Artificial Intelligence, Immersive Technologies, and Neurotechnologies in Breathing Interventions for Mental and Emotional Health: A Systematic Review

Abstract

Breathing is one of the most vital functions for being mentally and emotionally healthy. A growing number of studies confirm that breathing, although unconscious, can be under voluntary control. However, it requires systematic practice to acquire relevant experience and skillfulness to consciously utilize breathing as a tool for self-regulation. After the COVID-19 pandemic, a global discussion has begun about the potential role of emerging technologies in breath-control interventions. Emerging technologies refer to a wide range of advanced technologies that have already entered the race for mental health training. Artificial intelligence, immersive technologies, biofeedback, non-invasive neurofeedback, and other wearable devices provide new, but yet underexplored, opportunities in breathing training. Thus, the current systematic review examines the synergy between emerging technologies and breathing techniques for improving mental and emotional health through the lens of skills development. The Preferred Reporting Items for Systematic Reviews and Meta-Analyses (PRISMA) methodology is utilized to respond to the objectives and research questions. The potential benefits, possible risks, ethical concerns, future directions, and implications are also discussed. The results indicated that digitally assisted breathing can improve various aspects of mental health (i.e., attentional control, emotional regulation, mental flexibility, stress management, and self-regulation). A significant finding of this review indicated that the blending of different technologies may maximize training outcomes. Thus, future research should focus on the proper design and evaluation of different digital designs in breathing training to improve health in different populations. This study aspires to provide positive feedback in the discussion about the role of digital technologies in assisting mental and emotional health-promoting interventions among populations with different needs (i.e., employees, students, and people with disabilities).

1. Introduction

Humans can survive weeks without consuming food, days without drinking water, and just a few minutes without breathing oxygen. Although it is an automatic action, breathing is a pivotal component of the total human being and one of the most vital physiological processes. For a long time, the importance of functional breathing was underestimated, because there was limited knowledge and relevant research about the crucial role of breathing in human health [1].

According to Pratscher et al. [2], the respiratory system is one of the most interconnected systems in the body, and breathing has a bidirectional relationship with stress mechanisms and complex mental and emotional operations. During the COVID-19 pandemic as well as in the post-COVID-19 era, there has been a rapid increase in breathing rehabilitation programs as a means of recovery from the physical and psychological symptoms derived from breathing dysfunctions [3].

The physiological and psychological benefits derived from breathing practices indicate that such interventions may hold the key to improving various aspects of human health [2,4]. Thus, a global discussion has begun regarding the impact of breathing on various aspects of human health, highlighting the importance of making breathwork an integral part of our daily lives [5,6].

In recent years, a growing scientific interest has developed as regards the effectiveness of contemplation practices in human health [7]. An essential component in almost all mindfulness training programs is the awareness and voluntary control of breathing operations. According to practitioners, the voluntary regulation of breathing is achievable and has the potential to induce a state of relaxation and psychological balance [8]. It is a key principle in contemplation science that breathing, although automatic, can be under human conscious control after systematic practice. While there are numerous breathing techniques, the common objective of all breathing practices is to familiarize trainees with their breathing operations and effectively train them to voluntarily manipulate respiration according to their internal state of being and external demands and goals [9,10].

Other studies highlight the need to develop breath-control skills. They also recognize metacognition as an inherent component of breath-control interventions [6,10]. Metacognition integrates a set of self-regulatory abilities, skills, and strategies that allow subjects to raise self-awareness and voluntarily control psychophysiological operations to achieve a state of optimal balance at the physical, mental, and emotional level [11]. Thus, metacognitive training constitutes an essential component of breath-control interventions [6,10].

A growing body of research in the field of breath-control training has already provided promising evidence about the benefits of such interventions in physical, mental, and emotional health [6]. Improvements in physiological operations (i.e., hormonal balance), cognitive functions (i.e., attention and memory), and emotional operations (i.e., emotional regulations) have already been revealed [12,13,14]. Research has also demonstrated that breathing interventions can significantly support people with mental and emotional disorders [15].

Studies indicate that conventional breathing training can be challenging, especially for novices. Breathing training requires long-term training. In addition, novices need constant and real-time feedback to learn how to be aware of, monitor, and control breathing, by flexibly employing different breath-control strategies. Trainees may have difficulty remaining focused on breathing as well as employing attention as a tool for regulating breathing. As a result, a considerable number of novices are likely to abandon training. [10]. Geographical, and financial constraints as well as limited time may discourage subjects from taking part in breathing interventions [16].

It is widely recognized that information and communication technologies (ICTs) can be valuable assistive tools in mental and emotional health training interventions [17]. The employment of digital technologies in training programs for promoting physical, mental, and emotional well-being has already shown positive outcomes [18]. It is noteworthy that the existing evidence indicates that digitally assisted mental health interventions are often more effective than conventional training programs. In addition, the trainees often report that they feel more satisfied with digitally assisted training programs compared with traditional ones [19].

Artificial intelligence (AI) has been recognized as a promising tool in the transformation of mental health programs, especially for sensitive populations (i.e., people with disabilities), mainly because of constant accessibility, accuracy, personalization, and interaction [20,21]. Immersive technologies including virtual reality (VR), augmented reality (AR), and mixed reality (MR) offer innovative approaches to mental health training by creating interactive, multisensory, and engaging training experiences. The research has already demonstrated that immersive technologies can modify cognitive functions, behavior, mood, and perception [22]. Neurotechnologies provide innovative interfaces that allow the stimulation, monitoring, interaction, and voluntary manipulation of the nervous system [23]. Multimodal biofeedback is commonly used to help users gain voluntary control over their physiological processes [24]. Non-invasive wearable devices such as headbands can measure brainwaves with the use of electroencephalography (EEG) sensors [25]. These neurofeedback technologies are user-friendly, portable, and low-cost. It is also easy to link them to free applications which can be easily downloaded on the user’s smartphone [26].

In recent years, there has been a growing scientific interest as regards the role of digital technologies in breathing interventions [27]. Chittaro et al. [27] for instance, evaluated different designs of breathing training mobile applications. The results were promising, indicating that such applications have a significant potential to support breathing training by providing visual feedback and instructions, encouraging users to adopt functional breathing patterns and perceive training as a joyful experience. Kressbach et al. [28] confirmed the potential of breathing apps and wearables to support health. Other studies have outlined the crucial role that virtual reality can play in breathing intervention for promoting general health [6,10,29].

However, the role of advanced technologies in breathing training remains largely underexplored. In addition, there is limited knowledge about the health benefits of different digitally assisted breathing interventions. Finally, there is also limited knowledge about the impact of digitally assisted breathing training on various aspects of mental and emotional health with a special focus on skills development [6,27].

Considering the leading role that innovative technologies may have in future mental health programs, along with the importance of promoting breathing training as a positive daily habit, the current study aims to shed more light on the role of emerging technologies as assistive tools in breathing interventions for promoting mental and emotional health with a special focus on skills development.

Thus, the current study aims to summarize and synthesize the existing evidence about the effectiveness of different advanced technologies (i.e., artificial intelligence, virtual, augmented, and mixed reality, biofeedback, and neurofeedback) in breathing interventions. In addition, the review aims to record the beneficial effects of such interventions on mental and emotional health with a special focus on the skills domain.

We decided to conduct a systematic review and follow the methodological steps recommended by Prisma’s statement [30]. The current review includes studies that recruited healthy populations as well as populations with different health issues.

The central research questions of the current systematic review study are the following:

• What types of emerging technologies have the potential to assist breathing training for the promotion of mental and emotional health?
• Can digitally assisted breathing interventions improve trainees’ mental and emotional health?
• What types of mental and emotional health skills are mostly developed?

The current study hypothesizes that emerging technologies can assist breathing interventions in various ways, accelerating and optimizing the training outcomes for the benefit of mental and emotional health. The role of emerging technologies is hypothesized to assist the training objectives, by increasing trainees’ autonomy, motivation, and satisfaction. Most importantly, digitally assisting breathing interventions are expected to equip trainees with improved abilities and skills capable of enhancing mental and emotional health.

The results of the current review study aspire to provide positive feedback in the discussion about the potential role of advanced technologies in breathing training for the promotion of mental and emotional health.

2. Materials and Methods

2.1. The Study Design

The current review study was conducted following the guidelines of the Preferred Reporting Items for Systematic Reviews and Meta-Analyses (PRISMA) statement. The PRISMA statement aims to guide researchers in the process of identification, evaluation, selection, and synthesis of the chosen studies [30]. The review was registered with the Open Science Framework which is a platform that allows the registration of protocols for systematic reviews (https://osf.io/gmrqf/, assessed on 9 May 2024) [31]. The duration of the review lasted from August 2023 to March 2024. Three researchers were engaged in the review process.

2.2. Eligibility Criteria

In the current review paper, we decided to include experimental studies. A special focus was given to randomized controlled trials. Non-randomized controlled trials and quasi-experimental studies were also accepted, provided that they were consistent with the objectives of the review and followed a transparent methodology. Systematic reviews, meta-analyses, and chapters were not included. In addition, design frameworks or protocols for randomized controlled trials without evaluation were excluded too.

Regarding the type of population, the criteria were not strict. We included studies with both healthy participants and participants with health-related problems. Talking about the type of intervention, the review included only studies that applied breathing interventions assisted by digital technologies. We mainly focused on emerging technologies, including artificial intelligence, immersive technologies, and biofeedback technologies. Among the inclusion criteria was the selection of studies that provided evidence about the development of skills capable of improving mental and emotional health. In

Table 1. The main inclusion and exclusion criteria applied.

Inclusion	Exclusion
(a) Experimental studies (i.e., randomized controlled trials). (b) Published after 2014. (c) Breathing intervention as the primary type of intervention with the support of emerging technologies. (d) Studies that evaluated aspects of mental and emotional health or skills required for being mentally and emotionally healthy. (e) Healthy participants and participants with health-related problems.	(a) Systematic reviews, meta-analyses, and book chapters. (b) Protocols and design frameworks without evaluation. (c) Studies that examined the effectiveness of breathing interventions without the assistance of emerging technologies. (d) Studies that employed digital technologies only as a means of assessment.

, the inclusion and exclusion criteria are further analyzed.

2.3. Information Sources

The current systematic review was conducted with the use of four academic search engines: Web of Science, Pubmed, Scopus, and Scholar Google. The selected databases provide access to peer-reviewed and high-quality studies. They are widely recognized as valuable academic research engines for conducting systematic reviews. Among their potential benefits, we can outline that they provide quick access to relevant scientific papers. A wide range of search tools are offered, accelerating the process of identifying relevant papers. Most importantly, the selected databases can identify papers that fit with the objectives of this paper. More specifically, studies in the disciplines of computer sciences and digital health can be easily identified.

2.4. Search Strategy

The search was limited to papers published in the last ten years (between 2014 and March 2024). The duration of the review lasted from August 2023 to March 2024. Three researchers were engaged in the review process. The central search keywords employed in our search strategy included concepts related to breathing training techniques such as diaphragmatic breathing, nostril breathing, square breathing, mindful breathing, breath awareness, and pranayama breathing techniques.

The search also integrated terms relevant to emerging technologies, such as artificial intelligence, chatbots, intelligent agents, virtual coaches, virtual reality, augmented reality, mixed reality, metaverse, wearables, sensors, biofeedback, and neurofeedback technologies.

We also attempted to include in our search process keywords related to mental and emotional health or skills that are closely related to mental and emotional health. For instance, we searched for the following terms: self-awareness, emotional awareness, self-regulation, emotional regulation, mental or emotional flexibility, emotional recognition, self-observation, stress management, and resilience. The use of boolean operators facilitated the search process since we could integrate the central search keywords. Moreover, we used the filters provided in the academic databases to retrieve the papers that fit the objectives of the current review.

Table 2. The main search strings with the main keywords.

The Search Strings with the Main Keywords

“Diaphragmatic breathing” OR “pursed-lip breathing” OR “alternate nostril breathing” OR “square breathing” OR “buteyko breathing” OR “deep breathing” OR “mindful breathing” OR “kapalabhati” OR “brastrika Pranayama” OR “sitkari pranayamas” OR “bhramari Pranayama” OR “hold-breathing” AND “Artificial intelligence” OR “conversational agents” OR “chatGPT” OR “gemini” OR “virtual coaches”, “intelligent agents” OR “smartphone applications” OR “immersive technologies” OR “virtual reality” OR “augmented reality” OR “mixed reality” OR “metaverse” OR “biofeedback” OR “neurofeedback” OR “brain-computer interfaces” OR “brain-computer interfaces” AND “mental health” OR “emotional health” OR “resilience” or “metacognitive skills” OR “emotional intelligence skills” OR “self-regulation” OR “emotional regulation” OR “self-awareness” OR “emotional awareness” OR “emotional recognition” OR “impulse control” OR “attentional regulation” OR “flexibility” OR “stress management”

illustrates the search strings with the main keywords used in the search process.

