Metaverse-Related and Immersive Digital Technologies in Neurorehabilitation: A Scoping Review Authors: MohammadHossein Sahami Gilan, Azin Hasani, Negar Farajzadeh Dehkordi Published: 2026 ABSTRACT AbstractObjectiveTo delineate current research on the use of metaverse technologies in neurorehabilitation, pinpointing essential intervention attributes, target demographics, documented results, and evidence gaps in this nascent domain.Data SourcesA comprehensive search was performed across 7 electronic databases: PubMed, Scopus, Web of Science, Cochrane Library, Embase, CINAHL, and Google Scholar, including all papers published through April 19, 2025. However, only a subset of studies that employed immersive virtual reality (VR)/augmented reality or early metaverse-related technologies were included.Study SelectionOf 5681 initial records, studies were included if they investigated immersive technologies for the neurorehabilitation of human participants with neurological conditions and were published in English. Nineteen studies fulfilled the final eligibility criteria after screening.Data ExtractionTwo reviewers independently recorded data using a standardized form. The extracted items included study design, population characteristics, technology type, intervention specifics (duration, frequency), outcome measures, and key findings.Data SynthesisA narrative synthesis was conducted on the 19 included studies. VR was the most frequently used technology, accounting for 63% of studies. However, a subset of studies also incorporated early metaverse-related elements alongside VR for various neurological conditions. More than half of the studies (57.9%) reported significant improvements in motor, cognitive, or emotional domains. High methodological heterogeneity was observed, with only 26.3% of studies being randomized controlled trials.ConclusionsMetaverse-related and immersive digital interventions show promise for improving functional outcomes in neurorehabilitation; however, the evidence base remains preliminary, heterogeneous, and limited in terms of long-term follow-up. The therapeutic effectiveness and scalability of these therapies must be established via rigorous, large-scale trials with consistent procedures and long-term follow-up.KEYWORDS: Central nervous system disorders, Computer simulation, Nervous system diseases, Neurological rehabilitation, Rehabilitation, Virtual reality Neurorehabilitation is essential for restoring functional independence and improving quality of life in individuals with neurological impairments, including stroke, traumatic brain injury (TBI), and neurodegenerative diseases.1,2 A 2024 report by the World Health Organization indicated that more than 1 billion people are affected by neurological disorders globally, with stroke alone impacting 15 million individuals annually.3 Among these individuals, 5 million experience long-term disabilities requiring rehabilitation. Nevertheless, traditional neurorehabilitation methods face significant challenges, including limited access to specialized care, high costs, and difficulty maintaining patient engagement over extended periods.4 Individuals with severe neurological disabilities require intensive and long-term health care, rehabilitation, and social support. Therefore, updated population data and research on technological innovations are essential to inform clinical planning and optimize resource allocation, especially when proposing new technological advancements.5 Recent advancements in digital technology have introduced the metaverse, an immersive virtual environment, as a promising tool for health care applications.6 A study7 published in 2025 reported that virtual reality (VR)-based interventions, often considered foundational or precursor technologies for future metaverse ecosystems, have been integrated into 65% of mental health therapy programs in high-income countries, resulting in a 30% improvement in patient engagement compared with traditional methods. In this review, VR and augmented reality (AR) are not treated as synonymous with the metaverse; rather, they are considered metaverse-related or transitional technologies when they incorporate system-level integration, multiuser interaction, or other metaverse-like features. However, the metaverse’s application in neurorehabilitation remains largely unexplored, with only 12% of neurorehabilitation studies in 2024 examining its potential, according to a systematic review by Taylor et al.8 Metaverse-related technologies build on established immersive tools such as VR and AR.9 VR immerses users in a simulated environment for practicing motor or cognitive tasks,10 whereas AR overlays digital information onto the real world to support real-time guidance and feedback during functional activities.11 Extended reality is an umbrella term encompassing VR, AR, and mixed reality, which blend real and virtual elements with varying degrees of interactivity.12 While VR and AR have been explored independently for years, the metaverse represents a fundamental shift from isolated tools to an integrated ecosystem.13,14 Unlike standalone VR/AR tools, which serve as immersive interfaces, the metaverse is a collective, persistent, and interactive virtual space. Its clinical added value lies in creating a “rehabilitation community” where data and environments persist even when the user is offline, and where social interactivity enables real-time collaboration between patients and clinicians, offering a level of engagement that traditional, single-user VR cannot provide.15 The metaverse, as defined in this review, includes 3 key