1 Introduction

Stroke is major reason for over 10% of all deaths in many countries, accounting for over 10% of all deaths in the world. Stroke is the biggest reason of mortality and disability in adults, as well as the major cause of disability in countries with a middle- and high-income level of development (Torner et al. 2019). Stroke therapy is an important and continuing part of the healing process. Based on the adverse effect of the stroke, regaining a normal quality of life is often possible with the right assistance and the support of family and friends (Wang et al. 2019). Conventional therapy aims to activate relevant nerve tissue in stroke patients by continual muscle training, aiding patients in retaining existing body muscular strength and regaining their ability to function in daily life as much as feasible. It also uses an evaluation measure to assess the rehabilitation effects (Oh et al. 2019). Hemiparesis of the upper limb is the major disabilities following a stroke which decreases activity and a lower quality of life (Cai et al. 2021). As a result, complete upper-limb weakness recovery is required. Up to a year post-stroke, active rehabilitation tactics can speed up motor recovery on its own (Saposnik and Levin 2011).

Traditional upper extremity rehabilitation treatments after stroke typically involve a combination of physical therapy, occupational therapy, and/or speech therapy. Occupational therapy focuses on enhancing the patient’s ability to carry out daily living activities and regain independence. Studies have shown a disparity between functional recovery and daily use of the upper and prosthetic extremities during inpatient stroke rehabilitation, highlighting the importance of integrating occupational therapy into the rehabilitation process to address this gap (Rand and Eng 2012; Wang et al 2023b). Traditional rehabilitation treatments have limited results, necessitating the development of novel therapeutic approaches (Burdea et al. 2019; Chen et al. 2019).

Assistive devices, such as splints or braces, support the affected limb and improve function. Robotic and exoskeleton devices have been extensively studied for their effectiveness in rehabilitation, offering assisted movements and promoting the recovery of arm functions in post-stroke patients (Cerasa et al. 2018; Taveggia et al. 2016). Robotic devices, such as RUPERT, are developed to provide systematic therapy for upper extremity rehabilitation, offering a low-cost, safe, and easy-to-use option for patients and therapists. These devices facilitate repetitive physical rehabilitation, which is crucial for overcoming upper extremity deficits (Sugar et al. 2007).

Upper extremity rehabilitation post-stroke requires special attention because it is often the most affected part of the body after a stroke. The loss of motor function in the upper extremities can significantly impact a stroke survivor’s ability to perform activities of daily living, which can greatly reduce their quality of life. The upper extremities are critical for a wide range of functional activities, such as reaching, grasping, and manipulating objects. These activities are essential for self-care, work, and leisure activities. After a stroke, many patients experience hemiparesis, which is weakness or paralysis on one side of the body, including the upper extremities (Gauthier et al. 2017). The loss of motor function in the upper extremities can significantly impair a patient’s ability to perform activities of daily living, making rehabilitation of the upper extremities a critical component of post-stroke rehabilitation. The upper extremities have a greater impact on a person’s independence and participation in activities of daily living than the lower extremities (Wang et al. 2023a). Therefore, restoring the upper extremity function is a priority in post-stroke rehabilitation.

In this study, we adopt the following definition of Virtual Reality (VR) technology: VR is a computer-generated, interactive simulation of an environment that can be experienced by users through a specialized device called a head-mounted display (HMD). VR technology consists of both software and hardware components, with the former providing the content or the environment that the user will experience and the latter serving as the means of displaying the virtual environment. Software used in VR systems ranges from simple computer programs to complex game engines that enable the creation of highly realistic and interactive simulations. The hardware typically includes a high-quality display, a computer processor, and input devices such as hand controllers that enable users to interact with the virtual environment. VR can project an immersive 3D environment that enables users to interact with and manipulate virtual objects in real-time. The goal of VR in rehabilitation is to provide an engaging and interactive environment that promotes motor learning and functional recovery (Rutkowski et al. 2020).