2.5. Selection Processes

In the first stage, the eligibility criteria were determined. In addition, a keyword list was made to initialize our research in the academic databases, using boolean operators and various search filters. After employing various keyword strings, we gathered a significant number of candidate studies. In the next stage, we tried to further process the identified papers, prioritizing studies that included the main terms in the titles and the abstracts. After eliminating duplicate studies, references that did not meet the eligibility requirements were excluded in the abstract/title screening stage. The remaining studies followed the full-text screening phase, during which the content of the studies was evaluated in detail. At this crucial stage of screening, methodological parameters were also carefully examined. Two evaluators participated in the screening procedure independently. Each reviewer could choose one of the following decisions: “acceptance”, “rejection”, or “debatable”, providing relevant justification. When in doubt, the screeners could have a relevant discussion. In case of disagreement, the third reviewer provided his opinion.

2.6. Data Collection and Extraction

Initially, we collected the full text of the selected studies. Afterward, we determined the pieces of information we wanted to include, taking into consideration the objectives and the research questions of the current study. We decided to collect the following data: the authors’ information, the year of publication, the type of breathing technique, and the digital design of the intervention. In addition, the sample’s characteristics were collected. More specifically, we recorded information about the number, gender, age, and health condition of the participants. The type of the measurement was identified. Finally, the main findings were collected. Data were gathered from each chosen paper by two reviewers independently. The reviewers recorded the data in paper forms. After the data were extracted, responses were compared to ensure agreement or to identify discrepancies. In case of disagreements, the first solution was discussion between the reviewers. The second solution was arbitration by the third reviewer. Afterward, the collected data were recorded and classified in a single summary table to provide an overview of the findings and facilitate the synthesis procedures.

Table A1. Summary of the studies that employed emerging technologies in breathing training for mental and emotional health through the lens of skillfulness.

Reference	Country	Breathing Intervention	Digital Design	Sample	Health Condition	Duration	Type of Measurement	Research Design	Main Improvements
Mehta et al. 2021 [91]	USA	Breathing training combined with gratitude and self-compassion techniques	AI-powered smartphone app	n = 4.517, F = 3687, M = 693, NB = 155, Mage = 28.73	HS	4 weeks	GAD-7, PHQ-9	Longitudinal study	Emotion regulation, cognitive control adaptive behavior
Vertsberger et al. [93]	Israel	Breathing training combined with mindfulness practices	AI-powered conversational agent via smartphone	n = 10.387, no gender, Mage = 16	HS	16 weeks, 45.39 days	WHO-5	Longitudinal study	Emotional and psychological flexibility
Gabrielli et al. 2021 [92]	Italy	Mindfulness for presence and attention	Atena psychoeducational chatbot	n = 71, F = 48, M = 23, Mage = 20.6	HS	Eight sessions, 4 weeks, two per week for 10 min.	GAD-7, PSS-10, FFMQ	Mixed Methods Proof-of-Concept Study	Self- and emotional awareness, negative thoughts regulation
Leo et al., 2022 [94]	USA	Deep breathing, mindfulness	AI-based chatbot	n = 61, F = 53, M = 7, NB = 1, Mage = 55	HS	8 weeks	PROMIS	Feasibility prospective cohort study	Mood and anxiety regulation
Potts et al., 2023 [90]	UK	Mindfulness, breathing, gratitude	Multilingual chatbot	n = 348, F= 254, M = 94, Mage = 30,	HS	12 weeks	SWEMWBS, WHO-5, SWLS	Pre-Post Multicenter Intervention Study	Mental well-being
Tu et al., 2020 [95]	USA	Respiratory Sinus Arrhythmia biofeedback-based breathing training	Smart breathing training system supported by smartphone-based VR and sensors on a smartwatch	n = 10	HS	6 sessions (45 min.)	Physiological measurements	Single case study	Improved cognitive functions, stress reduction, improved attention
Shih et al., 2019 [116]	USA	Slow-paced breathing (six breathing cycles per minute)	Smart mobile app, gamified biofeedback	n = 43, F = 31, M = 12, Mage = 25.9	HS	50 min.	Physiological recording, self-reports	Pilot study	Adaptation breathing skills and improved cardiac functioning. Relaxation skills, perceived control, enjoyment
Chandler et al., 2020 [67]	USA	Breathing meditation	Smartphone app	n = 30, F = 15, M = 15, Mage = 45.0	Hypertension	12 months	measures of heart rate, PSS	RCT	Reduced blood pressure
Shamekhi et al., 2018 [69]	USA	Mindful breathing	Automated conversational agent	n = 24, F = 15, M = 9, Mage = 40.9	HS	Two 12-min sessions	MAAS, self-report measure	Pilot study	Relaxation, attention, body awareness, respiration self-regulation
Støre et al., 2022 [96]	Sweden	Deep breathing	Robot	n = 44, (nexp = 22, nclt = 22), M = 9, F = 35, Mage = 50.05	DEP, ANX, insomnia	3 weeks	ISI, HADS-A/D, PSAS	RCT	No significant improvements in arousal, anxiety, depression, and sleep problems
Asadi et al., 2022 [97]	Denmark	Deep breathing	Robot, breathing belts, EEG device	n = 28, M = 17, F = 11, Mage = 26.17 for males and 30.81 for female	HS	Single session	EEG headband assessment, questionnaire,	Pilot study, between-subject approach	Positive emotions, lower stress, improved regulation of breathing
Blum et al., 2019 [24]	Germany	Slow-paced breathing technique (6 br/min with a 5:5 I/E ratio)	VR HRV-biofeedback	n = 60, (nexp = 31, nclt = 29), M = 29, F = 31, Mage = 33.5	HS	1 session (10 min)	STAI-S, CIQ, SMS, Stroop task, heart data	RCT	Heart coherence, resilience, attention regulation
Hu et al., 2021 [98]	USA	Diaphragmatic slow breathing (6 br/min with 3:6:1 I/E ratio)	VR HRV-biofeedback	n = 40, (nexp = 20, nclt = 20), M = 19, F = 21, Mage = 28	HS	7 sessions	PANAS	RCT	Analgesic effects, interoception, and exteroception of attention
Shiban et al., 2017 [105]	Germany	Diaphragmatic breathing (6 br/min)	V6 HMD	n = 29, (nexp. = 15, nClt. = 14), F = 24, M = 5, Mage = 34.3	DEP, PB, PD	Single session	HR, SCL, RR, ASI, FFS, FSB	RCT	Emotional regulation, stress management
Rockstroh et al., 2019 [76]	Germany	Slow respiration (6 br/min)	Immersive VR HRV-biofeedback	n = 68, (nexp = 24, nclt1 = 22, nclt2 = 22), M = 27, F = 41, Mage = 22.9	HS	Single 10 min session	HR, MDBF, VAS	RCT	Mood regulation, attention regulation, self-awareness, relaxation
Waller et al. 2021 [100]	Canada	Breath awareness techniques	360° video-guided meditation via HMD	n = 82, M = 32, F = 50, 17 to 28 years	HS	Single session	MBAS, mDES, BASS	RCT	Experience of awe
Weerdmeester et al., 2021 [79]	Netherlands	Diaphragmatic breathing (6 br/min with 3:7 I/E ratio)	VR game-based biofeedback	n = 112, (nexp = 57, nclt = 55), M = 11, F = 101, Mage = 20.71	ANX	2 training phases of 2 sessions each (10 min)	STAI	RCT	Internal control, engagement, concentration, self-monitoring, self-regulation
Weibel et al. 2023 [108]	Switzerland	Slow-paced breathing (4.5–6.5 br/min)	VR HRV biofeedback	n = 107, (nHRV-BFscreen = 25, nHRV-BFHMD = 25, n sPBscreen = 29, nsPBHMD = 28), M =47, F = 60, Mage = 22.52	HS	Single session (120 min.)	MDMQ, ACTA, PQ, UTAUT	RCT	Stress management, resilience, self-motivation, beneficial psychological and cardiac effects, cardiac coherence
Lan et al., 2021 [26]	Taiwan	Slow abdominal breathing	VR headset, biofeedback	n = 20, (nexp = 10, nclt = 10), M = 10, F = 10, Mage = 20.6	HS	Eight 20 min. sessions within 2 weeks	EEG, Questionnaires about users’ satisfaction	RCT	Improved breathing rate, improved breathing control skills, greater focus, improved reaction, accuracy, less mind-wandering
Kang et al., 2021 [112]	Korea	Deep breathing	VR device (Oculus Rift CV1 head-mounted display and controller, soft stretch sensor	n = 50, (nexp = 25, nclt = 25), M = 23, F = 27, Mage = 42.52	HS	4 sessions	Spirometry measurements	RCT	Self-regulation
Rockstroh et al., 2021 [77]	Germany	Diaphragmatic breathing (7.5 breaths per minute)	VR-based respiratory biofeedback game	n = 45	HS	6 sessions	PSS-10, CBI	Longitudinal study	Greater breath awareness, stress management, burnout reduction,
Michela et al., 2022 [78]	Netherlands	Deep and slow diaphragmatic breathing	Biofeedback VR-based action game	n = 9, M = 9 Mage = 43.2	HS	4 weeks, 10 sessions, 15 min.	STAI	Single-Case Design Study	Increased self-awareness of physiological control, attention, breath-control skills, motivation, response inhibition under stress, improved performance in decision-making contexts
Bossenbroek et al. 2020 [111]	Netherlands	Mindful breathing	Virtual reality, biofeedback, serious game	n = 8 F = 1, M = 7, Mage = 14.67	ADHD	4 weeks, 6 sessions. 15 min.	STAI, disruptive classroom behavior scale	Single-case experimental study	Self-regulation
Blum et al., 2020 [106]	Germany	Diaphragmatic breathing	VR biofeedback	n = 72 (nexp = 36, nclt = 36), F = 56, M = 16, Mage = 21.6	HS	Single session	UEQ	RCT	Greater focus on breath
Roo et al. 2017 [113]	USA	Mindful Breathing	MR (immersive VR, AR, and tangible interface	n = 12, F= 12, Mage= 45	HS	2 sessions	FFMQ, STAI, TMS	Usability study	Increased relaxation and state mindfulness
Tinga et al., 2019 [109]	Netherlands	Mindful Breathing	VR biofeedback	n = 60, F = 37, M = 23, Mage = 22.07	HS	1 session (6 min.)	ECG signal, TSST	RCT	Increases short-term HRV, increased attention, and motivation
Seol et al. 2017 [101]	Korea	Mindfulness scenarios, breathing regulatory guidance	HMD, leap motion sensor, PSL-lecg2 and Falcon device	n = 5	PD	2 sessions	DASS	Usability study	Emotional stability, stress management
Cook et al., 2020 [103]	USA	Deep and slow-paced breathing (3 sec. inhale, 7 sec. exhale for 6 breaths a minute	VR headset	n = 15, F = 7, M = 8, Mage = 16.9	Concussion	6 weeks	PCSS, POMS, VOMS, pre- and post-vr symptom ratings,	Pilot study	Stress regulation, less tension and fatigue, improved attention and mental clarity
Ilioudi et al. 2023 [102]	Sweden	Βreathing combined with guided relaxation	VR HMD (an Oculus Go) running the mobile app	n = 60 (nexp = 40, nClt = 20), F = 35, M = 25 Mage = 39.1	ANX, DEP, BPD	1 session	MADRS-S, BAI	Quasi RCT	Self-relaxation
Sra et al., 2018 [110]	USA	Βreathing actions (inhalation, exhalation and hold)	Single and multiplayer virtual reality games	n = 16 (8 groups of 2), F = 4, M = 12, Mage = 26.8	HS	4 repetitions (5 min.)	SUS, GEQ	Pilot study, within-subjects design	Increased presence, attention, self-control skills, engagement, motivation
Kluge et al., 2021 [107]	Australia	Controlled breathing	VR application, biosensor hardware and software	n = 30, F = 17, M = 13, Mage = 30.5	HS	3 training sessions, 3 weeks	Breathing measurements	Pilot study	Increased awareness and control of breath, attention, motivation
Kojic et al., 2019 [104]	Germany	Exhale during the drive, inhale during the recovery	VR biofeedback exergame (rowing)	n = 23, F = 33, M = 53, Mage = 23.67	HS	5 sessions (60 min.)	SFSS-2, IPQ	Within-subjects design	Reduced helpfulness, increased flow, and sympathy
van Rooij et al., 2016 [81]	Netherlands	Diaphragmatic breathing	Biofeedback virtual reality game	n = 86, F = 33, M = 53, Mage = 10.1	ANX	Single session (7 min.)	STAIC, PANAS	Pilot study	Stress regulation
Yüksel et al., 2020 [99]	USA	Slow diaphragmatic breathing combined with relaxation/distraction strategies	Immersive VR environment	n = 29, F = 17, M = 12, 16–18 years old	SD	Single session (20 min.)	Polysomnography	Pilot study	Reduced worry, reduced heart rate, increased sleep efficiency, improved relaxation
Ponzo et al., 2020 [115]	UK	Diaphragmatic breathing	Mobile app and paired wearable device (BioBeam)	n = 262 (nexp = 130, nclt = 132), F = 92, M = 54, Mage = 19.9	ANX	4 weeks	STAI, WEMWBS, DASS-21, PHQ-9	RCT	Stress regulation, mood regulation, decreased depression, increased perceived well-being
Chelidoni et al., 2020 [114]	UK	Diaphragmatic breathing (6 breaths per minute for 5 min.)	Smartphone app	n = 75, F = 45, M = 27, Mage = 32.32	HS	5 min	FFMQ, SPFC, SSS, VAS, CPT	Randomized laboratory-based experiment	Relaxation and heart rate variability improved
Morris et al. 2018 [118]	USA	Diaphragmatic breathing	App developed for Android Wear smartwatches	n = 20, M = 12, F = 2, 29 to 50 years	PTSD, TBI	Daily for 2–4 weeks	Stress ratings through Google Analytics, GAS, PCL-5, BAI	Usability testing	Stress management, self-regulation, mood regulation, reduction in depression symptoms
Smith et al., 2019 [71]	USA	Guided breathing exercises	Clothing-attached, physiological monitoring device capable of sensing respiratory patterns and a smartphone application	n = 169 (nexp = 67, nclt = 102), F = 93, M = 76, Mage = 33.2	HS	4 weeks	PSS, MASQ, PANAS	RCT	Improved stress regulation. Emotional regulation
Al Rumon et al., 2023 [117]	USA	Guided breathing exercises (slow, normal, and fast breathing)	IoT-enabled smart T-shirt	n = 10, F = 3, M = 7, 27–34 years old	HS	2 sessions per exercise for a total of 30 sessions	EEG, HR, RR	Pilot study	Stress management
Schuurmans et al. 2021 [123]	Netherlands	Breathing training	Non-invasive headband headband, iPad	n = 77 (nexp = 40, nclt = 37), F = 31, M = 46, Mage = 15.25	PTSD	6 weeks 5 min. sessions	TRIER-C	RCT	Self-regulation, stress management
Crivelli et al., 2019 [122]	Italy	Focused attention, breathing awareness	Non-invasive EEGheadband and smartphone app	n = 50 (nexp1 = 15, nexp2 = 17, nClt = 18), Mage = 22.94	HS	2 weeks	EEG, PSS-10, STAI-trait, FFMQ	Three-branch pre-post experiment study	Attentional regulation
Bhayee et al., 2016 [119]	Canada	Breath-focused meditation	Non-invasive EEGheadband, mobile app	n = 26 (nexp = 13, nClt = 13), F= 12, M = 14 Mage = 33	HS	6 weeks	BSI, Stroop, digit span task, FMI, PANAS, BFI WHOQOL-BREF	RCT	Attention control, inhibition, and positive mood
Balconi et al. [120]	Italy	Breathing awareness	Lowdown Focus glasses and mobile app	n = 50 (nexp = 38, nClt = 12), Mage = 24.20	HS	3 weeks	Attentional matrices, DBQ, MFTC, Stroop test	RCT	Attention regulation
Hunkin et al., 2020 [121]	Australia	Breath-focused meditation	Non invasive EEG headband and iPad tablet	n = 68, F = 40, M = 28, Mage = 22.66	HS	2 weeks	EEG assessment, MEQ,	Crossover trial	Self-awareness, better sense of self-control