attributes that distinguish it from VR/AR applications: (1) convergence: the integration of technologies like VR, AR, artificial intelligence (AI), and the Internet of Things (IoT; a network of interconnected physical devices that collect and exchange data) into a unified system; (2) persistence: the continuous existence and development of virtual environments, even without user interaction; and (3) social interactivity: enables real-time collaborative experiences among multiple users, including patients, clinicians, and caregivers.16,17 This integrated framework provides innovative avenues for neurorehabilitation by replicating intricate, real-world scenarios in a controlled, data-rich, and socially interconnected environment.1,14 Substantial literature, including systematic reviews, has established the effectiveness of specific immersive technologies, such as VR, for poststroke upper-limb rehabilitation and for improving game-based motor function.18 However, these reviews typically focus on a single technology for a specific condition.19 For instance, foundational scoping reviews, such as those by Darekar et al,20 provided an early snapshot of VR’s potential but were naturally limited to older technologies and a narrow focus on poststroke balance and mobility. A recent scoping review by Ceradini et al21 synthesized evidence on fully immersive VR (IMVR) for upper-extremity rehabilitation, drawing on data from 13 studies across 3 databases. The review, while informative, reveals a gap in the literature due to its intentionally narrow scope, indicating the need for a broader synthesis. The metaverse, as a convergent paradigm integrating VR, AR, AI, and social platforms, offers a holistic approach that goes beyond task-specific immersive tools. The novelty of focusing on the metaverse, rather than on broader immersive technologies, is justified by its unique ability to facilitate a “rehabilitation continuum.” This paradigm enables a persistent digital space in which recovery is integrated into a social and data-rich environment rather than being limited to isolated, nonpersistent training sessions. No scoping review has, to our knowledge, mapped the landscape of integrated metaverse systems across various neurological disorders. This review was guided by the research question: “What are the characteristics of current studies on metaverse-related and immersive digital technologies in neurorehabilitation, which clinical populations and intervention types are addressed, and what outcomes and evidence gaps are reported?” This scoping review aimed to map the emerging landscape of metaverse-related and immersive digital technologies in neurorehabilitation by synthesizing current interventions, target neurological populations, reported outcomes, and key evidence gaps. Methods Data sources This scoping review adhered to the methodology outlined by Arksey and O'Malley,22 which aimed to map the literature, identify key concepts, and highlight research gaps. The review adhered to the Preferred Reporting Items for Systematic Reviews and Meta-Analyses extension for Scoping Reviews guidelines, ensuring transparency and comprehensive reporting throughout the review process.22 The process was divided into 5 stages: (1) the research question was defined; (2) relevant studies were identified; (3) studies were selected based on eligibility criteria; (4) data were extracted from the selected studies; and (5) the findings were collated, summarized, and reported.23 A systematic search was conducted across 7 electronic databases: PubMed, Scopus, Web of Science, Cochrane Library, Embase, CINAHL, and Google Scholar. This search was performed without time restrictions and included all studies published through April 19, 2025. To identify relevant studies, a combination of Medical Subject Headings and free-text terms was used as follows: ("Neurological Rehabilitation" OR "Nervous System Diseases" OR "Neurological Disorders" OR "Neuronal Disorders") AND ("Metaverse" OR "Metaverse Rehabilitation"). The comprehensive search strategy, encompassing tailored queries for each database, is outlined in supplemental table S2 (available online only at http://www.archives-pmr.org/). Study selection To ensure a rigorous selection process, eligibility criteria were framed using the Population, Intervention, Comparison, Outcome, and Study Design framework as follows: Population: individuals of any age with neurological disorders (eg, stroke, multiple sclerosis [MS], Parkinson’s disease) who require rehabilitation; Intervention: immersive digital technologies, including VR, AR, or platforms with metaverse attributes (multiuser environments, persistence, and social interaction); Comparison: conventional therapy, alternative digital interventions, or no control group; Outcomes: clinical efficacy, motor/cognitive improvement, feasibility, or user engagement; and Study Design : randomized controlled trials (RCTs), quasi-experimental studies, and observational studies. Studies were excluded based on the following criteria: non-English publications, participants with severe cognitive impairment that precluded informed participation in the intervention or assessment procedures, lack of original data (including protocols or preprints), and fewer than 2 full-text pages. Additional exclusions were applied to studies that employed nonrelevant or nonimmersive technologies, such as non-IMVR/AR, or engaged fewer than 3 senses. Studies on specific conditions outside of neurorehabilitation, such as spinal stenosis associated with chronic low back pain, were also excluded. This exclusion criterion was