Virtual therapy is an alternative method for post-stroke motor therapy that can be used in affiliation with other treatments (Lin et al. 2016). As per motor theory of learning, challenge, rigorous (i.e. more treatments and movements) and recurrent training is predominant for improving neuroplasticity and, therefore, motor functions (Ryselis et al. 2020). VR technology can be used to target specific movements and activities of daily living and provide a motivating and engaging environment for rehabilitation (Afsar et al. 2018). VR rehabilitation provides various advantages in terms of recovery speed and desire (P et al. 2017). Gamification—"the process of introducing play elements and play concepts for a chore or daily activities for motivation and involvement"—can help improve activity by boosting patient contentment (intensity) (Wenk et al. 2021). The methodical approach is achievable due to extensible and individualised rehabilitation designs are possible based on the patient’s limb disorder (Gamito et al. 2017). An economical VR rehab system could also be utilised as a part of the telehealth system (Vanagas et al. 2018) or as a residence rehabilitation tool, as well as a supplement to traditional rehabilitation with less monitoring from a therapist (Fini et al. 2021). Motion sensors combined with VR treatment systems enable for functional evaluation and digital monitoring of patients’ progress (Gervasi et al. 2010, Yates et al. 2016). Virtual reality (VR) has been demonstrated to be beneficial in a range of situations, ranging from military training to the treatment of anxiety disorders (Šalkevičius et al. 2019) and phobias to its ability to operate as an art form (EA et al. 2004). VR and interaction computer gaming have recently arisen as new therapy possibilities in healthcare, including the field of stroke rehabilitation and others (Maskeliunas et al. 2022). VR is a software environment in which users engage with virtual environments using multiple senses and receive “real-time” feedback (Chen et al. 2016a, b). It can be implemented as a part of assisted living environment and integrated with Internet-of-Things (IoT) in smart homes (Maskeliunas et al. 2019).

The use of virtual reality (VR) as an adjunct therapy to conventional rehabilitation approaches for post-stroke upper limb rehabilitation has gained attention in recent years. The use of VR techniques in stroke rehabilitation therapy has shown promising results in improving upper limb function and overall quality of life for stroke survivors (Domínguez-Téllez et al. 2020). Studies have shown that VR can provide a motivating and engaging environment for patients to perform repetitive exercises that are essential for motor recovery (Dias et al. 2019). VR can also allow for a more personalized and adjustable rehabilitation program that can be tailored to the specific needs and abilities of each patient (Laver et al. 2017). The use of VR can provide immediate feedback to patients and therapists, allowing for real-time monitoring and adjustment of therapy (Gaggioli et al. 2014). VR techniques have been found to be effective in improving range of motion, strength, and dexterity in the affected upper limb (Huang et al. 2019). The immersive and interactive nature of VR can also provide patients with a sense of accomplishment and boost their confidence, leading to increased participation and adherence to therapy (Kiper et al. 2022).

Several review studies have investigated the effectiveness of game-based VR upper limb rehabilitation systems in clinical settings. (Laver et al. 2017) conducted a review of randomized and quasi-randomized trials to investigate the effects of VR on upper limb function and activity, as well as gait and balance and global motor function in adults after stroke. The review included 35 new studies and found that VR was not statistically significant when compared to conventional therapy for improving upper limb function. However, when used in addition to usual care, VR showed statistically significant differences in the activities of daily living outcome. The study suggests that VR may be beneficial when used as an adjunct to usual care to increase overall therapy time. (Wang et al. 2022) conducted a systematic search for randomized controlled trials and identified 24 studies, including 793 patients. Their meta-analysis showed that game-based VR upper limb rehabilitation therapy is more effective than traditional rehabilitation in improving patients’ upper limb function and hand mobility. (Bui and Luauté 2021) suggest that VR may be particularly effective when used in combination with conventional rehabilitation approaches, and highlight key features integrated in VR systems that can help maximize therapy outcomes. (Wu et al. 2021) quantified the impact of VR training on stroke patients’ motor performance and found that VR training effectively improves upper limb function. (Zhang et al. 2021) reviewed the literature and concluded that VR interventions effectively improve upper- and lower-limb motor function, balance, gait, and daily function of stroke patients, but have no benefits on cognition. (Webster et al. 2021) investigated the effect of VR interventions on upper limb function in people with MS and found some evidence that VR is effective in improving motor function in the upper limb, but there is no clear consensus on which VR-based approaches are the most effective. (Domínguez-Téllez et al. 2020) conducted a meta-analysis of randomized controlled trials and found that VR interventions have potential benefits on the recovery of upper limb motor function and on quality of life after stroke. These studies suggest that VR interventions can effectively improve upper limb function in post-stroke patients, but more research is needed to determine the optimal conditions for implementing VR in clinical settings.