n  =  total sample, n_exp: total sample in the experimental group, n_clt: total sample in the experimental group, F: female, M: male, NB: non binary, br/min: breath per minute, I/E ratio: inhalation/exhalation ratio, M_age  =  mean age; RCT: randomized controlled trial, HS: healthy subjects, AΝΧ: anxiety, HRV: heart rate variability biofeedback, DEP: depression, TBI: traumatic brain injury, ADHD: attention deficit hyperactivity disorder, PB: phobia, PD: panic disorder, BPD: bipolar disorder, SD: sleep disturbances, PTSD: posttraumatic stress disorder, STAI-S: state-trait anxiety inventory, CIQ: cognitive interference questionnaire, SMS: state mindfulness scale, PANAS: positive and negative affect schedule, HR: heart rate, SCL: skin conductance level, RR: respiration rate, ASI: anxiety sensitivity index, FFS: fear of flying scale, FSB: flying phobia screening questionnaire, MDBF: multidimensional mood questionnaire, VAS: visual analogue scale, MBAS: meditation breath attention score, mDES: modified differential emotions scale, BASS: buddhist affective states scale, MEQ: meditative experiences questionnaire, MDMQ: multidimensional mood questionnaire, ACTA: autonomy and competence in technology adoption questionnaire, PQ: presence questionnaire, FSS: flow short scale, UTAUT: unified theory of acceptance and use of technology questionnaire, PSS-10: 10-item perceived stress scale, CBI: copenhagen burnout inventory, UEQ: user experience questionnaire, PCSS: post-concussion symptom scale, POMS: profile of mood states, VOMS: vestibular/ocular motor screening, SUS: slater–usoh–steed (SUS) presence questionnaire, GEQ: game experience questionnaire, STAIC: state-trait anxiety inventory for children, TRIER-C: trier social stress task for children, FFMQ: five-facet mindfulness questionnaire, MEQ: normative meditative experiences, BSI: brief symptom inventory, FMI: freiburg mindfulness inventory, WHOQOL-BREF: world health organization quality of life scale, BFI: big five inventory, MADRS-S: montgomery–åsberg depression rating scale-self assessment ΒAΙ: beck anxiety inventory, TMS: toronto mindfulness scale, DASS-21: 21-item Depression, Anxiety, Stress Scale, GAD-7: 7-item Generalized Anxiety Disorder, PHQ-9: patient health questionnaire-9, WHO-5: 5-item world health organization well-being index questionnaire, PROMIS: patient-reported outcomes measurement information system, SWEMWBS: short warwick–edinburgh mental well-being scale, SWLS: satisfaction with life scale, SFSS-2: short flow state scale, IPQ: igroup presence questionnaire, TSST: trier social stress task, WEMWBS: warwick–edinburgh mental well-being scale, DASS-21: depression, anxiety, and stress scale-21 items, GAS: goal attainment scaling, PCL-5: posttraumatic checklist-5, BAI: beck anxiety inventory, PSS: perceived stress scale, MASQ: mood & anxiety symptoms questionnaire, CPT: continuous performance task, SSS: stanford sleepiness scale, SPFC: samn–perelli fatigue checklist, MAAS: mindful attention awareness scale, HADS-A/D: hospital anxiety and depression scale, ISI: insomnia severity index, PSAS: pre-sleep arousal scale.

in Appendix A presents the collected data regarding the chosen studies’ characteristics and outcomes. The collected data were used to identify patterns or pieces of information that might be relevant to the research questions of the current study. The collected data were employed to make a narrative synthesis.

2.7. Risk of Bias Assessment

The selected studies were independently assessed by two reviewers, and conflicts were discussed with the third reviewer. The Cochrane Collaboration’s Risk of Bias Version 2 tool (ROB-2) was utilized for the identification of possible bias in the randomized controlled trials. This tool consists of the following five domains of bias (bias in the randomization process, deviations from the planned intervention, bias because of missing outcome data, bias in the measurement of the outcome, and bias in the selection of the reported results). The evaluator can select from the following judgments: low risk, moderate risk, or high risk. For instance, the study can be evaluated as low risk of bias for all domains for this result. The study can be judged to raise some concerns in at least one domain for this result but cannot be at high risk of bias for any domain. The study can be judged to be at high risk of bias in at least one domain for this result. The ROB-2 tool can also propose an overall judgment about the risk of bias arising from each domain, generated by an algorithm and based on answers to the signaling questions [32]. The remaining papers were evaluated with the Risk of Bias for Non-randomized Studies of Interventions tool (ROBINS-I) which consists of the following seven domains of bias: confounding, selection of participants, intervention, divergence from planned intervention, missing data, measurement of outcomes, and selection of the reported results. The risk can be characterized as “low”, “moderate”, “serious”, “critical”, or “no information” [33].

2.8. Studies Selection

After applying the inclusion and exclusion criteria, we identified a total of 845 studies. We removed 195 duplicate papers. The remaining 650 papers were evaluated based on titles and abstracts. A total of 296 studies were removed, and the full-text PDFs of the remaining 354 papers were retrieved. However, it was not possible to retrieve 25 papers. A full-text screening stage followed. The papers that did not meet the predefined methodological requirements were also excluded. After an in-depth processing, a total of 44 studies were selected. The screening operations through the different phases of the systematic review are presented in Figure 1.

2.9. Study Characteristics

The selected studies examined a total of 17,408 subjects who participated in breathing interventions assisted by emerging technologies. Most participants were female. Six studies did not provide information about the gender possibly for anonymity. Most participants were adults. Three studies mainly focused on children and adolescents. Three studies provided no information regarding the age of the participants. The majority of the digitally assisted breathing interventions focused on healthy populations (n = 31). The remaining 13 studies recruited samples with several health issues such as depression and anxiety disorders. The duration of the interventions varied from single-session interventions to annual interventions. As regards the studies’ design, seventeen studies were randomized controlled trials. The remaining 27 studies were non-randomized experimental studies.

The countries where the experimental processes took place were the following: USA (n = 14), Netherlands (n = 6), Australia (n = 2), Canada (n = 2), the UK (n = 3), Germany (n = 6), Korea (n = 2), Italy (n = 3), Israel (n = 1), Switzerland (n = 1), Denmark (n = 1), Sweden (n = 2), and Taiwan (n = 1).

The current review study mainly focused on studies conducted in the last ten years, from 2014 until March 2024. The selected studies revealed a rapid increase in digitally assisted breathing interventions in the years 2019, 2020, and 2021. During this period, COVID-19 influenced people’s respiratory, mental, and emotional health. Thus, it can be hypothesized that there was an increased need for such health programs.

The selected studies mainly focused on slow-paced relaxation breathing techniques (i.e., diaphragmatic breathing). Regarding digital design, the selected studies revealed innovative digital tools and digital designs including chatbots, intelligent coaches, smart applications, virtual reality, biofeedback, and neurofeedback technologies.

2.10. Quality of the Studies

The process of evaluating the selected studies’ quality indicated that the majority of the studies were conducted according to high-quality standards. Regarding the 17 randomized controlled trials, 16 had a low risk of bias, whereas for one study we raised some concerns. As regards the remaining 27 non-randomized studies, the assessment indicated 20 low-risk studies, 5 studies with a moderate risk, and 2 studies with a serious risk.

3. Background Knowledge

3.1. The Type of Breathing Exercise Matters

Breath-control practices require proper modification of various breath-related parameters including (a) the respiration rate, (b) the depth rate of respiration, (c) air velocity (d) inspiratory and expiratory phase, duration, and pause, (e) the respiration rhythm, and (f) primary area of movement (i.e., upper or lower chest and abdomen). Practitioners must acquire the experience and the skills to modify the above parameters, on a case-to-case basis, to achieve the desired outcome [34].

The proper modification of breathing parameters can significantly influence the biomechanical system of the organism. It can also influence the mental and emotional states of humans. Dysregulation in breathing patterns can be a sign of physiological, psychological, or mental dysfunction and may predict pathological conditions [1].

