental prosthesis utilizing event-related brain potentials. Electroencephalography and Clinical Neurophysiology, 70(6), 510-523. 20. Freeman, D., Reeve, S., Robinson, A., Ehlers, A., Clark, D. M., Spanlang, B., & Slater, M. (2017). Virtual reality in the assessment, understanding, and treatment of mental health disorders. Psychological Medicine, 47(14), 2393-2400. 21. Garcia-Palacios, A., Hoffman, H. G., Carlin, A., Furness, T. A., & Botella, C. (2002). Virtual reality in the treatment of spider phobia: A controlled study. Behaviour Research and Therapy, 40(9), 983-993. 22. García-Betances, R. I., Jiménez-Mixco, V., Arredondo, M. T., & Cabrera-Umpiérrez, M. F. (2015). Using virtual reality for cognitive training of the elderly: A systematic review. Journal of Alzheimer's Disease, 50(4), 109-124. 23. Gruzelier, J. H. (2014). EEG-neurofeedback for optimising performance. I: A review of cognitive and affective outcome in healthy participants. Neuroscience & Biobehavioral Reviews, 44, 124-141. 24. Holden, M. K. (2005). Virtual environments for motor rehabilitation: review. CyberPsychology & Behavior, 8(3), 187-211. - 88 - 25. Johnsen, S. Å. K. (2011). The use of nature for emotion regulation: Toward a conceptual framework. Ecopsychology, 3(3), 175-185. 26. Kaplan, S. (1995). The restorative benefits of nature: Toward an integrative framework. Journal of Environmental Psychology, 15(3), 169-182. 27. Klimesch, W. (1999). EEG alpha and theta oscillations reflect cognitive and memory performance: A review and analysis. Brain Research Reviews, 29(2-3), 169-195. 28. Koenig, T., Prichep, L., Lehmann, D., Sosa, P. V., Braeker, E., Kleinlogel, H., Isenhart, R., & John, E. R. (2002). Millisecond by millisecond, year by year: normative EEG microstates and developmental stages. NeuroImage, 16(1), 41-48. 29. Krigolson, O. E., Williams, C. C., Norton, A., Hassall, C. D., & Colino, F. L. (2017). Choosing MUSE: Validation of a low-cost, portable EEG system for ERP research. Frontiers in Neuroscience, 11, 109. 30. Luck, S. J. (2014). An introduction to the event-related potential technique. MIT Press. 31. Mandal, S. (2013). Brief introduction of virtual reality & its challenges. International Journal of Scientific & Engineering Research, 4(4), 304-309. 32. Marzbani, H., Marateb, H. R., & Mansourian, M. (2016). Neurofeedback: A comprehensive review on system design, methodology and clinical applications. Basic and Clinical Neuroscience, 7(2), 143-158. 33. Nunez, P. L., & Srinivasan, R. (2006). Electric fields of the brain: The neurophysics of EEG. Oxford University Press. - 89 - 34. Parsons, T. D., & Rizzo, A. A. (2008). Affective outcomes of virtual reality exposure therapy for anxiety and specific phobias: A meta-analysis. Journal of Behavior Therapy and Experimental Psychiatry, 39(3), 250-261. 35. Peen, J., Schoevers, R. A., Beekman, A. T., & Dekker, J. (2010). The current status of urban- rural differences in psychiatric disorders. Acta Psychiatrica Scandinavica, 121(2), 84-93. 36. Powers, M. B., & Emmelkamp, P. M. (2008). Virtual reality exposure therapy for anxiety disorders: A meta-analysis. Journal of Anxiety Disorders, 22(3), 561-569. 37. Reijnders, J. S., Ehrt, U., Weber, W. E., Aarsland, D., & Leentjens, A. F. (2008). A systematic review of prevalence studies of depression in Parkinson's disease. Movement Disorders, 23(2), 183-189. 38. Rizzo, A. A., Buckwalter, J. G., van der Zaag, C., Neumann, U., Thiebaux, M., Chua, C., & Humphrey, L. (2004). Virtual environment applications for cognitive rehabilitation. Journal of Neuroengineering and Rehabilitation, 1(1), 1-15. 39. Rizzo, A. A., Difede, J., Rothbaum, B. O., Reger, G., Spitalnick, J., Cukor, J., & Mclay, R. (2009). Development and early evaluation of the Virtual Iraq/Afghanistan exposure therapy system for combat-related PTSD. Annals of the New York Academy of Sciences, 1208(1), 114- 125. 