The current systematic review study differs from the previous reviews in several ways. Firstly, it includes only randomized controlled trials and excludes quasi-randomized trials, which were included in some of the previous reviews. This increases the quality of the evidence being reviewed. Secondly, the study focuses specifically on interactive VR technologies, which involve active participation and feedback from the user, rather than passive VR technologies such as watching videos or images. This allows for a more targeted analysis of the effects of interactive VR technologies on upper extremity rehabilitation post-stroke. Finally, the study aims to identify the most effective interventions within the realm of interactive VR technologies, rather than simply assessing the overall effectiveness of VR interventions compared to traditional rehabilitation approaches.

The aim of this study is to review the research studies on virtual reality (VR) training for upper limb rehabilitation in patients following stroke. The main contributions of this study are:

  • This study distinguishes itself from prior research by exclusively focusing on randomized controlled trials (RCTs), enhancing the quality of evidence compared to reviews that include quasi-randomized trials.

  • It specifically examines interactive Virtual Reality (VR) technologies that necessitate active user participation, rather than passive VR forms, providing a deeper understanding of how interactive engagement impacts rehabilitation outcomes.

  • The methodological rigor, including adherence to PRISMA guidelines and extensive database searches, ensures comprehensive insights into VR’s potential in clinical settings. In future research, a high-standard meta-analysis needs to done to fully evaluate its effectiveness.

The organization of the paper is as follows. Section 2 focuses on the search methodology. The technologies employed in VR rehabilitative therapy, clinical evidence as well as outcomes for VR in stroke motor rehabilitation are discussed in Sect. 3. The paper ends with conclusions in Sect. 4.

2 Search methodology

2.1 Search queries and databases

Following the rigorous standards set forth by the PRISMA guidelines for systematic reviews, our research methodology was designed to ensure a comprehensive and transparent review process. We initiated our systematic review by employing a targeted search strategy, utilizing specific keywords:

“virtual reality rehabilitation," “upper limb," “lower limb," “rehabilitation," “interactive gaming system," and “VR based games."

This strategic keyword selection was aimed at capturing the broad spectrum of virtual reality applications within the domain of rehabilitation. Our search extended across LILACS, PUBMED, IEEE, Web of Science (Clarivate Analytics), and Scopus databases, to ensure a wide-ranging collection of relevant literature. The temporal scope of our search was deliberately set from 2005 to 2021, a period chosen to encompass the evolution and maturation of VR technology in medical rehabilitation contexts.

2.2 Inclusion and exclusion criteria

The initial corpus of 820 articles was subjected to an exhaustive screening process tailored to isolate those most relevant to VR’s applications in medicine. From this pool, 539 articles were deemed outside the scope of this review, 109 were categorized under non-medical uses of VR, and 76 were excluded based on inappropriate titles and abstracts. The remaining 96 studies that most closely aligned with the medical applications of VR were rigorously analyzed. These studies spanned a range of topics including body balance, gait enhancement, Parkinson’s disease, stroke rehabilitation, and other neurological conditions. The criteria for exclusion were strictly adhered to, ensuring the dismissal of articles not directly related to the medical uses of VR, those with irrelevant content, and those poorly characterized by their titles and abstracts. This meticulous selection process culminated in the identification of 96 pertinent studies, primarily randomized controlled trials (RCTs) published since 2014. This highlights the cutting-edge developments in the use of VR for rehabilitation (Campo-Prieto et al. 2021).