According to the literature, several breathing practices can have a variety of health benefits. Some studies categorize breathing activities as relaxing breathing techniques, excitatory breathing techniques, or warming or cooling breathing techniques [6]. In the current section, we briefly describe several commonly used breathing interventions.

Diaphragmatic breathing utilizes the diaphragm, which divides the chest cavity from the abdomen. This type of breathing maintains the vital capacity of the lungs to the maximum. Diaphragmatic breathing modulates the autonomic nervous system and induces a state of relaxation and mental clarity [4].

In the deep-breathing technique, breathing patterns become slow, long, and deep. Slow breathing enhances the vital capacity of the lungs and induces positive feelings in the trainee [35].

Square breathing refers to a slow-rhythm breathing technique that follows four steps, including breathing in, holding, breathing out, and holding one’s breath. This technique has been shown to improve mental functions including attention. It also acts as a powerful tool for stress management [36].

In alternate-nostril breathing, inhalation is executed from one nostril, whereas the other nostril remains closed. This breathing type oxygenates the blood properly and enhances the feeling of peacefulness [35].

The pursed-lip-breathing technique consists of exhaling through tightly pursed lips and inhaling through the nose with the mouth closed. This technique slows down breathing by increasing exhalation time, allowing more air to be expelled with each breath cycle [37].

Hold-breathing techniques combine deep breathing and breath holds. This type of exercise aims to activate the sympathetic nervous system, increase excitatory hormones and tidal volume, and increase blood flow to the muscles to enhance energy and endurance [13].

Kapalabhati is considered an advanced yogic breathing exercise that improves the lungs’ capacity, oxygenates the blood, and provides trainees with additional energy [35].

Right-nostril breathing increases inner energy and is characterized as heat-generating, because it activates the sympathetic nervous system and elevates the body’s temperature [38].

Sheetali and Sitkari pranayamas lower body temperature and induce a more relaxing mental state. Inhalation occurs through the mouth, with the tongue rolled into a semi-tubular form. Expiration is executed from both nostrils [39]. Figure 2 presents the most commonly used breath-control techniques.

3.2. Breath-Control Skillfulness: It Is a Matter of Metacognitive Training

Breath-control techniques can be described as a set of strategies appropriately designed to enhance the effective management of breathing patterns. They aim to familiarize practitioners with techniques that encourage the conscious manipulation of breath-related parameters such as the rate, rhythm, depth, and breathing pattern to achieve physical, mental, and emotional wellness [6,9].

Research has already revealed that breathing training requires the practitioner to acquire relevant knowledge regarding the complex functioning of breathing. Even more importantly, practitioners should realize that breathing can be under human control and can be used as a powerful strategy for self-regulation [9]. However, systematic practice is a prerequisite to equipping all practitioners with the skills needed to exercise control over respiratory operations. Several researchers describe this type of control as a form of metacognitive control [6,10].

Metacognition refers to awareness-raising skills that allow individuals to voluntarily monitor and control psychophysiological operations as a means to induce and maintain optimal internal stability [40]. Metacognitive training is considered an inherent characteristic of breath-control practices [10]. In such training interventions, practitioners learn to reflect on and understand the healing power of breathing. In addition, they are systematically trained to employ attention as a tool for regulating breathing operations. They are also trained to effortlessly pay attention to breathing, monitor breathing operations, and make regulations and adjustments accordingly (i.e., the rhythm and the depth of respiration) to achieve an optimal breathing pattern that fits with the internal resources and the external demands of a given task. Practitioners are also trained to voluntarily get themselves into a state of increased awareness that permits them to recognize all the latent conditions, either internal or external, that tend to dysregulate breathing. Moreover, metacognition in breathing enables practitioners to make wise judgments about their breathing status. Last but not least, practitioners learn to transfer all these breath-control skills into real-life situations [6,10,11].

According to Figure 3, metacognitive training constitutes an inherent component of breath-control practices. Metacognition provides trainees with a wide range of skills (i.e., observation, regulation, and adaptation) which in turn allow them to effectively manage breathing operations according to the desired goals [10].

3.3. The Healing Power of Breathing: Benefits in Physiological, Mental, and Emotional Health

3.3.1. The Impact of Breath-Control Interventions in Physiological Parameters: The Building Elements of Mental and Emotional Health

Research has demonstrated that systematic breathing training is associated with a wide range of improvements in respiration health, including the expansion of the lungs’ capacity and functionality. For instance, Jansang et al. [37] investigated different types of breathing exercises. It was revealed that breathing training effectively strengthened respiratory muscles and increased the volume of the air that the lungs could hold, leading to an enhanced oxygenation of the body’s tissues and organs. In a similar study, Budhi et al. [41] enhanced respiratory efficiency was revealed after breathing training with improvements in the elastic qualities of the lungs.

Research also indicates that breath-control interventions ensure a balanced exchange between oxygen and carbon dioxide which in turn influences overall health, including positive mental functioning and emotional stability [9]. Paced breathing (6 breaths/minute) has been proven to restore oxygen and carbon dioxide balance, improving homeostasis, mental clarity, and emotional stability [42].

Apart from the oxygen and carbon dioxide balance, proper breathing allows the regulation of acid–base balance, which in turn is associated with reductions in the inflammatory indicators that increase anxiety and an increase in the anti-inflammatory markers that induce relaxation [12,13]. By enhancing the antioxidant status, breathing training can prevent oxidative stress and protect the immune system [12,13].

Thus, it seems that systematic practice with breath-control exercises can prevent neurodegeneration, which is related to the development of various disorders including mental and emotional problems, and most importantly boost neurogenesis (i.e., hippocampal neurogenesis) which is linked with good mental health and well-being [43,44]. Research indicates, for instance, that specific types of breathing techniques such as breath-holding techniques can increase growth factors including neurotrophins such as brain-derived neurotrophic factor (BDNF) and vascular endothelial growth factor (VEGF) [45].

Research provides evidence that breath-control interventions like pranayama balance the functioning of the autonomous nervous system. More specifically, autonomous nervous systems consist of the sympathetic and parasympathetic nervous subsystems that compete with each other to induce alertness or relaxation accordingly [40]. The sympathetic system prepares the organisms for “fight-or-flight” responses. On the contrary, the parasympathetic system is responsible for inducing a state of relaxation [40]. Studies indicate that inspiration is closely linked with the sympathetic nervous system and expiration with the parasympathetic nervous system [10].

Research also indicates that breathing interventions can significantly influence the balance between the sympathetic and parasympathetic nervous systems. More specifically, Pal et al. (2004) [46] found that slow breathing can help people have better control over the functions of the autonomous nervous system. In addition, it was revealed that slow-paced breathing can both decrease the functioning of the sympathetic nervous system and stimulate parasympathetic nervous system activity. In another study conducted by Komori (2018) [47], it was found that breathing techniques with prolonged expiration stimulate parasympathetic nervous functions, whereas rapid breathing suppresses parasympathetic nervous function and elevates stress-related responses. Pal et al., 2014 [48] found that people who practiced right-nostril breathing activated their sympathetic nervous system, whereas subjects who practiced left-nostril breathing activated their parasympathetic nervous system.

By using breath-control exercises such as diaphragmatic breathing, practitioners can improve the balance of hormones and neurotransmitters that influence cognitive functions and mood. Studies have shown that relaxation breathing intervention reduces cortisol, the hormone that is closely related to anxiety, panic, and depression [14,49]. Martarelli et al. [49] found that breathing training not only reduces cortisol but also increases melatonin which competes against cortisol and is responsible for preventing oxidate stress, repairing DNA, boosting good mood, sleep, and homeostasis balance. In a study conducted by Jerath et al. (2015) [50], it was found that breathing techniques stimulate gamma-aminobutyric acid (GABA) pathways that induce a state of calmness.

3.3.2. The Impact of Breathing Exercises on Cognition, Higher Mental Abilities, and Consciousness

Attention constitutes one of the most vital cognitive functions [11]. According to many researchers, attention is considered a higher mental ability because it allows subjects to utilize attention as a means of self-regulation [10]. Melnychuk et al. [51] provided evidence that respiration and attention work in synergy as a coupled dynamical system. According to their study, fluctuations in breathing can influence attention. Disturbances in breathing lead to irregularities in attention. On the contrary, regulations in breathing can restore attentional dysfunctions. According to their conclusions, breath-control practices can achieve optimal synchronicity between attention and respiration.

Indeed, Ma et al. [14] found that an 8-week diaphragmatic breathing training improved sustained attention in a total of 40 participants. Kiselev et al. [52] found that diaphragmatic breathing can improve attention regulation among people with attention deficit hyperactivity disorder. Telles et al. (2017) [53] revealed that alternate-nostril breathing can improve attention without hyperactivation of the sympathetic nervous system. The study conducted by Telles et al. (2013) [54] indicated that high-frequency yoga breathing can improve attentional flexibility.

Zelano et al. (2016) [55] found that proper breathing improves the synchronization between brain networks that influence attention, memory, emotion, and consciousness. In addition, it was observed that inhalation was associated with improved recognition of fearful faces and better recollection of images. Wang et al. [56] demonstrated that breathing training can improve short-term memory and attention among people with cognitive impairments. Naveen (1997) [57] demonstrated that uni-nostril breathing training improved spatial memory among one hundred children.

It is noteworthy that studies outline that breathing can alter mental states and can induce altered states of consciousness [6,44]. Miller et al. [58] investigated the impact of a breathing technique known as holotropic breathwork™ in a total of 20 subjects. This technique involves subjects in a voluntary, prolonged deep over-breathing procedure. The results showed alterations in the self-transcendence scale. Vialatte et al. (2009) [59] employed a breath-control technique known as Bhramari Pranayama. The EEG data of all the participants indicated a significant increase in high frequencies (gamma oscillations). The participants reported that they experienced a feeling of extreme peacefulness and blissfulness.

3.3.3. The Impact of Breathing Exercises on Emotional Health and Resilience

Studies provide a growing body of evidence that breathing intervention can help people to improve their emotional health. Beblo et al. [60] demonstrated that breathing exercises can help healthy subjects to more effectively regulate emotions, especially those with negative emotional valence. Novaes et al. [61] revealed that pranayama practices can modulate the activity of brain regions related to emotional processing and emotional awareness. Moreover, it was found that this type of yoga breathing exercise can decrease negative emotions and reduce anxiety. He et al. [62] found that breathing can help subjects to be aware of internal sensations and be more empathetic.

Scroggins et al. [63] demonstrated that relaxing breathing helped a child with Asperger syndrome who exhibited behavioral problems to improve social skills and emotional well-being. Sharma et al. (2016) [64] conducted a randomized controlled trial to examine the impact of breathing training on depression. The results indicated that breathing helped subjects with depression to better regulate mood.

DeGraves et al. [65] investigated whether a breathing intervention could help employees reduce stress and increase resilience. Indeed, the results indicated breathing interventions are promising strategies for promoting resilience.

Table 3. Summary of the benefits of breath-control training in physical, mental and emotional health.

Benefits of Breath-Control Interventions in Physical, Mental, and Emotional Health	Reference
Improvements in respiratory health (i.e., strengthened muscles, lungs’ capacity, endurance, and balanced gas exchange)	[37,41,64]
Ιmprovements in blood’s acid–base balances and blood pH	[13]
Improvements in voluntary control over the autonomic nervous system	[46,47,48]
Reductions in inflammatory indicators, free radicals, oxidative stress and increase in anti-inflammatory mediators, growth factors, and antioxidant status	[6,12,13,49]
Modifications in hormones and neurotransmitters related to anxiety, calmness, and good mood (i.e., cortisol, melatonin, and GABA)	[13,14,49]
Benefits of breath-control interventions in cognition, higher mental abilities, and consciousness	[6,58,59]
Improvements in attentional operations (concentration, vigilance, attentional regulation, and attentional flexibility)	[14,53,54]
Improvements in memory skills (i.e., spatial memory).	[55,57]
Improvements in higher mental abilities (i.e., executive functions and metacognitive abilities)	[52]
Alterations in mental states and induces altered states of consciousness (i.e., self-transcendence)	[58,59]
Benefits of breath-control interventions in emotional health, emotional intelligence skills, and resilience	[14,60,61,62,63,64,65]
Improvements in emotional regulation	[60]
Improvements in emotional awareness	[61]
Improvements in empathy	[62]
Improvements in stress management	[14,61]
Improvements in resilience	[65]

below summarizes the main benefits of breath-control training in physical, mental, and emotional health.