established to ensure conformity with our understanding of the metaverse as a profoundly immersive paradigm. Since low-immersion or noninteractive digital technologies do not adequately capture the rich, multimodal worlds at the heart of the metaverse concept, interventions that rely solely on low-immersion or noninteractive digital technologies without meaningful sensory or experiential engagement were judged to be outside the scope of this review. The search process involved systematic screening of titles and abstracts, followed by a full-text review of potentially relevant articles. The selection was guided by predefined inclusion and exclusion criteria, and the final number of included studies was documented using a Preferred Reporting Items for Systematic Reviews and Meta-Analyses flow diagram, as detailed in the Results section (fig 1).23 The screening and management of retrieved records were conducted using Rayyan software.a The Preferred Reporting Items for Systematic Reviews and Meta-Analyses flow diagram of the study. Data extraction A standardized data extraction form was developed and rigorously tested on a subset of studies to ensure consistency. The extracted data included vital details such as authors, publication year, and country of origin; study design and sample size; population characteristics (including neurological diagnosis, age, and sex); the type of metaverse technology used (eg, VR, AR, multisensory technology, or metaverse platforms); intervention specifics (including duration, frequency, and setting); outcome measures (eg, motor function, cognitive performance, or psychosocial well-being); and key findings from each study, which delineated the reported advantages and limitations of the interventions. To maintain rigor and reduce bias, each phase of the review process, including study selection and data extraction, was conducted independently by 2 reviewers (N.F.D. and M.H.S.G.). Discrepancies at both stages were addressed through dialogue to achieve consensus. A third reviewer was consulted to resolve any outstanding disputes. Data synthesis The extracted data were synthesized in a narrative format. Studies were organized thematically according to the type of metaverse technology employed, the neurological conditions targeted, and the primary domains of reported outcomes, including motor, cognitive, and psychosocial aspects. The characteristics and findings of the included studies were summarized in tables and figures, thereby providing a comprehensive overview of the evidence base. Gaps in the literature were identified, and potential directions for future research were underscored during the synthesis process.24 Due to the high level of methodological heterogeneity regarding intervention protocols, technological platforms, and clinical outcome measures across the included studies, a quantitative meta-analysis was not feasible. Instead, a narrative synthesis was performed to map the emerging landscape of metaverse applications in neurorehabilitation. Results General characteristics of the included studies This scoping review initially identified 5681 records through comprehensive database searches. After 1164 duplicates were removed, 4517 articles remained for title and abstract screening. A total of 686 full-text articles were subsequently assessed for eligibility, 19 of which met the inclusion criteria and were selected for final analysis (fig 1); their key characteristics are summarized in supplemental table S1 (available online only at http://www.archives-pmr.org/). These studies were published between 2022 and 2025, with most appearing between 2023 and 2024. Geographically, South Korea contributed the most studies (n=6; 31.6%), followed by Italy (n=4; 21%). Additional studies originated from countries such as the United States, Australia, China, Iran, India, and the United Kingdom, with 1 multinational collaboration. The 19 studies included exhibited significant design heterogeneity. The largest group comprised 10 review articles of different types,25, 26, 27, 28, 29, 30, 31, 32, 33, 34 followed by 5 interventional studies, including both RCTs35, 36, 37 and experimental trials,38,39 and the remaining 4 encompassed various designs, including cohort studies,40 perspectives,15,41 and commentaries.42 The sample sizes varied considerably, from as few as 2 participants in a feasibility case report35 to 627 individuals in a large-scale study on metaverse-related cognitive rehabilitation.30 On average, nonreview studies reported sample sizes of approximately 150 participants. Target populations and neurological conditions The populations studied spanned a broad spectrum of neurological conditions. The most frequently investigated group comprised individuals with cognitive impairments, including Alzheimer’s disease and mild cognitive impairment (MCI), which featured in 6 studies.15,29,30,34,38,39 Other populations included patients with cerebral palsy (2 studies),36,41 stroke survivors (3 studies),15,31,41 those with anxiety and depressive disorders (3 studies),26,37,42 individuals with spinal cord injuries (1 study),40 and patients experiencing chronic pain (eg, back or neck pain).37 Notably, 2 studies focused on caregivers’ training and on enhancing empathy for individuals with dementia.34,43 Age ranges varied widely across studies, encompassing both pediatric (eg, children with cerebral palsy; mean age=11 years) and geriatric populations (eg, older adults with MCI or Alzheimer’s disease). For example, Oh et al38 examined metaverse-based cognitive training in seniors aged >70 years, whereas Moon et al36 assessed