2.3 Screening procedure

The screening procedure entailed a dual-stage assessment—initially by title and subsequently by title and abstract—conducted by a primary researcher, with a subset independently reviewed by a second researcher to ensure thoroughness. Articles that successfully passed these stages were subjected to a full-text review for final inclusion, with the findings double-checked by two additional researchers. This rigorous selection mechanism, rooted in the PRISMA guidelines, ensured that our study was built on a foundation of high-quality, relevant research, focusing on VR’s transformative potential in treating various medical conditions through rehabilitation Fig. 1.

Fig. 1
figure 1

A PRISMA diagram for exploring articles in databases

Full size image

3 Results of review and analysis

VR technology permits consumers to immerse themselves in a computer-generated environment. Users of VR are immersed in an artificial (virtual) universe, with realistic impressions of a virtual software environment where both genuine and implausible occurrences might occur (Archambault et al. 2014). As a result, end users could engage and interact organically with virtual items, as if they are in a real-life situation, and they might not even realise they’re in one. Since VR technology reacts to a user input in a realistic aspect, like revealing the relevant video or picture on the screen during the usage time, the interaction produces a sense of living in a virtual world (Gibbons et al. 2016). The user’s ability to control the character’s body motions can lead to a sense of owning the character’s body parts as proxy for their own, which is known as “virtual immersion” occurs. Users are totally immersed because of these characteristics, allowing them to have a realistic experience of where they’re from what they are doing there (Garcia-Hernandez et al. 2021).

Studies have shown that VR can improve upper limb motor function and daily living activities in stroke patients. The Khymeia VR Rehabilitation System (VRRS) has been found to be effective in restoring upper limb motor function in acute stroke patients (Turolla et al. 2013), while the Jintronix VR system using Microsoft Kinect has been shown to be effective in sub-acute and chronic stroke patients (Norouzi-Gheidari et al. 2013; Archambault et al. 2014). TUIs, such as the PneuGlove, have been designed to train grasp-and-release movements, which are important for daily living activities (Connelly et al. 2010).

3.1 Techniques for movement tracing and input in virtual rehabilitation

VR is a form of game in which software content is synchronised with the user’s movements or a motion guide is provided. Several research studies have been completed to enquire the use of VR in upper limb motor restoration (Table 1). Motion detection sensors integrated with VR system can be used to capture the area of the patient’s body, which is done by synchronising and transmitting the acquired data to the VR system for limb motor therapy (Connelly et al. 2010; Xie et al. 2021). For movement visualisation, you’ll need a mouse and joystick, as well as intensity cameras, electromagnetic sensors, inertial measurement units, bending detectors, monitoring devices, and other detectors that track the participant’s motion (Glueck and Han 2020; Silva et al. 2020). Sensor performance is critical for accurate motion tracking (Babaei et al. 2022), but subjective opinion and choices, as well as cost, are also key considerations. Motion-detecting sensor technology is required for virtual rehabilitation (Choi and Paik 2018; Zahabi and Abdul Razak 2020). Wearable and nonwearable devices that detect upper-limb rehab movements are two examples of this type of technology. Nonwearable devices also are subdivided into those using a visual detector alone or in conjunction with an automation control or a three degree-of-freedom controller (DOF) (Rogers et al. 2021). Wearable devices are often separated into two types: data gloves and exoskeletons. Both forms have been utilised in several investigations (De Araújo et al. 2019; Finley et al. 2021; Weber et al. 2019). The tracking of markers or colour patches with webcams in nonwearable devices has lately shifted to monitoring body or hand gestures using proximity sensing technologies. Within a confined space with no barriers, the users’ movements are sensed in this way (Foulds et al. 2008). The sensor is connected to wearable devices to gather high-frequency data, including force, torque, position, and movement (Arienti et al. 2019).