3.4. The Potential of Emerging Technologies in Assisting Breathing Interventions

3.4.1. The Potential of Artificial Intelligence in Assisting Breathing Interventions

Research indicates that AI has a significant potential to revolutionize mental health care by making mental health training programs even more accessible, accurate, personalized, interactive, and effective for populations with different needs and capabilities [20]. AI algorithms can assist breathing interventions by analyzing breath-related data (i.e., respiratory patterns and lung capacity), making assessments of respiratory operations, providing real-time feedback, recording progress, and keeping medical history [66]. By collecting respiration-related and other health-related data, AI can make ongoing adaptations to follow the users’ progress. AI can be utilized in the process of designing breathing training interventions [20]. More specifically, it can make use of large-scale respiration-related databases and recognize patterns and correlations. Thus, it can contribute to the process of developing new breathing training protocols based on accurate and evidence-based information [67].

Users can also ask AI assistants for help, instructions, and suggestions [68]. Virtual coaches are embodied conversation agents that simulate the verbal and non-verbal behaviors of a human. They intend to promote users’ training, providing constant guidance and support [69]. A significant advantage is that AI-powered tools can encourage users to continue their practice, by identifying users’ personal preferences, providing appropriate triggers, and using positive reinforcement and rewards [66]. Therapists can also remotely monitor short-term as well as the long-term training progress. AI can be a significant solution for populations who deal with geographical constraints, limitations due to physical disabilities, or psychological factors (i.e., disliking in-person interactions) [70]. It can also support group training or community training. AI can be combined with other technologies such as biofeedback devices including a wide range of low-cost sensors and wearables [71]. AI-driven biofeedback technologies can increase the amount of the collected physiological data which in turn can elevate the quality of the breathing intervention. AI-driven virtual reality technologies can provide extraordinary and motivational breathing training experiences capable of fitting with the trainee’s needs [66].

AI can take various forms. Chatbots, known as conversational agents, constitute human–machine interactive interfaces allowing the interaction of users with a computer program capable of creating meaningful text- or speech-based conversations [21]. Chatbots can be used to provide users with instructions and guidance, teach users breath-related skills, and provide them with continuous coaching, positive reinforcement, and reminders during training [72]. ChatGPT, Gemini, and Co-pilot, for instance, utilize a sophisticated language model that employs deep learning techniques to generate human-like replies that resemble those of a human [68,73]. They can provide users with relevant and simplified information about breathing techniques and their suitability for specific health conditions. They can clearly explain to users step-by-step how to execute a breathing exercise and analyze the benefits or the possible risks of practicing a specific type of breathing exercise [74]. It can also design training programs and answer questions about breathing interventions. ChatGPT can communicate with wearable devices (i.e., smartwatches or fitness trackers), maximizing the training benefits [73].

Smartphone applications can offer innovative solutions for breathing training. Users can find constant support anywhere and at any moment. They are cost-effective tools and user-friendly. In general, smartphones are increasingly seen as versatile m-health instruments for treatment and training, and some authors predict that the mobile phone will emerge as the preferred personal coach for the 21st century. However, mobile apps for breathing training lack formal evaluation in the literature [27].

3.4.2. The Potential of Immersive Technologies in Assisting Breathing Interventions

Virtual reality refers to interactive three-dimensional systems that have the potential to replace a real-world environment with a simulated one, immersing users in a computer-generated environment via the use of interactive devices (i.e., goggles, headsets, gloves, or body suits) that allow the exchange of information [75]. Various low-cost devices can be used in combination with other technologies such as smartphones, biofeedback, and neurofeedback devices [76]. They can be used among the masses, but at the same time, they can be programmed to fit the different needs of trainees [18].

Immersive technologies are commonly used to induce a state of presence, confidence, and relaxation. In addition, the peaceful landscapes presented in virtual environments are considered a significant factor in restoring attentional operations and reducing mental fatigue [10,76].

Recent studies outline the benefits of integrating VR and biofeedback sensors in breathing intervention protocols [77]. Sensors can monitor users’ physiological responses, while VR can provide visual and auditory cues within the virtual environment. In addition, users can use biofeedback as a tool for self-regulation within the VR environment [10,78,79].

VR offers digital environments specifically designed to promote calmness (i.e., peaceful landscapes or serene natural environments) which can facilitate deep-breathing training [18]. VR can employ gamified elements that can make breathing training more engaging and enjoyable [78,79]. The use of virtual characters known as avatars can also boost the training process, allowing users to experience a feeling of embodiment [80]. VR environments have the advantage of encouraging intuitive learning, which allows subjects to effortlessly train their self-regulation skills [81].

The metaverse, although in its early stages, holds promise for applying innovative breathing interventions. The metaverse refers to an immersive virtual environment where users interact in real-time both with other users as well as digital items. This multiuser ecosystem integrates physical reality with digital virtuality and works as a networked and socially interactive environment that promotes real-time, seamless embodied user contact with digital artifacts as well as continuously evolving interactions with them [82].

3.4.3. The Potential of Wearables, Biofeedback, and Non-Invasive Neurofeedback Technologies in Assisting Breathing Interventions

Wearable technologies refer to devices that can be worn on specific body parts (e.g., the wrist, hand, or neck), as they consist of sensors that provide continuous measurement of physiological signals (i.e., heart rate, temperature, and galvanic skin response) [83]. Wearables not only identify physiological signals but also provide users with real-time feedback regarding non-conscious or hard-to-identify psychophysiological functions (i.e., respiratory rate) [24,71,83]. Wearables also provide reminders, track long-term changes, boost motivation, and cultivate positive beliefs about human potential for achieving new goals [71].

Biofeedback provides users the advantage of raising awareness over their physiological responses when executing breathing exercises [84]. The presentation of visualizations can provide significant feedback on how to respond when practicing various breathing techniques [24]. Respiratory biofeedback can help the trainees to maintain focus on breathing [84]. Equipped with information provided by biofeedback, trainees can also learn how to self-regulate and adjust their respiratory operations to achieve the desired state of being [77,85]. Moreover, biofeedback can increase trainees’ motivation to maintain focus and effort over training [84].

The brain is capable of pulsating at many different frequencies, with each frequency level accurately reflecting quantifiable states of mental and emotional health [86]. Neurofeedback technologies refer to computer-based tools that directly track, process, and evaluate brain-specific neuro-data, which are converted into interpretable outputs [87]. The neurofeedback techniques are classified into the following three categories: non-invasive, semi-invasive, and invasive [88].

In this review study, we examine non-invasive methods that use scalp-mounted sensors to detect the electrical potentials generated by the brain (EEG). Non-invasive wearable devices such as headbands can measure brainwaves with the use of electroencephalography sensors [25]. These neurofeedback technologies are user-friendly, portable, and low-cost. It is also easy to link them to free applications downloaded on the user’s smartphone [26]. As mentioned, smartphones or virtual worlds are used to transfer the collected feedback to users in the form of visual or auditory cues, motivating users to be engaged in the training tasks [26,89]. Training protocols based on neurofeedback can be customized to fit the trainees’ needs, goals, and capabilities [26]. Training sessions can be tailored based on factors such as the individual’s baseline brain activity, specific breathing challenges, and desired outcomes, allowing for personalized intervention plans [26,89]. Neurofeedback-based training can focus on brain regions and neural networks that are closely linked with respiration operations. By measuring brain waves and providing users with relevant feedback, this type of technology can gradually train subjects to be aware of and voluntarily control brain activity to achieve regulation in physiological operations including respiration [87]. Thus, trainees can gradually develop better control skills over their breathing and, as a result, promote mental and emotional health. Repeated practice can help individuals develop lasting skills, allowing them to internalize the techniques instructed through neurofeedback training, and apply them in various situations to promote relaxation, emotional well-being, and optimal breathing patterns. Figure 4 illustrates the emerging technologies with a high potential to assist breathing training, while

Table 4. Summary of the potential benefits of emerging technologies in assisting breathing interventions.

Emerging Technologies	Potential Benefits
Artificial intelligence-based tools (i.e., conversational agents and ChatGPT)	Unlimited accessibility, interaction, accuracy, reliability, customization, personalized coaching, intelligent coaches, multilingual support, human-like responses, training data, intervention planning, positive reinforcement, progress tracking, scalable and adaptive training, anonymity, group training, interoperability, confidentiality, distance training, 24/7 support, user-friendly, cost-effective solution [20,66,67,68,69,70,72,73,74]
Immersive technologies (i.e., virtual reality, metaverse)	Immersive, training via simulations, various scenarios, calming environments, interactive feedback, personalization, engaging, multisensory, exposure therapy techniques, intuitive training, multimodal feedback, induce a state of presence, attentional restoration, attract attention, direct attention, distracts from unhelpful stimuli, group sessions, anonymity, gaming elements, virtual coaches, combined with other technologies, remote training, data collection, data analysis [10,76,77,78,79,80,81,82]
Biofeedback and neurofeedback technologies (i.e., wearables, sensors)	Improve awareness of physiological operations, train self-regulation, and adaptation based on the provided feedback, provide customized training protocols, real-time quantitative feedback, offer positive reinforcement, and target specific brain regions or neural networks [24,25,26,83,84,85,86,87,88,89]

presents the main advantages that emerging technologies can offer in breathing interventions.

Although emerging technologies provide a wide range of potential benefits for assisting mental and emotional health interventions such as breathing interventions, their potential role in breathing interventions is yet underexplored. Thus, there is a need to collect, summarize, and synthesize the existing experimental studies that investigate the effectiveness of emerging technologies in breathing training. The current review also investigates the effectiveness of the synergy between emerging technologies and breathing training on various aspects of mental and emotional health.

4. Main Findings: Emerging Technologies in Breathing Intervention for Mental and Emotional Health

4.1. Artificial Intelligence in Breathing Interventions: Conversational Agents blended with Biofeedback, Intelligent Coaches, Sensors, Robotics, and Gamification

The assistance of AI chatbots in breathing training that provided instructions and guidance was found to positively influence users’ mental well-being, especially for populations who are at risk of mental health problems. Potts et al. (2023) [90], for instance, investigated the effectiveness of a multilingual AI chatbot in a sample of 348 subjects to promote relaxation, mental health, and general wellness among people living in rural areas under stressful conditions. The chatbot employed instructional content and breathing techniques. After 12 weeks of training, 348 participants significantly improved their psychological balance.

The selected studies indicated that AI-based breathing training can also be effective for people with anxiety and emotional disorders. Youper is an AI chatbot that is designed to train users to monitor, recognize, and manage their respiration along with their thoughts and emotions. The users can receive mindful breathing training combined with cognitive behavioral therapy (CBT) and acceptance and commitment therapy (ACT). Mehta et al. [91] analyzed the data of 4517 users based on measurements of anxiety and depression. Users’ data indicated improvements in emotional regulation.

Gabrielli et al. [92] assessed an AI psychoeducational chatbot for reducing stress among university students. The Atena chatbot provided breathing instructions to the users. It was revealed that the bot significantly helped 71 students regulate breathing in stressful situations. Participants reported that they found the AI breathing chatbot an effective intervention for novices in breath-control techniques.

AI-powered breathing coaches were found to be promising tools in breathing interventions. An AI conversational agent was developed to take the role of a virtual breathing coach by Shamekhi et al. [69]. The intelligent breathing coach could interact with the users and adapt according to the users’ respiration behavior, taking into account the input from the sensors used. It was found that the virtual coach had a significant influence on the users’ breathing behavior as well as their mood. In addition, they improved their awareness and control over physiological operations. They could better regulate breathing to achieve a state of relaxation. It is noteworthy that most participants found the responses as well as the voice of the coach acceptable and relaxing, whereas a limited number of participants characterized the voice as too robotic and distracting.

Vertsberger et al. [93] assessed an AI-based personal companion that provided 10,387 adolescents with both breathing instructions as well as reflective conversations according to the users’ preferences. The personal AI companion taught users how to breathe through their noses, using their diaphragm, and note their posture. The findings revealed improvements in emotional and psychological balance.

Empathetic conversational agents were found to positively influence trainees during breathing training. Wysa is a conversational agent, capable of delivering breathing training. It has the advantage of providing users with empathetic responses. The AI app also employs written and structured conversations to respond to trainees’ emotions and suggest self-regulation strategies based on mindfulness, cognitive behavioral therapy, and positive psychology. Leo et al. [94], using a sample of 61 subjects with symptoms of pain, anxiety, and depression, found that deep-breathing training with the support of a conversational agent could effectively relieve pain and reduce anxiety and depression. The researcher concluded that the intervention helped users develop self-regulation skills resulting in improvements in physical, mental, and emotional health.