physical therapy interventions in children with cerebral palsy. Intervention characteristics and technologies used The interventions were grounded in a variety of metaverse-related technologies and digital tools. The most commonly used technology was VR, which was implemented in 12 studies (63%). This predominance of standalone VR over fully integrated metaverse platforms suggests that the field is still in its infancy, with many interventions focusing on immersive simulations rather than on persistent, multiuser digital ecosystems. AR was used in 4 studies, whereas 3 presented more integrative platforms combining AI, digital twins (DTs; virtual models of physical objects or systems), and extended reality. Several studies, such as Zhou et al29 and Lan et al,30 employed hybrid models that integrate VR, AR, AI, IoT, and DTs to offer highly personalized cognitive rehabilitation programs. The intervention dosage and delivery settings varied considerably. For example, Moon36 used 30-minute sessions 3 times/wk over 4 weeks, whereas Mascanzoni et al40 implemented a long-term program with a mean duration of 166 days. Most interventions followed a standard 30-60-minute session, 2-3 times/wk, over 4-12 weeks. The settings included hospitals, rehabilitation centers, participants’ homes, and fully virtual environments such as Oculus Questb and Second Life.c Reported outcomes and observed patterns Outcomes measured across the studies included motor function, cognitive performance, psychosocial well-being, pain reduction, and enhanced social participation. The 19 studies revealed distinct patterns in outcomes across neurological conditions. VR-based interventions were frequently associated with improved cognitive and motor functions, particularly in Alzheimer’s disease, MCI, and cerebral palsy.30,36 Studies incorporating AI or DT technologies29 often emphasized enhanced personalization of interventions. Additionally, psychosocial outcomes such as reduced anxiety and improved caregiver empathy were commonly reported in studies using avatar-based or social VR platforms.37 Specific investigations elucidated these trends by offering more profound insights into the observed results. Among the 19 studies, 11 reported significant improvements in 1 or more domains (representing 57.9% of the included studies). However, these findings should be interpreted with caution; as per scoping review protocols, no formal quality assessment or risk-of-bias analysis was conducted. Therefore, these results provide a descriptive overview of potential benefits rather than a definitive confirmation of clinical efficacy. For example, Yu et al37 reported notable reductions in anxiety, depression, and pain with a VR-based yoga intervention, whereas Ziaee et al26 reported significant decreases in symptoms of anxiety and depression with a VR-integrated cognitive behavioral therapy model. Moon36 reported that metaverse-oriented physical therapy improved motor and cardiopulmonary function in children with cerebral palsy more effectively than conventional physiotherapy. Kokorelias et al34 demonstrated that IMVR-based training improved caregivers’ empathy, communication, and perceived quality of life. Conversely, several studies15,35 focused primarily on feasibility or technological development and did not include quantitative outcome measures. Without empirical data, Nawaz et al42 offer a conceptual critique of metaverse psychiatry. The 19 studies exhibited considerable heterogeneity in their aims, technological approaches, and outcome measures. This variability was evident in the intervention protocols, which ranged from intensive short-term sessions (eg, 30-minute sessions over 4 weeks) to prolonged programs exceeding 160 days. Furthermore, the measured outcomes were highly diverse, ranging from objective motor metrics, such as gait speed, to subjective psychosocial scales, reflecting the broad and exploratory nature of metaverse applications in this field. Notably, participant adherence to the intervention protocols was poorly and inconsistently reported across the included studies, highlighting a significant gap in the current literature regarding the feasibility of these interventions. For detailed information on the intervention protocols, settings, and additional outcomes, please refer to supplemental table S1. Discussion The goal of this scoping review was to document the state of knowledge on the use of metaverse technologies in neurorehabilitation. This review answered its guiding research question on the features of existing therapies, the groups being targeted, and the key outcomes and evidence gaps described in the literature by combining the results of 19 studies. According to the findings, there is a growing emphasis on using interactive and immersive digital environments to address the complex problems related to neurorehabilitation. The reviewed literature revealed a wide range of neurological diseases that are addressed by metaverse therapies. These conditions include neurodegenerative diseases, including Alzheimer’s disease, MCI,14 pediatric cerebral palsy,44 poststroke sequelae, spinal cord injury, chronic pain, and psychiatric disorders such as anxiety and depression.45 The diversity of target demographics demonstrates the adaptability of metaverse applications and their capacity to facilitate tailored rehabilitation across multiple therapeutic environments. The predominance of VR-based approaches likely reflects their greater technological maturity, longer research history, and wider commercial availability relative to fully integrated metaverse platforms. However, it