Table 1 Summary of various sensors and VR types used in the analyzed studies
Full size table

Most studies have used content-based visual and aural input, with some experimenting with touch and force feedback as well. The research was divided into two categories: visual—motor input and visual-haptic feedback (Norouzi-Gheidari et al. 2013; Threapleton et al. 2016). Visual—motor input provides visual feedback by using sensors to apply measured motions to material in real time. Visual-haptic feedback is the combination of visual information and haptic feedback (Gauthier et al. 2017). Tactile feedback uses vibration, skin deformation, or minor forces to deliver input to end-users via their touch sense (Kerous et al. 2018). Force feedback, also known as kinesthetic force feedback, uses motorised motion or resistance to imitate real-world physical contact rather than precise touch (Rutkowski et al. 2021, Palma et al. 2017). If fine motor tracking is performed during upper limb rehabilitation, and if wearable or nonwearable equipment are used, virtual rehabilitation research can be characterised (Stetz et al. 2011; Shum et al. 2019). Hand tracking is possible using a VR HMD, which has already been released as a commercial product (e.g., Oculus Quest) in the case of camera approaches among nonwearable devices (McDermott et al. 2021).

3.2 VR in motor recovery after stroke: clinical evidence

Table 2 shows the key characteristics of the meta-analyses that were included. Nine studies included randomized control study, one study used Tangible user interface, five studies used Microsoft Kinect camera, one study used Khymeia VR Rehabilitation System (VRRS), two studies used mobile VR games, two studies used Nintendo Wii game therapy, one study used Leap Motion-based 3D Immersive Virtual Reality, one study utilized Control weight therapy, one study used Joystim (3-dimensional manipulator) for control trial (Ahmed et al. 2020). Table 2 also provides details of aim, methodology, number of participants, age range and intervention period of the study.

Table 2 Survey of the clinical evidence and characteristics used in the analyzed studies
Full size table

Several studies have investigated the effectiveness of using virtual reality (VR) technology for stroke rehabilitation. One study by (Connelly et al. 2010) tested the PneuGlove, a pneumatic glove that provides independent extension help to every finger, for training grasp-and-release movements in stroke patients with chronic hand hemiparesis using immersive VR. (Hilton et al. 2011) used a tangible user interface with a 3D model of a kitchen operated by a touch screen to provide spatial and orientation data to the computer to test VR rehabilitation in acute stroke patients. (Norouzi-Gheidari et al. 2013) used the Jintronix VR system with Microsoft Kinect Camera for sub-acute hemiparesis stroke patients, while (Turolla et al. 2013) used the Khymeia VR Rehabilitation System (VRRS) in acute stroke patients to improve upper limb motor function and daily task performance. (Archambault et al. 2014) tested the Jintronix VR system in chronic stroke patients for arm mobility rehabilitation, and (Gamito et al. 2015) developed a severe gaming program for psychological treatment based on VR in patients above 40 years. (Lin et al. 2016) used interactive VR games to improve the enthusiasm of patients undergoing upper extremity rehabilitation treatment and to store patients’ rehabilitation data, while (Saposnik et al. 2016) compared VR to recreational therapy for limb movement recovery in acute ischemic stroke patients. (Gauthier et al. 2017) compared conventional hospital-based constraint-induced mobility therapy with video game-based constraint-induced mobility therapy in chronic upper extremity hemiparesis, and (Choi and Paik 2018) developed a mobile VR upper extremity rehabilitation program leveraging game applications. Finally, (Afsar et al. 2018) evaluated the effects of the Microsoft Xbox 360 Kinect video gaming system on stroke patients’ upper extremity movement functions.

