Advancing Trunk Control and Balance in Rehabilitation: A Quantitative Approach Using VR Head-Mounted Display and Motion-Sensing Technologies

doi:10.21203/rs.3.rs-4984010/v1

Virtual reality (VR) technology is becoming increasingly vital across various sectors, including healthcare, engineering, and science. Its applications extend to training, education, clinical evaluations, and rehabilitation. Particularly in rehabilitation, VR is instrumental for assessing and treating a range of conditions. It holds promise for enhancing balance and gait in patients with neurological impairments and offers added benefits when integrated with conventional rehabilitation therapies. Despite its widespread use, there is a notable absence of methods and technologies for the functional quantification of training performance within current VR systems. This study addresses this gap by employing VR head-mounted displays and motion-sensing recognition technologies. This integration facilitates precise positioning and free movement within a VR environment, leveraging motion interaction functions to target trunk control and balance training. The study introduces a functional quantification framework that encompasses algorithms for analyzing movement trajectories, trunk activity, and three-dimensional spatial motion performance. This framework digitally records and updates the rehabilitation journey of participants in real-time, generating functional reports. Such reports enable medical professionals and patients to monitor rehabilitation progress continuously, thereby enhancing the overall efficacy of the treatment process.

Health sciences/Diseases/Neurological disorders

Health sciences/Health care/Therapeutics/Rehabilitation

Virtual reality

Rehabilitation

Functional quantification

Trunk Control

Balance

Virtual Reality (VR) creates immersive 3D environments that offer an engaging alternative to our conventional reality. These environments are increasingly leveraged across training and education sectors, including medicine, engineering, and science, propelling forward advancements in areas like research, virtual archaeology, and remote operations [1–6]. VR stands out by enabling the design of dynamic, ecologically valid environments where interactive responses can be meticulously recorded and assessed, presenting a novel paradigm for clinical assessment and rehabilitation beyond traditional methods [7–10]. Central to VR's effectiveness is its capacity for immersive user experiences and the simulation of virtual embodiment, which are key to its expanding utility in cognitive development and physical rehabilitation therapies [11–14]. Specifically, VR rehabilitation is transformative in the management of neurological conditions such as stroke, multiple sclerosis, spinal cord injury, and Parkinson’s disease, showing promise in enhancing balance and gait. Its integration with conventional rehabilitation techniques can yield additional benefits and adapt well to home-based post-discharge scenarios [15–18]. Empirical evidence suggests that VR rehabilitation can match or exceed the outcomes of standard practices, particularly in improving gait and balance in Parkinson’s patients [19–21]. Balance training through VR can significantly bolster postural and neuromuscular control, diminish sway, and elevate overall balance functionality [22]. Moreover, trunk movement training, whether conducted on stable or unstable surfaces, has demonstrated effectiveness in enhancing trunk control, dynamic sitting balance, and potentially standing balance post-stroke [23, 24]. VR-assisted seated trunk balance exercises have also been linked to improved ambulatory capabilities [25]. The strength and composition of trunk muscles are moderately associated with balance, functional performance, and fall risk in the elderly. Exercise programs like core strength training and Pilates have proven to be practical and highly adherent activities for this demographic [22]. Despite these advances, a gap remains in the quantitative assessment of VR training performance in trunk control and balance exercises. Addressing this, our study presents VR-driven programs that utilize headsets and motion sensing for precise positioning and free movement within VR settings. These programs incorporate diverse interaction modes to aid in trunk control and balance training, underpinned by an algorithmic functional quantification assessment framework.

Equipments

The Pico neo3 Pro Eye head-mounted display system features a Qualcomm XR2 processor with an octa-core 64-bit architecture, peaking at a clock speed of 2.84GHz. It boasts 6GB RAM and an ample 256GB of flash storage capacity (figure 1). The device accommodates WIFI connectivity over dual-band frequencies of 2.4GHz and 5GHz and runs on the Android 10 operating system. Its compact design measures 190mm by 135mm by 112mm. The display impresses with a high-resolution of 3664 by 1920 pixels, and the screen refresh rate alternates between 72Hz and 92Hz, ensuring a 4K ultra-high-definition visual experience. The system is designed to provide a comfortable fit suitable for enterprise use.

The trunk control training software, named 'Dodge-ball Training Software' is an in-house development by Shenzhen Yiwei Technology Co., Ltd. It engages users in an immersive VR environment where they maneuver a virtual sphere to avoid incoming obstacles from various directions. This interactive simulation is aimed at enhancing users' trunk control and balance through dynamic training.