Breathcoach is a smart biofeedback VR game for breathing training. It employs sensors on a smartwatch and a smartphone. More specifically, it provides continuous assessments of key biosignals and automatically calculates the optimal breathing rate based on measurements. After six sessions, the participants could better control attention and manage stress. The researchers concluded that a smart breath coach could be more effective than traditional breathing training, especially for the improvement in cognitive function and stress management [95].

Breathing robots were among the smart devices used in breath-control interventions. Støre et al. (2021) [96] examined the effectiveness of a robot-assisted breathing intervention on insomnia, anxiety, depression, and learning difficulties. The Somnox sleep robot resembles a bean-shaped cushion. Users hug the robot which, in turn, gives them physical and auditive guidance to regulate their breathing rate. Forty-four participants were randomized to a 3-week intervention program with the robot (n = 22) or a waitlist-control group (n = 22). The effect of the sleep robot on the participants’ insomnia, anxiety, and depression were not statistically significant. The researcher concluded that a 3-week intervention with daily at-home use of the robot was not found to be an effective method to relieve the symptom burden in adults with insomnia.

In a similar study, Asadi et al. [97] employed a soft robot to help 28 trainees improve their breathing patterns by synchronizing their breathing rhythm with the robot’s rhythm. The robots simulated breathing. During the training, data were gathered via two belts and an EEG device. The participants in the experimental condition could better regulate their breathing, slowing down their breathing rate, a finding that indicates reduced stress levels. The findings also indicated increased emotional valence for participants in the robot-assisted condition.

Table 5. A summary of the types of AI-based breathing interventions and the main results.

AI-Based Breathing Intervention	Mental and Emotional Health Benefits	Reference
Mindful breathing via multilingual chatbot	Mental wellness	[90]
AI-powered smartphone breathing app	Emotional regulation, cognitive control and adaptive behavior, self-regulation of physiological operations	[67,91]
Mindful breathing via psychoeducational chatbot	Self and emotional awareness, negative thoughts regulation	[92]
Mindful breathing via AI-powered conversational agent via smartphone	Emotional and psychological flexibility	[93]
Deep breathing via AI-based chatbot	Mood and anxiety regulation, relaxation, attentional control, body awareness	[69,94]
Smart breathing training system supported by smartphone-based VR and sensors on a smartwatch	Improved cognitive functions, stress reduction, improved attention	[95]
Deep breathing via breathing robots	Positive emotions, lower stress, improved regulation of breathing	[96,97]

and Figure 5 present a summary of AI-based breathing interventions through the lens of skills development.

4.2. Immersive Technologies in Breathing Interventions: Virtual, Augmented, and Mixed Reality Blended with Biofeedback, Gamification, and Virtual Agents

Studies indicated that VR-based breathing interventions can be more effective than conventional interventions in modifying subjective feelings (i.e., pain) and synchronizing internal with external attention. Hu et al. (2021) [98] compared traditional mindful breathing with VR breathing in a sample of 40 healthy subjects. In the VR breathing group, 3D lungs were synchronized with the participants’ breathing cycles, providing them with an immersive visual–auditory exteroception of their breathing. The results indicated that both interventions had analgesic effects. However, two distinct analgesic mechanisms were identified. More specifically, traditional breathing practice altered sensory interoception, while VR breathing modulated sensory exteroception.

Adolescents often complain about experiencing increased tension and insomnia, which in turn increases the risk of physical and mental disturbances. VR breathing interventions were found to be an effective strategy for promoting physiological downregulation and resilience among adolescents with such symptoms. Yüksel et al. (2020) [99], for instance, implemented a VR-based slow-breathing program among 29 adolescents. The results showed improvements in mental relaxation and reductions in mind-wandering. Waller et al. [100] randomly allocated 82 undergraduate students to two guided mindful breathing sessions, one VR-assisted and the other non-VR-assisted. A focused breathing intervention in virtual reality increased students’ positive emotions. More specifically, they experienced a heightened connectedness that in turn allowed them to better regulate stress, mind-wandering, and mental fatigue.

Studies indicated that VR-based breathing can help subjects with different disorders. Seol et al. [101] employed a VR-based relaxation and breathing intervention for dealing with panic disorder. The participants were asked to hold a virtual heart in one hand, while a haptic device provided them with the simulated kinesthetic feedback of the beating heart. The aural and textual instructions were given to enable trainees to consciously regulate their breathing. The five participants in the training session reported that they could better regulate their mood under stressful conditions.

VR-based breathing was found to be beneficial for decreasing aggressiveness and developing more flexible behaviors among people with depression and bipolar disorder. In a study conducted by Ilioudi et al., 2023 [102], mindful breathing training in a VR calm room was compared with staying in a quiet room. The results indicated that the experimental group was more able to relax and develop adaptive behaviors [102].

Digitally assisted breathing interventions were found to be effective for people with health problems. Cook et al. [103] examined the effectiveness of a virtual reality-based breathing intervention among 15 young people with concussion. The subjects took part in brief, paced, deep-breathing exercises with the assistance of a VR headset. Nearly all the trainees (93%) could stay focused. In addition, they reported that stress, tension, and mental fatigue were much more manageable. Kojic et al. [104] outlined that some participants find it more difficult to synchronize respiration with visual cues in VR

Promising effects were found after the use of VR breathing blended with exposure therapy. VR and breathing training helped individuals with phobias better manage stressful stimuli. Shiban et al. [105] randomly divided twenty-nine individuals with phobias into the experimental group, which received a VR exposure treatment accompanied by diaphragmatic breathing, whereas the control group received the VR exposure therapy without the support of breathing training. The results demonstrated that the VR breathing condition improved participants’ self-control skills.

Immersive technologies, combined with biofeedback, were found to have a positive impact on breathing training. More specifically, Blum et al. (2019) [24] compared immersive VR-based nature scenery breathing training combined with biofeedback and slow-paced breathing with a conventional biofeedback intervention. The results showed that the participants in the VR-based biofeedback condition demonstrated better outcomes in terms of heart coherence, perceived anxiety, relaxation, and self-efficacy. The participants were more able to pay attention and generally regulate attentional operations with flexibility, achieving a significant reduction in distraction and mind-wandering.

Similarly, Blum et al. 2020 [106] compared the effects of a focused breathing condition with a VR-based biofeedback breathing intervention. It was revealed that the VR breathing biofeedback resulted in heightened breath awareness. The authors outlined that breathing can act as an anchor for improving focused attention, while VR can effortlessly direct attention to breathing. The participants also reported that biofeedback as well as the feedback from the abdominal movements significantly helped them to maintain attention to breathing.

Rockstroh et al. (2019) [76] compared the effectiveness of VR-based biofeedback with conventional slow-breathing training. Both interventions were effective in increasing heart rate variability (HRV). Kluge et al. [107] evaluated a VR-based biofeedback intervention to enhance breath-control skills and reduce anxiety among 30 healthy adults. After three sessions, the participants were more able to be aware of and control breathing in stressful situations. Participants also reported that their attention improved.

VR-based biofeedback breathing interventions may be more effective than conventional breathing interventions or breathing interventions that employ only biofeedback or virtual reality breathing without biofeedback. Weibel et al. [108] recruited 107 healthy subjects to participate in a four-condition experiment. The conditions were the following: heart rate variability biofeedback (BF), HRV-BF via a head-mounted display (HMD), conventional paced-breathing without feedback on a screen, and slow-paced breathing with the support of HMD. The VR BF condition immersed trainees in a mountainous landscape. They experienced the VR environment from the perspective of sitting on a large tree trunk and looking at a grassy meadow. The soundscape included bird sounds and waterfall sounds. A breathing pacer was located in the middle of the meadow. It was found that all the techniques and digital designs were effective. The brighter and more colorful the environment, the more effectively the breathing exercise was executed. The results indicated that all the techniques and digital designs were effective. However, the most effective intervention was the VR-based biofeedback breathing condition. More specifically, beneficial effects were found in terms of cardiac coherence, mood regulation, stress management, and resilience.

It is noteworthy that VR biofeedback breathing is not always an effective combination. Tinga et al. [109] compared the effectiveness of a VR biofeedback condition to a no-biofeedback condition in lowering subjective and objective arousal after stress. Participants were presented with a breathing session in VR while subjective and objective arousal were assessed via electrocardiography (ECG) and electroencephalography (EEG). The results showed that VR effectively reduced stress. However, biofeedback did not reduce arousal as expected.

The employment of virtual characters that imitate the trainee’s actions during training was found to be effective in a VR-based breathing intervention. Lan et al. [16] employed a virtual avatar of the practitioners in a slow-breathing VR-based multimodal biofeedback training to increase participants’ breath awareness and motivation for systematic practice. The study showed that the trainees improved their breath awareness, and as a result, they could achieve a more slow and stable breathing rate. In addition, the Stroop task revealed reductions in mind-wandering, improvements in the allocation of attentional resources, greater concentration, and more accurate reactions. Similarly, the use of a virtual breathing avatar from a first-person perspective allowed trainees to regulate their respiratory rate and improve sustained attention [80].

Virtual reality, biofeedback, and gamification in breathing training were found to be effective for increasing internal control, self-regulation, stress management, attentional control, and intrinsic motivation. More specifically, Weerdmeester et al. (2021) [79] recruited 86 children and examined the effectiveness of a VR-based biofeedback video game in breathing training. The participants were immersed in a peaceful environment that aimed to enhance their sense of embodiment, intuition, and desire for exploration. The intervention also exposed participants to mildly stressful situations (i.e., dark caves) to promote trainees’ self-regulation skills. Trainees could participate in the VR game with the power of their deep breathing. The results indicated that VR biofeedback gaming improved the sense of internal control and engagement. Rooij et al. (2016) [81] employed the same intervention, and it was confirmed that VR gaming biofeedback breathing training can raise trainees’ breath awareness, allowing the conscious regulation of breathing. It was also revealed that the development of breath-control skills can improve trainees’ attention.

Results also indicated that breathing with VR games can induce higher mental states. Sra et al. [110] explored the effectiveness of employing breathing as a training element in virtual reality games. Sixteen young participants reported a higher sense of satisfaction and engagement. In addition, the results indicated that improvements in breath-control abilities allowed participants to enter a state of flow that improved attentional operations and consciousness.

Behavioral control was another skill developed after breathing training in VR games. Rockstroh et al. (2021) [77] implemented a VR-based respiratory biofeedback game intervention to train diaphragmatic breathing. The intervention improved participants’ breath awareness, stress regulation, resilience, and self-efficacy. In addition, the participants reported that the experience was pleasant. Similarly, mindful breathing training with VR gaming and biofeedback was found to assist young people with ADHD. The players were asked to explore an underwater virtual world with the use of their respiration. After six sessions, participants could better control anxiety and aggressive behaviors [111].

VR breathing was found to be an effective strategy not only for relaxation but also for alertness. A ten-session VR-based action game with biofeedback and slow diaphragmatic breathing was implemented among nine police officers for four weeks. The VR game evaluated response inhibition in high-arousal situations. The results indicated that each time trainees effectively controlled their breathing, their peripheral vision improved, resulting in accurate responses. In addition, they could make fast and accurate decisions. The results also showed increased attention regulation during breathing training [78].

Kang et al. [112] developed and evaluated a VR-based biofeedback breathing training system with gamification elements and a soft stretch sensor among 50 adults. The unique characteristic of the intervention was the use of thoracic expansion from inspiration using a sensor instead of a mouthpiece. The visual feedback allowed trainees to observe their breathing and make accurate judgments about their breathing patterns. In addition, they received audiovisual feedback as a reward. The trainees characterized the training as authentic and very funny. However, several participants found the headset device heavy and inconvenient.

Exergaming, combined with VR biofeedback breathing interventions, was found to be an alternative method with positive outcomes for trainees. Kojic et al. [104] simulated a rowing exercise providing users with constant visual feedback (i.e., animation of lungs) regarding their breathing patterns. The main objective of the game was to maintain focus on breathing rhythm. The results indicated improvements in terms of flow, sympathy, and helpfulness. The results suggested that integrating breathing biofeedback in VR exergames enhances user experience, motivation, and engagement, despite the possible challenges in fine-tuning breathing with visual cues, indicating the need for additional research on different sports and biofeedback modalities.