3.3 Measured outcomes and conclusions from VR activities

VR rehabilitation was used as an intervention, while traditional therapy or no intervention was used as a control. Table 3 gives the summary of the measured outcomes, findings, and conclusion in the analyzed studies. In the selected meta-analyses, both specially designed virtual worlds and professional video gaming systems are used in VR rehabilitation (Killane et al. 2015). (e.g., Nintendo Wii or Microsoft Xbox Kinect). Most of the time, the results were a composite of upper limb function or activity (Saposnik and Levin 2011; Rogers et al. 2021). The Fugl–Meyer Assessment of Upper Extremity Function (FMUE), Functional Independence Measure (FIM) scale, The Wolf Motor Function Test (WMFT), The Brunnström stage (Bstage) for the arm and hand, the Box & Block Test (BBT Score), Chedoke McMaster Arm and Hand Activity Inventory (CAHAI), Stroke Impact Scale (SIS) and Performance Assessment of Self-Care Skills (PASS) were the major outcomes measured in the studies analyzed.

Table 3 Summary of the measured outcomes, findings, and conclusion in the analyzed studies
Full size table

The effectiveness of upper extremity rehabilitation was measured in five studies using different outcome measures. (Connelly et al. 2010) measured improvements in hand component of the Fugl–Meyer Assessment of Upper Extremity Function, palmar pinch and Box and Blocks test, and found that the PneuGlove group improved twice as much compared to the non-PneuGlove group immediately after training. (Turolla et al. 2013) utilized the Fugl–Meyer Upper Extremity and Functional Independence Measure scales, and found that VR rehabilitation was more effective than ULC therapy alone in increasing scores. (Lin et al. 2016) found that group C had the highest rehabilitation effectiveness compared to group B and group A using the Fugl–Meyer Assessment of Upper Extremity. (Saposnik et al. 2016) used the Wolf Motor Function Test and found that simple low-cost leisure activities were as effective as VR systems. Finally, (Choi and Paik 2018) utilized the Fugl–Meyer Assessment of Upper Extremity, Brunnström stage, and manual muscle testing and found that mobile phone-based VR app significantly encourages stroke patients to regain use of their upper extremities without experiencing any negative side effects.

The findings from the studies presented in Table 3 provide insightful comparisons between various interventions for upper limb rehabilitation post-stroke, particularly focusing on VR and its enhancements with additional tools like haptic gloves or conventional therapies. The addition of a haptic glove to VR therapy may significantly enhance outcomes such as grip strength and lateral pinch, suggesting that tactile feedback could play a vital role in rehabilitation effectiveness. For instance, grip strength improved markedly from 113 to 161 when a haptic glove was used alongside VR compared to VR alone (Connelly et al. 2010).

Comparisons between VR-enhanced interventions and traditional therapies show differing impacts. For example, the Fugl–Meyer Assessment of Upper Extremity (FMA-UE) scores and Functional Independence Measure (FIM) scale scores generally improved more in VR-enhanced settings than in conventional therapy alone, suggesting that VR may facilitate more significant improvements in motor function and independence (Lin et al. 2016; Choi and Paik 2018; Norouzi-Gheidari et al. 2020). Outcomes such as the Wolf Motor Function Test (WMFT) illustrate the effectiveness of VR over recreational activities, with the VR group showing a more substantial improvement (Saposnik et al. 2016; Lee et al. 2021). This highlights VR’s potential to provide targeted, effective rehabilitation exercises that can be more engaging and yield better outcomes than less structured activities.

Across various settings, groups receiving VR interventions (whether immersive or non-immersive supplemented devices like the Leap Motion Controller) often showed greater improvements in measures like the Brunnstrom stages of arm and hand recovery and the FMA-UE scores (Choi and Paik 2018; Afsar et al. 2018). This suggests that VR interactive rehabilitation methods can be comparatively more effective in advancing motor recovery rates than tedious traditional methods. Studies incorporating additional technology with VR, such as motion-tracking gloves (Connelly et al. 2010) or other sensor-based feedback mechanisms such as Kinect (Afsar et al. 2018), often report higher efficacy in rehabilitation outcomes. This suggests that the integration of VR with other technologies might provide compounded benefits by enhancing sensory feedback and motor learning.