is crucial to reiterate the conceptual distinction between VR/AR and the metaverse: VR and AR serve as foundational immersive interfaces, whereas the metaverse is an interconnected, persistent network of these realities. VR’s ability to create immersive, controlled settings for repetitive, feedback-driven training aligns conceptually with several established neurorehabilitation principles, including task-specificity, repetition, intensity, and the salience of practice, which are recognized as key drivers of experience-dependent neuroplasticity.46 Its effect on motor performance was established even before the ‘metaverse’ concept emerged.47 Research conducted before the emergence of the metaverse concept, including the study by Calabrò et al,16 has shown that VR can enhance motor performance. Cognitive and motor function improvements through VR applications for conditions such as Alzheimer’s disease, MCI, and cerebral palsy align with a broader spectrum of external research findings. Maggio et al48 examined the computer-assisted rehabilitation environment system, an advanced VR platform, and demonstrated its potential to improve motor and cognitive skills in patients with neurological conditions. However, the review48 also emphasized the need for additional research, citing small sample sizes in current studies. Although only 4 studies directly involved AR technologies, the field is rapidly progressing, and these studies highlighted the potential of AR to enhance rehabilitation, particularly when combined with VR or integrated into metaverse approaches. The unique ability of AR to superimpose digital information onto the user’s real-world perspective presents notable benefits. These include enabling task practice in ecologically valid settings and potentially mitigating concerns such as cybersickness, which may affect certain vulnerable patient groups in fully IMVR environments. A recent scoping review49 on AR in physical therapy indicated that interventions often focus on gait and balance among neurological and older adult populations. While AR has occasionally demonstrated greater effectiveness, the overall efficacy outcomes have yet to be conclusively determined. A separate systematic review50 highlighted the use of AR in addressing balance issues, stroke rehabilitation, and Parkinson’s disease, emphasizing its promising potential while acknowledging the variability among studies and the necessity for more extensive trials. The external observations correspond with the encouraging, although constrained, results from the AR studies featured in our review. The choice between VR and AR in neurorehabilitation depends on treatment goals and patient characteristics. VR can provide realistic virtual worlds for instruction, whereas AR may improve patient contact with their real surroundings. According to a narrative review by Khokale et al,51 VR and AR may improve neuroplasticity poststroke, although patient customization is generally insufficient. A systematic review and meta-analysis of IMVR for upper-limb stroke rehabilitation revealed statistically significant improvements with IMVR compared with conventional rehabilitation across several outcome measures (eg, Fugl-Meyer Assessment for Upper Extremity, Box and Block Test). Nevertheless, these improvements are not always clinically significant, highlighting the need for optimal treatment protocols.52 The findings of this scoping review indicate preliminary, though cautiously optimistic, evidence supporting the application of both VR and AR in neurorehabilitation contexts. Given that neither technology has demonstrated universal superiority across all neurological conditions and patient profiles, a technology-specific and patient-centered approach, in which selection between VR, AR, or hybrid modalities is guided by therapeutic goals, clinical presentation, and individual patient characteristics, is warranted and should inform the design of future trials. The external literature suggests that AR is a rapidly advancing field with unique benefits, and the future may involve tailoring technologies to patient needs and therapeutic goals or using hybrid models that combine the strengths of different immersive modalities.53 The emphasis on cognitive impairments, such as Alzheimer’s disease and MCI, as well as stroke, is warranted due to their significant prevalence and associated disability burden.54 Immersive environments are effective for cognitive and motor training, as demonstrated by studies by Lan30 and Oh,38 which confirmed that VR enhances motivation and cognitive performance. This is consistent with the research priorities outlined by the Hertie Institute for Clinical Brain Research in Tübingen, which focuses on neurodegenerative disorders and stroke. Additionally, it reflects the efforts of researchers such as Dr Hannah Gullo, who explore cognitive impairment resulting from TBI and stroke through technology-enhanced therapy.54 The evaluated studies also addressed cerebral palsy, anxiety, depression, and spinal cord damage, although metaverse therapies might assist a far wider range of neurological diseases.48 Despite VR research being more established for Parkinson’s disease, MS, and TBI, there appears to be a comparative scarcity of metaverse research for these conditions.55 Despite the growing body of VR research on conditions such as Parkinson’s disease and MS, including studies demonstrating improvements in upper-limb function through VR-based serious game interventions,56 the application of integrated metaverse frameworks to TBI rehabilitation remains notably underexplored. This