Immersion and interaction in mixed reality-based breathing training were found to increase trainees’ curiosity and motivation for developing self-regulation abilities. Roo et al. [113] recruited 12 subjects to evaluate a mixed reality breathing and relaxation system that was supported by augmented reality. It was a multi-modal, tangible artifact that resembled an AR sandbox. A sandbox was connected to physiological sensors to create a relaxing interactive experience where the participants could shape their world that evolved according to their breathing and heart rate. The user had the choice to be immersed inside this world via a head-mounted display to focus attention on breathing and achieve a state of peacefulness. The results indicated that the mixed reality breathing intervention elevated trainees’ curiosity and motivation. Most importantly, subjects were connected to their physiological operation of respiration to voluntarily achieve a state of relaxation.

Table 6. A summary of the types of breathing interventions based on immersive technologies and the main results.

Breathing Interventions Assisted by Immersive Technologies	Mental and Emotional Health Benefits	Reference
VR-based app for mindful breathing	Self-relaxation, stress regulation, reduced tension, and fatigue, improved attention, mental clarity	[99,102,103]
Slow breathing via VR HRV-biofeedback	Heart coherence, resilience, self-regulation, attention regulation, mood regulation, self-awareness, relaxation, stress management, self-motivation, breathing control skills, greater focus, improved time reaction, less mind-wandering	[24,26,76,101,106,108,109,112]
Diaphragmatic breathing via VR games based on biofeedback	Increased awareness of physiological operations, attentional control, breath-control skills, motivation, response inhibition under stress, improved performance in decision-making contexts, increased sense of internal control, engagement, concentration, self-monitoring, self-regulation, stress regulation	[77,78,79,81,98,104,110,111]
Breathing awareness training via 360° video-guided meditation via HMD	Experience of awe	[100]
VR-based exposure treatment accompanied by diaphragmatic breathing	Emotional regulation, stress management	[105]
MR-based tangible interface for mindful breathing	Increased relaxation and state mindfulness	[113]

and Figure 6 present a summary of the breathing interventions based on immersive technologies through the lens of skills development.

4.3. Neurotechnologies and Wearables in Breathing Interventions: Biofeedback, Neurofeedback, Non-Invasive Wearables, Brain-Sensing Headbands Assisted by Mobiles, and Gamification Elements

Smart mobile applications with the support of sensors were found to be an effective solution for achieving physiological recovery and improving mental capacity. Chelidoni et al. (2020) [114] evaluated the effectiveness of a smart app-based breathing intervention in improving physiological recovery using a sample of 75 employees who experienced cognitive and emotional stress. Heart rate variability (HRV) changes were measured to indicate cardiovascular health, self-regulation strength, emotional stability, and psychological adaptability. The app offered guided breathing adjusted to the trainees’ heart rate. Significant differences were found in the app-based intervention group compared to the control group. The authors concluded that biofeedback breathing interventions delivered by smart devices can support stress recovery and boost positive health. The same application was evaluated by Ponzo et al. [115] which was paired with a wearable device among university students with anxiety. Data, as regards physical activity, sleep quality, and heart rate were gathered via a wrist-worn wearable device. Feedback was received via the mobile device. After 4 weeks of training, the university students were more able to regulate stress.

Gamified biofeedback was found to improve adaptive skills among healthy adults. More specifically, Shih et al. [116] implemented gamified biofeedback breathing training with the support of a mobile application among 43 healthy subjects. The trainees underwent slow-paced biofeedback training. The smartphone’s microphone detected breathing phases in real time. Afterward, the trainees could receive a gamified biofeedback to adapt their breathing. The participants reported that the training helped them to relax and feel self-controlled.

Smartphone applications, combined with innovative sensors, were found to promote stress management and mental wellness. Smith et al. [71] employed a clothing-attached device that monitors respiratory effort and physical activity. A vibrotactile motor was embedded in the device to provide silent feedback to the trainees without the constant support of a smartphone device to receive feedback. The smartphone app provided users with auditory guided breathing training as well as notifications about their progress. The results demonstrated improvements in stress regulation and emotional regulation.

Al Rumon et al. [117] designed and evaluated an IoT-based smart T-shirt to provide personalized breathing training. A sample of 10 adults took part in the intervention. During the intervention, they received real-time guidance and instructions for adapting their breathing at various speeds. The participants were satisfied and relaxed after the intervention.

Wearables provide individualized solutions for people with anxiety disorders. Morris et al. [118] evaluated a stress management mobile app developed for Android Wear smartwatches to promote self-regulation and stress management skills among subjects with post-traumatic stress disorder. The results demonstrated that 86% of the trainees would very likely make daily use of the intervention. They reported that the elements that they liked the most were quick accessibility, customizability, the biofeedback received, and the reminders used.

Non-invasive neurofeedback-assisted breathing training was found to equip trainees with new skills essential for being mentally and emotionally balanced. In addition, it was found to be more effective when compared with conventional breathing training. Bhayee et al. [119] evaluated a neurofeedback-based training intervention that guided 13 healthy adults to manipulate their breath. The researchers compared the results with those of an active control group (n = 13) that solved online math problems. A wearable brain-sensing headband was utilized, accompanied by a mobile phone. Neurofeedback was delivered via auditory cues (i.e., wind and storm sounds), which increased in intensity when trainees were distracted and subsided towards calm when trainees’ attention was stabilized. The findings indicated that the neurofeedback-based breathing condition showed better improvement in attentional regulation and inhibition control compared with the control group.

Similarly, Balconi et al. [120], implemented a similar intervention in a sample of fifty subjects. The intervention group received mindful breathing training with a wearable brain-sensing device, whereas the control group received conventional breathing training. The experimental group outperformed in terms of attention and behavioral regulation.

Significant improvements in the ability to be aware and regulate mental functions such as attention were found after a neurofeedback-assisted breathing intervention conducted by Hunkin et al. [121]. The neurofeedback-based breathing intervention helped 68 trainees induce a state of increased awareness and perceived control over attentional operations as measured by the non-invasive EEG headbands.

Crivelli et al. [122] investigated the impact of neurofeedback breathing training on attention control, self-awareness, and psychological wellness. The experimental group received breathing training with the support of a wearable neurofeedback device, whereas the control group followed conventional breathing practices. After fourteen training sessions, participants in both conditions showed significant improvements in all indicators of attention regulation.

The selected studies indicated that neurofeedback-based breathing training can be an effective self-regulation technique for people with anxiety disorders. In a randomized control trial, Schuurmans et al. [123] used neurofeedback game-based mindful breathing training with a sample of 77 young people with posttraumatic stress disorder. After six weeks, the trainees improved their ability to self-regulate anxiety symptoms.

Table 7. A summary of the types of breathing interventions based on neurotechnologies and wearable technologies and the main results.

Neurotechnologies and Wearables in Breathing Interventions	Mental and Emotional Health Benefits	Reference
Breath-focused training via a mobile EEG device	Self-regulation, stress management, attentional regulation, inhibition control, positive thinking, self-awareness, attentional control	[119,121,122,123]
Diaphragmatic breathing via a wrist-worn wearable device and a mobile device	Stress regulation, mood regulation, decreased depression, increased perceived well-being	[114,115]
Diaphragmatic breathing via a mobile app and smartwatch	Stress management, self-regulation, mood regulation, reduced depressive symptoms	[118]
Clothing-attached, physiological monitoring device capable of sensing respiratory patterns and a smart-phone application	Stress regulation, emotional regulation	[71]
IoT-enabled smart T-shirt	Stress management	[117]
Slow breathing via biofeedback VR HRV	Heart coherence, resilience, self-regulation, attention regulation, mood regulation, self-awareness, relaxation, stress management, self-motivation, breathing control skills, greater focus, improved time reaction, less mind-wandering	[24,26,76,101,106,108,109,112]
Diaphragmatic breathing via biofeedback-based VR games	Increased awareness of physiological operations, attentional control, breath-control skills, motivation, response inhibition under stress, improved performance in decision-making contexts, increased sense of internal control, engagement, concentration, self-monitoring, self-regulation, stress regulation	[77,78,79,81,98,104,110,111]

and Figure 7 present a summary of the breathing interventions based on neurotechnologies and wearables through the lens of skills development.

5. Discussion

5.1. Major Findings and Final Considerations

The current review investigated the role of emerging technologies in breathing training for promoting mental and emotional health through the lens of skill development. Overall, the results of this paper indicated that emerging technologies can effectively assist in breathing interventions. It is noteworthy that digital designs in breathing interventions demonstrated better outcomes compared to conventional interventions [98,119]. It was also found that digitally assisted breathing interventions can promote physical, mental, and emotional help.

One of the main research questions concerned the types of technologies used in digitally assisted interventions. The results indicated that artificial intelligence, immersive technologies, and non-invasive neurotechnologies are promising assistive technologies. More specifically, the selected studies provided evidence about the following digital breathing designs:

• AI multilingual chatbots that provided instructional content and breathing guidance [91,92];
• Empathetic conversational agents that blended breathing training with relaxation and positive psychology techniques [94];
• AI-based personal companions and coaches that provided breathing guidance along with breath-related reflective conversations [69,93];
• Breathing robots that simulated breathing patterns [96,97];
• VR-based breathing intervention [99,100,101,102,103];
• Biofeedback VR-based breathing interventions [24,106,107,108,109];
• Biofeedback VR-based breathing interventions with the support of avatars [16]
• Breathing training with biofeedback VR games (or exergames) [77,78,79,81,98,104,110,111,112]
• Mixed reality breathing training systems that blended virtual reality and augmented reality with the support of physiological sensors [113];
• Gamified biofeedback breathing training [116];
• Smart app-based biofeedback breathing interventions assisted by sensors and IoT-based technologies [71,114,117,118];
• Neurofeedback-based breathing training [119,120,121,122,123].

The second part of our investigation focused on the impact of digitally assisted breathing training on mental and emotional health through the lens of skills development.

Table 8. The main breathing training benefits in mental and emotional health through the lens of skills development.

Training Benefits in Mental and Emotional Health Through the Lens of Skills Development	References
Physiological awareness	[69,77,78,107]
Physiological control	[69,107]
Self-awareness	[76,77,78,92,107,121]
Self-monitoring	[79]
Self-regulation	[24,69,91,94,98,111,112,115,118]
Attention regulation	[16,24,79,95,99,103,106,107,110,119,120]
Cognitive control	[91]
Inhibition control	[78,119]
Mental clarity	[103]
Adaptative behavior	[91,102]
Emotional awareness	[92]
Emotional flexibility	[93]
Emotional stability	[101]
Emotional regulation	[71,91,92,105]
Mood regulation	[24,94,101,115,119]
Positive thinking and feeling	[97,100]
Flow	[104]
Transcendence, state mindfulness	[100,113]
Self-efficacy	[24]
Self-relaxation, stress management skills	[71,77,81,94,95,96,99,101,102,103,105,108,113,114,115,116,117,118]
Decision-making	[78]
Self-motivation	[78,79,107,109,110]

demonstrates the training benefits of digitally assisted breathing training on mental and emotional health focusing on skill development.

Overall, most of the studies confirmed our initial hypothesis that digitally assisted breathing intervention can improve mental and emotional health. Indeed, the results indicated significant improvements in various metacognitive and metaemotional skills. More specifically, after training, participants were more able to be aware of and control physiological operations such as respiration [69,77,78,107]. They improved their awareness about themselves and their emotions [76,77,78,92]. Thus, they could more effectively observe and regulate negative thoughts and emotions [24,69,71,76,79,91,92,93,94,98,105,111,112,115,118]. Systematic breathing with the assistance of advanced technologies allowed practitioners to reduce cognitive load and mental fatigue, to think with clarity, and to control mental functions with increased flexibility [24,26,79,91,95,99,103,106,107,109,110,119,120]. The training helped subjects to inhibit impulses, stabilize mood, and adopt more flexible behaviors [76,78,91,93,94,102,115,119]. It is noteworthy that the integration of digital technologies in breathing training helped trainees to self-induce more positive states of mind (i.e., flow and awe) and, in general, think more positively [100,104,113]. The selected studies also demonstrated that digital technologies in breathing can be an excellent strategy for training stress-management skills [71,77,81,94,95,96,99,101,102,103,105,108,113,114,115,116,117,118]. Finally, it is essential to mention that such training interventions increase trainees’ self-efficacy and internal motivation to continue practicing [24,78,79,107,109,110].