These findings underscore the importance of specifying intervention types and comparing them rigorously within the context of rehabilitation studies. By clearly stating what each intervention group receives and the duration of these interventions, we can better assess the specific contributions of VR technologies to rehabilitation outcomes. This is crucial for minimizing biases and ensuring that conclusions drawn about VR’s effectiveness in rehabilitation are based on robust, comparative data.

3.4 Latest research trends

The most recent studies, published after the systematic review was completed, highlight a growing trend in the use of Virtual Reality (VR) for upper limb rehabilitation across various neurological conditions, including cerebral palsy, Parkinson’s disease, and post-stroke recovery. These articles collectively demonstrate an evolving research stream that seeks to enhance traditional rehabilitation methods through technological integration.

The studies by Alrashidi et al. (2023) and Bissolotti et al. (2024) demonstrate the diverse applications of VR in treating cerebral palsy and Parkinson’s disease, respectively. Alrashidi et al. highlight the uncertain benefits of VR compared to traditional therapy in cerebral palsy, advocating for more adaptable VR solutions, while Bissolotti et al. examine VR’s effectiveness in assessing and enhancing motor performance in Parkinson’s disease. Chen et al. (2023) and Castillo et al. (2024) further explore the efficacy and customization potential of immersive VR in post-stroke rehabilitation, with Chen et al. affirming its safety and Castillo et al. focusing on user-centered design for better alignment with patient needs. Aguilera-Rubio et al. (2024) discuss low-cost VR options like the Leap Motion Controller® that enhance patient engagement and motor function affordability. Masmoudi et al. (2024) introduce innovative methods using camera-based sensors to measure engagement, highlighting the significance of emotional insights in VR rehabilitation. Hao et al. (2023) compare immersive and non-immersive VR, finding the former more beneficial for upper extremity recovery in stroke survivors. Lastly, Banduni et al. (2023) detail the synergistic use of VR with rTMS, suggesting enhanced neuroplasticity and motor relearning in stroke rehabilitation.

These studies suggest several key trends: the necessity for personalized and adaptable VR systems that cater specifically to the therapeutic needs and conditions of patients; the importance of integrating innovative technologies to enhance the effectiveness and engagement of rehabilitation processes; and the potential for VR to complement or even surpass traditional rehabilitation methods in certain scenarios.

3.5 Discussion on the usage of VR in stroke therapy

One significant challenge in the use of VR is motion sickness, which occurs due to discrepancies in sensory input and user interaction. HMDs need to be lightweight, stable, and comfortable, as prolonged use can induce visual fatigue or physiological tiredness, often resulting in motion sickness. Advances in HMD technology and interaction techniques could mitigate these effects and enhance user comfort during extended rehabilitation sessions (MF et al. 2020; Mirelman et al. 2014; Hogan et al. 2021; Liu et al. 2020).

The ecological validity of VR interventions—how well the skills learned in VR transfer to real-world contexts—is critical. Although VR creates a controlled environment that standardizes rehabilitation, the skills acquired must be applicable to everyday activities. The disparity in performance between virtual and real tasks suggests the need for VR systems to better simulate real-life scenarios to ensure the transferability of rehabilitative gains (Heinrich et al. 2021; Lee et al. 2017; Rogers et al. 2019; In T, 2016; Montoya et al. 2020). Aligning the VR interface with the real-world requires work aspect could improve the VR therapeutic effect’s transfer. Therefore, it is essential to use outcome measures that are specific to the skills trained in VR, such as tasks involving object manipulation and reaching. Furthermore, incorporating tasks that require interaction with a dynamic environment, such as obstacle avoidance or dual-tasking, may further improve ecological validity.

Gamification in VR rehab has shown promise in increasing patient engagement and motivation by making rehabilitation more enjoyable. However, the design of these systems must cater to varied abilities, particularly in older patients, to prevent frustration and disengagement (Martinez-Cesteros et al. 2021; Lin et al. 2021; Keshner et al. 2019).