represents a significant gap, given the complexity and heterogeneity of cognitive and motor sequelae associated with TBI, which may benefit, in particular, from the personalized, persistent, and socially interactive features that distinguish metaverse ecosystems from standalone VR applications. The greater engagement, individualized training, and remote access associated with metaverse technology may benefit various patient populations.57 The focus of public health research on stroke and Alzheimer’s disease is reasonable. However, the versatility of metaverse technology suggests broader adoption. The lack of metaverse-specific studies (which often imply more integrated, persistent, or socially connected virtual experiences beyond standalone VR) in other prevalent neurological conditions creates a considerable research gap and an opportunity to explore how tailored metaverse interventions may meet the unique needs of these populations. The finding that more than half of the studies analyzed (11 out of 19) demonstrated significant improvements in motor, cognitive, or emotional domains is a positive indication for the discipline. This aligns with broader patterns observed in the field of digital therapeutics. In our review, the studies by Yu37 and Ziaee26 emphasized psychosocial benefits, whereas Moon36 reported improvements in motor function in children with cerebral palsy. These positive findings align with sentiments expressed in the literature. A study by Villa et al47 highlights the potential of integrated approaches, specifically combining motor imagery with robotics, to enhance gait and balance in patients with stroke while promoting brain plasticity. Recent research by Choi et al58 explored the efficacy of non-interactive immersive virtual reality training in improving standing function and gait in patients with stroke in their home settings. Evidence from the external literature further supports these observations; IMVR exergaming has been associated with improvements in motor function and cortical reorganization in individuals poststroke.59,60 A recurring theme, apparent in the studies reviewed, eg, Lan,30 indicates that VR increases motivation. In the external literature, these immersive technologies have the capacity to improve patient engagement and motivation. Ozturk et al61 emphasize that AR and VR offer significant advantages, including high repetition, task-specificity, objective feedback, and enhanced motivation. Research conducted by Nath et al62 shows that electromyogram-based visual biofeedback, which can be effortlessly incorporated into metaverse systems, enhances user control and motivation during robotic hand exoskeleton exercises. The personalization of therapy is frequently enabled by AI, as illustrated by Jovari et al,63 and is a vital element in enhancing therapeutic outcomes. Given the field’s early stage, this review found limited high-level evidence to confirm clinical efficacy. Of the 19 studies, only 5 (26.3%) were RCTs, the criterion standard for therapeutic research. This low proportion of RCTs relative to established digital health fields indicates that metaverse neurorehabilitation research remains largely exploratory. The prevalence of non-RCT designs, including pilot studies, feasibility trials, and conceptual papers, significantly contributes to the observed methodological heterogeneity. Variability in study designs, outcome measures, and intervention protocols significantly limits cross-study comparability and precludes a quantitative synthesis of the evidence. The profound methodological heterogeneity observed limits the generalizability of current findings and precludes definitive clinical guidelines. Consequently, a major research gap exists regarding the optimal dosage, hardware specifications, and long-term efficacy of these interventions. To move beyond exploratory proof-of-concept studies, future research must prioritize methodologically rigorous, large-scale, multicenter RCTs. Furthermore, future directions should focus on establishing standardized protocols to directly compare standalone VR with fully integrated, socially interactive metaverse platforms and delineate their specific added value.64 Advanced computing technologies will increase the neurorehabilitation capacity of metaverse systems. Previous studies, including those by Jovari,63 Lan,30 and Bazargani et al,39 have highlighted the usefulness of AI in customizing rehabilitation programs and enhancing diagnostic accuracy. Bazargani39 used VR and machine learning classifiers to detect Alzheimer’s disease early. Advances in health care AI complement this trend toward AI integration. Neuromorphic AI is a computational paradigm inspired by the architecture and processing principles of the biological brain, in which artificial neural networks are run on specialized hardware that mimics the structure and function of neurons and synapses. Unlike conventional deep learning systems, neuromorphic processors operate through event-driven, spike-based computation, enabling real-time adaptive processing with substantially lower energy consumption. In the context of neurorehabilitation, this has considerable implications: neuromorphic systems could enable continuous, low-latency adaptation of therapeutic environments in response to a patient’s physiological signals, such as electroencephalography or electromyography, thereby facilitating closed-loop, brain-machine interface-driven rehabilitation on metaverse platforms. The integration of neuromorphic AI into metaverse-based neurorehabilitation may therefore represent a significant step toward