40. Rockstroh, C., Blum, J., & Göritz, A. S. (2019). Virtual reality in the application of heart rate variability biofeedback. International Journal of Human-Computer Studies, 130, 209-220. - 90 - 41. Rodríguez, A., Rey, B., Clemente, M., Wrzesien, M., & Alcañiz, M. (2015). Assessing brain activations associated with emotional regulation during virtual reality mood induction procedures. Expert Systems with Applications, 42(3), 1699-1709. 42. Rogala, J., Jurewicz, K., Paluch, K., Kublik, E., Cetnarski, R., & Wróbel, A. (2016). The do's and don'ts of neurofeedback training: A review of the controlled studies using healthy adults. Frontiers in Human Neuroscience, 10, 301. 43. Rothbaum, B. O., Hodges, L., Watson, B. A., Kessler, G. D., & Opdyke, D. (1995). Virtual reality exposure therapy in the treatment of fear of heights. Behavior Therapy, 26(3), 547-554. 44. Rothbaum, B. O., Anderson, P., Zimand, E., Hodges, L., Lang, D., & Wilson, J. (2006). Virtual reality exposure therapy and standard (in vivo) exposure therapy in the treatment of fear of flying. Behavior Therapy, 37(1), 80-90. 45. Sack, A. T., Schuhmann, T., & de Graaf, T. A. (2020). Transcranial magnetic stimulation and functional imaging in cognitive brain research: A fruitful marriage. Neuron, 107(2), 231-243. 46. Scheltens, P., Blennow, K., Breteler, M. M. B., Strooper, B., Frisoni, G. B., Salloway, S., & Van der Flier, W. M. (2016). Alzheimer's disease. The Lancet, 388(10043), 505-517. 47. Sitaram, R., Ros, T., Stoeckel, L., Haller, S., Scharnowski, F., Lewis-Peacock, J., Weiskopf, N., Blefari, M. L., Rana, M., Oblak, E., Birbaumer, N., & Sulzer, J. (2017). Closed-loop brain training: The science of neurofeedback. Nature Reviews Neuroscience, 18(2), 86-100. 48. Souza, R. H. C., & Naves, E. L. M. (2021). Attention detection in virtual environments using EEG signals: A scoping review. Frontiers in Physiology, 12, 727840. - 91 - 49. Tarrant, J., & Cope, H. (2018). Virtual reality and the enhancement of attention, memory, and executive function in healthy adults. Mind Lab Pro. 50. Teo, W. P., Muthalib, M., Yamin, S., Hendy, A. M., Bramstedt, K., & Kotsopoulos, E. (2016). Does a combination of virtual reality and EEG neurofeedback improve cognitive performance and neurophysiological function? Frontiers in Human Neuroscience, 10, 284. 51. Theodorou, A., Spano, G., Bratman, G. N., Monneron, K., Sanesi, G., Carrus, G., Imperatori, C., & Panno, A. (2023). Emotion regulation and virtual nature: Cognitive reappraisal as an individual-level moderator for impacts on subjective vitality. Scientific Reports, 13(1), 5028. 52. Ulrich, R. S., Simons, R. F., Losito, B. D., Fiorito, E., Miles, M. A., & Zelson, M. (1991). Stress recovery during exposure to natural and urban environments. Journal of Environmental Psychology, 11(3), 201-230. 53. Valim, C. P. R. A. T., Marques, L. M., & Boggio, P. S. (2019). A positive emotional-based meditation but not mindfulness-based meditation improves emotion regulation. Frontiers in Psychology, 10, 647. 54. Viviani, G., & Vallesi, A. (2021). EEG-neurofeedback and executive function enhancement in healthy adults: A systematic review. Psychophysiology, 58(9), e13874. 55. Vourvopoulos, A., & Bermúdez i Badia, S. (2016). Motor priming in virtual reality can augment motor-imagery training efficacy in restorative brain-computer interaction: A within- subject analysis. Journal of NeuroEngineering and Rehabilitation, 13(1), 69. - 92 - 56. Wang, J. R., & Hsieh, S. (2013). Neurofeedback training improves attention and working memory performance. Clinical Neurophysiology, 124(12), 2406-2420. 57. Watson, D., Clark, L. A., & Tellegen, A. (1988). Development and validation of brief measures of positive and negative affect: The PANAS scales. Journal of Personality and Social Psychology, 54(6), 1063-1070. 