Despite its benefits, VR implementation faces several barriers, including the cognitive and technological capabilities of patients, and therapist familiarity with VR systems. Overcoming these challenges requires tailored training for therapists and the development of user-friendly VR systems that accommodate the specific needs of post-stroke patients (Liberatore et al. 2021; Selves et al. 2020; Safir and Wallach 2011)

The generalizability of VR research is limited by small sample sizes, short intervention durations, and a lack of long-term follow-up, which may not capture the sustained benefits of VR or represent the broader stroke survivor population. Future research should focus on larger, more diverse populations and extended study periods to more comprehensively evaluate the efficacy of VR interventions (Cai et al. 2021; Chen et al. 2016a, b; Yates et al. 2016; Maggio et al. 2019).

Therefore, VR represents a significant advancement in the rehabilitation of stroke survivors, offering customizable, engaging, and effective interventions that can significantly enhance motor recovery and improve quality of life. However, more research is needed to optimize these systems, ensure their ecological validity, and determine the best implementation strategies for diverse patient populations.

4 Conclusions

This study presents a systematic review of the clinical evidence and techniques to adopt virtual reality in post-stroke upper limb rehabilitation. The review period covered the years up to 2021. The research gap found in the already existing methods are not customized according to the patients’ needs. The main challenge will be designing of VR rehabilitation system based on the ability and needs of the patients. The user interface for patients and therapists must be improved, considering both the patient’s physiological and cognitive limitations as well as the therapist’s needs. VR can be incorporated into new therapeutic methods that increase neuroplasticity (e.g., BCI and non-invasive neural stimulation) and results in improved recovery when used in tandem, which needs more investigation.

Ecological validity is essential to ensure that the benefits of VR rehabilitation translate to functional gains in daily living. VR training has the potential to optimize rehabilitation programs and outcomes in several ways. Firstly, VR training allows for a high degree of customization and variability, which can help to address the individual needs and preferences of patients. Secondly, VR training can provide a more engaging and motivating environment for patients, which can lead to increased participation and effort during therapy sessions. Thirdly, VR training can provide a safe and controlled environment for patients to practice functional tasks and movements, which can be challenging or risky in the real world. The effectiveness of VR training compared to traditional therapy is not yet fully understood. While some studies have shown that VR training can lead to greater improvements in upper extremity function and daily activities, others have found no significant differences between VR and traditional therapy. It is possible that the benefits of VR training are related to the unique features of the platform, such as the ability to provide real-time feedback and variability, as well as the motivating and immersive environment. However, it is also possible that the amount of time provided in VR therapy plays a role in the outcomes, as patients may receive more focused and intensive therapy in VR sessions.

The sample size of the included studies varied widely, ranging from 5 to 96, with most of the studies being conducted in adults and seniors with chronic stroke. The interventions varied in terms of duration, frequency, and the types of virtual reality systems used. The study only focused on outcomes related to quality of life, dynamic steadiness, upper extremity movement function, and daily tasks, which may limit the generalizability of the results. Finally, the researchers identified a gap in the existing methods, highlighting the need for customized VR rehabilitation systems that consider both the patient’s physiological and cognitive limitations and the therapist’s needs. Despite these limitations, the study provides insights into the potential of virtual reality as an adjunct to traditional rehabilitation approaches for post-stroke upper limb rehabilitation.

There are several future directions that could enhance VR adoption in this field. One direction is to personalize VR rehabilitation by creating interventions that are tailored to each individual’s needs and adapt to their progress. Another direction is to incorporate other senses such as touch, proprioception, and haptic feedback to enhance the immersive experience and provide a more realistic simulation of functional tasks. There is a need for more research on the effectiveness of VR rehabilitation for chronic stroke patients who have plateaued in their recovery. Combining VR with other rehabilitation techniques such as constraint-induced movement therapy or mirror therapy could also enhance its effectiveness. Lastly, there is a need for larger-scale clinical trials to establish the efficacy and cost-effectiveness of VR compared to traditional rehabilitation methods.