The results of this study confirm the findings as well as the hypotheses of previous studies that digital technologies may have a supportive role in breathing training [27]. For instance, the results of this study are in line with previous studies that outlined the potential of digitally assisted breathing to raise trainees’ motivation and provide effective guidance and valuable multimodal feedback. In addition, this study confirmed previous studies that recognized the beneficial role of virtual reality, wearables, and non-invasive sensors in supporting such training for promoting mental health and emotional health [28,29].

AI chatbots which provided breathing instructions and guidance were found to improve users’ psychological balance [90]. By blending breathing exercises with relaxation techniques, reflective conversations, and empathetic responses, chatbots allowed novices to feel intrinsically motivated, relaxed, and more able to be aware and manage breathing operations [91,92,93,94]. The selected studies revealed that chatbots, conversational agents, and intelligent coaches helped trainees develop a wide range of skills that improved mental and emotional health (i.e., emotional regulation, self-regulation, and positive thinking). However, breathing robots were found to be less effective [96].

VR-based breathing training improved various aspects of mental and emotional health. Fundamental mental abilities, such as attention, were improved [109]. Immersion in VR combined with biofeedback and slow-paced breathing techniques was found to be promising, showing improvements in practitioners’ cardiac coherence, mood regulation, and resilience [24,106,107,108]. In general, it was observed that VR and biofeedback can be an excellent combination, except for one study [109]. In trying to interpret these findings, we assume that VR covered the limitations derived from the use of biofeedback (i.e., monotony, reduced motivation and rewarding, and negative reactions to biofeedback due to a lack of experience) [124]. VR provided the benefits of immersion, engagement, motivational instructions, interactive feedback, and gamification elements [24]. Virtual reality breathing enabled subjects to liberate attentional resources by inducing a state of presence and directing users’ attention toward relaxing cues [24]. In the same direction, biofeedback minimizes VR limitations (i.e., VR may provoke stress). More specifically, biofeedback allowed trainees to relax and breathe more deeply and rhythmically.

Deep-breathing techniques assisted by biofeedback and VR games enhanced trainees’ sense of control, self-monitoring capacity, and internal motivation. In addition, heightened levels of attention were observed [79,81]. Participants increased their breathing awareness. They were more able to use attention as a tool for regulating breathing, to induce a state of mental clarity and emotional stability. The use of avatars in VR-based breathing interventions encouraged embodiment, which in turn increased practitioners’ attention and motivation to continue efforts to achieve the desired breathing pattern [79,81].

The integration of gamification elements in digitally assisted breathing interventions was found to be an effective strategy because it raised trainees’ motivation for practice. The use of engaging badges, leaderboards, points, different levels of difficulty, and challenges, as well as music elements, were the most mentioned [125].

Biofeedback-based breathing training with the support of various sensors was found to be positive for stress management [114,115,117]. Gamified biofeedback was also an effective method for developing adaptive skills as well as the sense of having self-control [116].

Non-invasive neurofeedback-based breathing training effectively trained subjects to regulate their breathing patterns by fine-tuning brain activity with respiration and attention. By fine-tuning brainwave activity with the rhythm and depth of breathing, neurofeedback allowed subjects to develop voluntary control over respiration [122]. It was observed that neurofeedback enabled trainees to be aware of, observe, and control attentional operations more effectively, encouraging focus to remain active in breath-control training [119,121]. By providing real-time feedback on brain wave patterns, trainees developed the skills to voluntarily induce deeper states of relaxation allowing breathing to become slower and more stable [123]. Neurofeedback-based breathing interventions also had a significant positive influence on behavioral regulation [120].

5.2. Discrepancies, Adverse Effects, and Other Limitations of the Selected Studies

The investigation of the selected studies revealed the following limitations: For instance, several interventions lacked randomized controlled environments [90,91]. Other studies did not include follow-ups and long-term assessments [76]. The heterogeneity of the measurements was another difficulty in the interpretation of the results. Especially in the case of AI-powered applications, there were doubts about the reliance on the self-report measures utilized [91]. A limited number of studies recruited a small sample of participants [78]. The brief duration of several studies can be a risk factor for the reliability of our results [71]. In addition, several interventions, especially AI apps, combined additional interventions such as relaxation techniques. Thus, it was not easy to identify the possible influence of the concurrent interventions [91]. However, it was hypothesized that the blending of breathing training with relaxation and self-motivational techniques can help breathing training provide positive outcomes.

In addition, in interventions that mostly employed AI-powered applications, it was not possible to gather or share demographic data about age and gender. Thus, the process of generalizing the results was more difficult. In addition, anonymity limited the specific feedback analysis [90]. The reports of participants in several studies indicated that a source of displeasure from the use of virtual coaches was the fact that they lacked human-like features [90].

In most studies, the participants reported satisfaction with the use of VR without mentioning adverse effects [77,103]. In the study conducted by Rockstroh et al. (2019) [76], participants evaluated the VR training program positively and reported low values of simulation sickness. Waller et al. [100] mentioned that some participants characterized the VR breathing conditions as tiring. In the study conducted by Cook et al. [103], a limited number of subjects reported a mild increase in headache, dizziness, and nausea. Kang et al. [112] mentioned that some participants found the headset devices heavy, inconvenient, or too hot. In addition, the researchers mentioned that the connection line with the sensors was cumbersome. As regards the use of breathing interventions supported by wearable devices, Morris et al. [118] mentioned difficulties regarding the operation of devices by people with physical difficulties.

Regarding the neurofeedback-based breathing interventions, we can mention as a limitation the use of low-cost consumer-grade EEG devices. Moreover, several studies recruited a limited number of participants, and the duration of the interventions was short.

5.3. Emerging Technologies in Breathing Interventions: Challenges, Future Research, and Implications

AI-powered tools in breathing training inevitably raise ethical and legal issues as regards privacy and data security [90]. For that reason, clear guidelines and strict policies are considered a necessity. In addition, the collaboration among healthcare professionals, digital technology professionals, trainers, and policymakers must be a requirement in the process of designing, implementing, and evaluating digitally assisted breathing interventions [72].

AI algorithms utilize big data without always being able to differentiate between reliable and unreliable sources. In addition, AI tools may not have all the information needed to make an accurate evaluation of the trainees’ mental and emotional health needs. Thus, AI-assisted breathing interventions must be used as complementary interventions, especially for people with health problems. Remote screening of anonymous populations in the absence of in-person formal assessments may raise risks for trainees. The supervision of trainees and therapists should be considered a prerequisite to avoid inaccurate or biased assessments. However, it is encouraging to note that many AI-powered apps connect users both with virtual assistants and healthcare professionals [126].

Immersive technologies can provoke diverse effects such as cybersickness, nausea, fatigue, eye strain, and disorientation [127]. VR may elevate users’ anxiety. However, breathing training can help users overcome stress-related symptoms derived from the use of VR. Other significant topics concern the cost of equipment, the therapists’ training, the cultivation of positive attitudes about the role of technologies in breathing intervention, the design of customized breathing training sessions, and the formal evaluation of the immersive technologies [127]. Poorly designed VR breathing interventions might distract trainees from the objectives of the breathing exercises [76]. Thus, special focus must be given to the design process.

The use of biofeedback in breathing interventions may have some risks. The use of biofeedback may be challenging for novices. For instance, negative feedback may be intimidating and increase the stress of trainees. Trainees often report that the use of biofeedback in breathing training may be monotonous, less motivational, and less rewarding [76,124]. Non-invasive brain-sensing wearable devices may cause anxiety, irritability, mental fatigue, and mental fogginess [128]. Similarly to the other technologies, neurofeedback-based breathing, especially for clinical purposes, must be performed under the supervision of licensed practitioners.

Table 9. The main challenges derived from the use of advanced technologies in breathing training.

Emerging Technologies in Breathing Interventions	Risks, Challenges, and Ethical Considerations
Artificial intelligence tools and methods	Biased content, over-dependence, data insecurity, lack of human connection, lack of privacy, discriminatory access, unreliable resources, unvaried responses, medical and legal issues, and unsafe content for sensitive groups [73,90,126]
Immersive technologies	Adverse effects (i.e., cybersickness), heavy devices, safety concerns for people with health issues, limited experts’ training, cost challenges, design limitations, lack of comprehensive manuals, limited technical knowledge, unethical utilization [76,127]
Biofeedback and neurofeedback technologies	Adverse effects (i.e., headaches, anxiety, mental fatigue, emotional disturbances), negative feedback may be stressful, safety issues, monotonous, less rewarding, and motivational, data inaccuracy [76,124,128]

summarizes the main challenges derived from the use of advanced technologies in breathing training.

The use of emerging technologies in breathing training is in its early stages. More evidence is needed focusing on the potential health benefits as well as the dangers derived from the employment of advanced technologies in such training. It would be important to develop deeper knowledge as regards the effectiveness of digitally assisted breathing interventions in different social groups (i.e., employees and students), and most importantly, in sensitive populations (i.e., people with different disorders).

Additional research with formal evaluations should be conducted to provide more reliable evidence about the effectiveness as well as the safety of the digital tools employed in breathing interventions [27]. The development of intelligent and virtual breathing training coaches should emphasize the development of human-like characteristics to make the intervention procedure more appealing and motivational [69]. As regards biofeedback, more research is needed to examine different biofeedback modalities combined with different breathing exercises or breathing games [104].

An interesting topic for future research concerns the blending of different techniques and technologies in breathing interventions. Future research should emphasize the evaluation of different combinations of technologies. For instance, the use of VR-based biofeedback interventions was found to be very promising [108].

Our investigation helped us realize that the effectiveness of digitally assisted breathing interventions depends significantly on the design of the interventions. Thus, this research outlines the emergence of appropriate digital designs with the support of professionals from different fields [76].

Digitally assisted breathing interventions can help both populations with clinical and non-clinical conditions. Such interventions can be an essential tool for promoting mental and emotional health among people living in rural areas where the conditions of life are very stressful [90]. In addition, according to the results of the current review, digital breathing interventions can be used in the workplace to promote employees’ stress-management skills [71]. Mobile-based conversational agents can deliver engaging and effective breathing interventions among adolescents to reduce stress and improve well-being conditions [93].

As a final comment, we must outline that the achievement of digitally assisted breathing interventions depends on the human’s willpower to recognize and manage physiological operations such as breathing to improve mental and emotional health. Humans are required to raise awareness of the healing powers of breathing and the various ways that breath-control practices can make their lives better. In addition, it is essential to understand the potential role of digital technologies to assist them in the journey of discovering the healing powers of breathing. Finally, they should be aware of the benefits, the limitations, and the risks derived from the use of technologies. People who receive such interventions should be gradually familiarized with the use of advanced technologies, always with the support of professionals.

6. Conclusions

After the COVID-19 pandemic, a global discussion has begun about the potential role of emerging technologies in breath-control interventions. The current systematic review study investigated the effectiveness of digitally assisted breathing interventions on mental and affective health through the lens of skill development. Artificial intelligence tools, immersive technologies, and non-invasive neurotechnologies were mostly identified, providing promising results for a wide range of mental and emotional health conditions. The results indicated that emerging technologies can significantly assist breath-control interventions, allowing the development of a wide range of metacognitive and metaemotional skills that may promote mental and emotional health among people with clinical and non-clinical conditions. Self-awareness, self-regulation, attentional control, inhibition control, emotional awareness, emotional regulation, positive thinking, stress management, self-efficacy, and self-motivation were the skills that the trainees mostly developed.

The results did not indicate significant adverse effects. However, special emphasis should be given not only to the benefits but also to the risks and ethical considerations that arise from the use of advanced technologies.

The results of the current review aspire to contribute to the discussion about the crucial role of advanced technologies in breathing training programs to promote mental and emotional health. In addition, this study can provide feedback about the potential role of emerging technologies in breathing training, including the benefits and the risks.

Future research should shed more light on the significant impact of digitally assisted breathing interventions on the development of skills that can help trainees voluntarily enhance their mental and emotional health. Additional digitally assisted breathing interventions can be implemented in populations with different needs, such as students, employees, and people with mental and emotional disorders. An interesting topic for future research concerns the blending of different techniques and technologies in breathing interventions. Future research should emphasize the proper design and evaluation of digitally assisted breathing interventions based on the combination of different technologies and breathing techniques, considering the advantages and risks of the technologies used and, most importantly, the special needs and characteristics of the trainees.