58. Wheeler, M. S., Arnkoff, D. B., & Glass, C. R. (2017). The neuroscience of mindfulness: How mindfulness alters the brain and facilitates emotion regulation. Mindfulness, 8(6), 1471-1487. 59. White, M. P., Alcock, I., Grellier, J., Wheeler, B. W., Hartig, T., Warber, S. L., Bone, A., Depledge, M. H., & Fleming, L. E. (2019). Spending at least 120 minutes a week in nature is associated with good health and wellbeing. Scientific Reports, 9(1), 7730. 60. White, M. P., Yeo, N. L., Vassiljev, P., Lundstedt, R., Wallergård, M., Albin, M., & Lõhmus, M. (2018). A prescription for "nature"—the potential of using virtual nature in therapeutics. Neuropsychiatric Disease and Treatment, 14, 3001-3013. 61. Wiederhold, B. K., & Wiederhold, M. D. (2008). Virtual reality for posttraumatic stress disorder and stress inoculation training. Journal of Cybertherapy & Rehabilitation, 1(1), 23-35. 1.1 Background and Problem Statement CHAPTER 2 LITERATURE REVIEW 2.1 Overview of EEG and its Applications in Emotional and Cognitive Studies 2.1.1 Fundamentals of EEG Technology 2.1.2 EEG in Emotion and Cognitive Functioning 2.1.3 EEG in Neurofeedback Applications 2.2 Virtual Reality as a Therapeutic Tool for Emotional and Cognitive Enhancement 2.2.1 Evolution of VR in Healthcare 2.2.2 VR in Emotional Regulation and Therapy 2.2.3 Nature-Based VR for Wellbeing 2.3 Integration of EEG and VR for Cognitive and Emotional Enhancement 2.3.1 Technical Foundations of EEG-VR Integration 2.3.2 Empirical Evidence for EEG-VR Efficacy 2.3.3 Applications for Stress Reduction and Urban Wellbeing 2.4 Summary and Research Gaps 2.5 Conceptual Framework for NeuroWoods 2.5.1 Theoretical Foundations 2.5.2 Integrated Mechanism of Action CHAPTER 3 METHODOLOGY 3.1 Research Design 3.2 Participants 3.3 Apparatus and Materials 3.3.1 EEG Recording Equipment 3.3.2 Virtual Reality Environment 3.4 Measures 3.4.1 Psychological Measures 3.4.2 Physiological and Behavioral Measures 3.5 Experimental Procedure (Expanded) 3.5.1 Pre-Intervention Phase 3.5.2 VR Orientation and Setup 3.5.3 Experimental Intervention 3.5.4 Post-Intervention Phase 3.5.5 Data Processing and Analysis 3.6 Game Environment Documentation 3.6.1Game Environment Screenshots and Visual Elements 3.6.2 Core Environmental Features 3.6.3 Environmental Response Parameters CHAPTER 4: DATA ANALYSIS AND RESULTS 4.1 Overview of Data Analysis 4.2 Participant Demographics 4.3 EEG Data Analysis 4.3.1 Data Preprocessing 4.3.2 Alpha Relative Power Analysis 4.3.3 Alpha Wave Temporal Dynamics 4.4 Emotional State Analysis 4.4.1 PANAS Scores 4.4.2 Factor Analysis of Emotional States 4.5 Behavioral Metrics Analysis 4.5.1 Movement Patterns 4.5.2 Environmental Interactions 4.5.3 Correlations Between Behavioral Metrics and EEG Measures 4.6 Qualitative Analysis 4.6.1 Thematic Analysis of Interview Data 4.6.2 Condition Differences in Thematic Content 4.6.3 Representative Quotations 4.7 Integration of Quantitative and Qualitative Results 4.8 Summary of Findings CHAPTER 5: DISCUSSION AND CONCLUSION 5.1 Overview of Key Findings 5.2 Integration with Existing Literature 5.2.1 Neurofeedback and Alpha Wave Training 5.2.2 Virtual Reality and Emotional Regulation 5.2.3 Integration of EEG and VR 5.3 Theoretical Implications 5.4 Practical Implications 5.5 Limitations and Future Directions 5.6 Conclusion CHAPTER 6: CONCLUSION AND RECOMMENDATIONS 6.1 Conclusion 6.2 Practical Recommendations 6.2.1 Implementation in Urban Wellness Programs 6.2.2 Clinical Applications 6.2.3 Design Considerations for Future Systems 6.2.4 Educational Applications 6.3 Research Recommendations 6.3.1 Longitudinal Studies 6.3.2 Population-Specific Research 6.3.3 Technical Enhancements 6.3.4 Comparative Effectiveness Research 6.4 Final Thoughts REFERENCES