truly personalized and responsive digital therapeutics.65 Researchers at SPECS Laboratories are conducting research to develop trustworthy, secure AI-enhanced stroke neurorehabilitation solutions. Our scoping review includes hybrid models that use IoT, such as Jovari’s63 conceptual framework. In real time, wearable sensors and smart home technologies can monitor patients’ progress, exercise compliance, and physiological reactions. Metaverse systems may use this data to provide adaptive, tailored therapy, which is essential for telerehabilitation. A comprehensive review by Lee et al66 highlights how wearable motion sensors and mobile devices, key components of IoT, are opening new avenues for balance and gait training in telerehabilitation, with studies demonstrating persistent benefits over many weeks. A study by Athilingam et al67 provides preliminary evidence that IoT-enabled mHealth technology may improve exercise adherence among patients with heart failure. VR and AR alone may not alter neurorehabilitation, but their synergistic convergence with AI, DTs, and IoT may. AI can provide customization and adaptability, DTs can provide predictive modeling, and IoT can bridge the essential gap between clinical or virtual training settings and patients’ everyday lives. Neurorehabilitation ecosystems should become more adaptable, predictive, accurate, and successful with this combination. This review maps the current landscape of metaverse-related technologies in neurorehabilitation, demonstrating that while IMVR remains the predominant modality, the field is progressively transitioning toward more integrated, socially connected, and AI-driven platforms. This illustrates the methodological variability in existing research and the early development of empirical evidence, especially in the few RCTs. The design and priority of future studies on this subject depend on a critical review of current research. This review outlines the existing background, laying the groundwork for the field. It describes the current state of knowledge, evaluates methodological flaws in the literature, and identifies critical areas for future rigorous, high-quality research. Our review advances metaverse-related neurorehabilitation by directly communicating research goals to funding agencies, clinical innovators, and the scientific community. Despite the clinical promise, several critical real-world implementation factors remain underreported. The feasibility and scalability of metaverse-based interventions are significantly hindered by the high upfront costs of hardware (eg, advanced head-mounted displays and sensor arrays) and the requirement for robust technological infrastructure, which limit accessibility, particularly in low-resource settings.68 Furthermore, potential physiological and technological risks must be acknowledged. “Cybersickness,” characterized by nausea, dizziness, and visual fatigue, poses a major barrier to adherence, especially among older adults or patients with preexisting vestibular and sensory deficits resulting from stroke or TBI. Additionally, the technical literacy required to navigate these platforms often necessitates ongoing caregiver assistance, challenging the goal of independent home-based rehabilitation.69 Although these outcomes are promising, a significant obstacle to the therapeutic application of metaverse technology is the evident neglect of real-world implementation issues in the existing literature. Our review identified a substantial lack of evidence in this field, particularly the omission of reporting on participant adherence. Most studies failed to provide quantifiable data on compliance rates, session completion rates, or the causes of participant attrition. This is problematic for multiple reasons. First, in the absence of adherence data, it is infeasible to assess the actual feasibility and acceptability of these innovative interventions from the patient’s perspective. Second, it complicates the interpretation of efficacy findings: favorable results may be ascribed to a highly motivated and selected subgroup of participants rather than to the intervention itself.70,71 Third, while metaverse technologies can augment participation, they may also pose distinct adherence challenges, such as cybersickness, technical complexities for older adults, or the need for ongoing caregiver assistance. To advance the field beyond proof of concept, it is essential that forthcoming research meticulously monitor, document, and evaluate adherence measures. Using mixed-methods approaches to understand the obstacles and enablers of prolonged use will be essential for developing interventions that are both successful and durable in practical clinical and domestic settings.72,73 Study limitations This scoping review has several limitations. Despite a thorough and rigorous selection process, only 19 papers met the inclusion criteria, indicating the novelty of the field and the limited empirical research with well-delineated outcomes. Significant variability in study design, technology type, population, and outcome measures has limited the comparability of investigations. Furthermore, most interventions do not adequately report session frequency, duration, immersion level, or user interface, complicating the identification of the essential factors affecting effectiveness. As a scoping review, no meta-analysis or bias assessment was conducted; although independent screening twice reduced errors, subjective judgment may have influenced the inclusion of marginal papers. A critical limitation of this review is that none of the 19 included studies systematically collected or reported data on adverse reactions to the interventions. Potential adverse effects, such as cybersickness (including nausea, dizziness, and visual fatigue), musculoskeletal discomfort, psychological distress, or technology-related harms, were not formally assessed or documented in any of the included studies. The absence of safety monitoring and adverse event reporting represents a major gap in the current evidence base, as it precludes any evaluation of the safety profile of metaverse-related and immersive digital interventions in neurological populations. This is of particular clinical concern given that several target groups in this review, including older adults with Alzheimer’s disease, individuals poststroke, and children with cerebral palsy, may be especially vulnerable to such adverse effects. Conclusions This scoping review provides a comprehensive mapping of the emerging landscape of metaverse-related technologies in neurorehabilitation, highlighting a transformative shift from standalone simulations to persistent, data-driven digital ecosystems. By synthesizing evidence from 19 studies, our work shows that while VR remains the predominant interface, the integration of AI, IoT, and social connectivity is paving the way for a more holistic “rehabilitation continuum.” This study contributes to existing knowledge by conceptually distinguishing the metaverse paradigm from traditional immersive tools and by identifying critical implementation barriers, such as hardware costs, technical literacy, and cybersickness. While the reported clinical improvements across various neurological conditions are promising, the high methodological heterogeneity and limited number of RCTs necessitate a cautious interpretation of efficacy. To move beyond the current exploratory phase, future research must prioritize large-scale, methodologically rigorous trials and standardized protocols. Ultimately, the successful transition to metaverse-based care could enhance patient motivation and facilitate accessible, remote neurorehabilitation, provided that real-world feasibility and scalability are addressed. Subsequent research should concentrate on creating standardized intervention regimens and reporting technical and clinical parameters. Longitudinal studies are essential for evaluating the sustainability of outcomes beyond immediate benefits. Integrating mixed-methods approaches that combine clinical data with user experience would improve understanding of usability, adherence, and accessibility, especially for older adults and individuals with cognitive impairments. Furthermore, investigating economic and mobile-oriented iterations of metaverse treatments could enhance scalability. Meaningful progress in this field necessitates sustained interdisciplinary collaboration among neurologists, rehabilitation physicians, physiotherapists, occupational therapists, speech-language therapists, neuropsychologists, biomedical engineers, human-computer interaction specialists, and digital ethicists. Such collaboration is essential to ensure that future metaverse-based interventions are not only technologically advanced but also clinically grounded, functionally relevant, ethically sound, and accessible to the diverse populations they aim to serve in neurorehabilitation practice. Suppliers a.Rayyan; Rayyan Systems Inc. Rayyan; Rayyan Systems Inc. b.Oculus Quest; Meta Platforms, Inc. Oculus Quest; Meta Platforms, Inc. c.Second Life; Linden Lab. Second Life; Linden Lab. Disclosure The authors assert that they have no conflicts of interest. Acknowledgments Acknowledgments The authors would like to thank the scientific and academic community for their continuous contributions to the subject, which laid the groundwork for this study. Researchers whose work was included in the survey inspired and guided this work. Authorship contributions M.H.S.G. formulated and devised the structure of the scoping review. N.F.D. and A.H. performed the literature review and the preliminary evaluation of titles and abstracts. N.F. and M.H.S.G. cooperatively conducted data extraction and synthesis. All authors participated in the analysis and interpretation of the results. M.H.S.G. composed the paper, and N.F. meticulously reviewed it. All authors reviewed and approved the final version of the manuscript. Data statement No new data were created or analyzed in this study. Data sharing does not apply to this article. Declaration of Generative AI and AI-Assisted Technologies in the Writing Process During the preparation of this work the authors used Gemini Pro and QuillBot Pro in order to improve the grammatical structure, style, and overall fluency of sentences. After using this tool/service, the authors reviewed and edited the content as needed and take full responsibility for the content of the publication. Footnotes List of abbreviations: AI, artificial intelligence; AR, augmented reality; DT, digital twin; IMVR, immersive virtual reality; IoT, Internet of Things; MCI, mild cognitive impairment; MS, multiple sclerosis; RCT, randomized controlled trial; TBI, traumatic brain injury; VR, virtual reality. This research did not receive financial assistance from governmental entities, commercial groups, or academic institutions. Supplementary material associated with this article can be found, in the online version, at doi:10.1016/j.arrct.2026.100658. Contributor Information MohammadHossein Sahami Gilan, Email: Sahami3621@gmail.com. Negar Farajzadeh Dehkordi, Email: negar.farajzadeh@nm.mui.ac.ir. Supplementary materials References Associated Data Supplementary Materials