Trunk Control Training System with Virtual Reality Integration

The trunk control training system encompasses a comprehensive VR setup, including a main control terminal in the headset that generates immersive virtual scenes. Integrated within this is a body sensing recognition module, capturing the trainee's head position and movement to create a corresponding virtual body model. This model is then presented in the virtual environment by the visual presentation module, aligned with the trainee's actual physical location for a seamless interactive experience. Central to the system is the motion interaction module, exemplified by applications like Dodge-ball training software, where the trainee engages in dynamic activities, dodging virtual obstacles by controlling the virtual body model. The system also features a motion mode setting module that offers diverse training modes including head-neck control, sitting, and standing exercises. Completing the setup is a function quantification analysis module that meticulously tracks and records the body's six degrees of freedom, encompassing both displacement and rotation across three axes each. This sophisticated analysis enables precise assessment of trunk control and balance capabilities.

Ethics approval and consent to participate

This study has been approved by medical ethics committee of shenzhen prevention and treatment center for occupational diseases. All subjects have signed written informed consent for participation. Additionally, all methods employed in this study were performed in accordance with the relevant guidelines and regulations.

Control of Virtual Objects by Users in Virtual Reality Training Environment

In the Dodge-ball training (figure 2) for trunk control, a virtual object in the form of a ball is presented in different VR training scenarios, positioned in front of the user’s viewpoint. As the user moves the VR headset, the ball follows the movement. The user controls the movement of the ball by moving the VR headset, in order to dodge incoming attacks.

Adaptive adjustment of training difficulty

In the Dodge-ball training module of the virtual reality environment, the system generates attacking objects at random positions along the boundary of a horizontal plane, encompassing a radius of 2 meters and a height of 3 meters around the virtual object ball. The position of these attacking balls is dynamically relative to the virtual object ball's current location, with new balls appearing at each time interval based on a predetermined frequency. For instance, if the frequency is set to generate a ball every second, an attacking ball will appear at coordinate A (the position of the user's virtual object ball at the first second) and move towards it. Subsequently, at the next second, another attacking ball will emerge at coordinate B (the virtual ball's position at that moment) and so on.

The velocity of these attacking balls, measured in meters per second, varies according to the difficulty level of the training session (table 1). Different training modes, including head and neck control, seated trunk movement, and standing trunk movement evaluations, offer varying difficulty settings. The movement of the attacking ball is proportional to the VR headset's displacement, modulated by a specific coefficient: 300% for head training (e.g., a 5cm leftward movement of the headset translates to 15cm for the ball), 200% for seated training, and 100% for standing training, where the ball mirrors the headset's movement without any coefficient.

Table 1

Detailed Levels of Difficulty for Head and Neck, Seated, and Standing Training. This table presents the varying parameters across different types of training categorized by levels of difficulty. Each training type includes five key factors: Diameter of the Virtual Sphere, Diameter of the Attacking Sphere, Speed of the Attacking Sphere, Service Interval, and Movement Coefficient Bonus. Each factor is subdivided into 12 levels of difficulty (L), offering targeted training schemes. This data is crucial for understanding the specific requirements and challenges of each training level, aiding in the design of more precise and effective rehabilitation training programs.

Head and Neck Training / Factors

L1

L2

L 3

L 4

L 5

L6

L7

L8

L 9

L10

L 11

L12

Virtual Sphere Diameter

10cm

11cm

12cm

13cm

14cm

15cm

16cm

17cm

18cm

19cm

20cm

21cm

Attacking Sphere Diameter

5cm

6cm

7cm

8cm

9cm

10cm

11cm

12cm

13cm

14cm

15cm

16cm

Attacking Sphere Speed

0.50m/s

0.55m/s

0.60m/s

0.65m/s

0.70m/s

0.75m/s

0.80m/s

0.85m/s

0.90m/s

0.95m/s

0.100m/s

0.105m/s

Service Interval

3.0s

2.6s

2.3s

2.0s

2.3s

2.6s

2.9s

2.12s

2.15s

2.18s

2.21s

2.24s

Movement Coefficient Bonus

200%

190%

180%

170%

160%

150%

140%

130%

120%

110%

100%

90%

Seated Training / Factors

L1

L2

L 3

L 4

L 5

L6

L7

L8

L 9

L10

L 11

L12

Virtual Sphere Diameter

20cm

21cm

22cm

23cm

24cm

25cm

26cm

27cm

28cm

29cm

30cm

31cm

Attacking Sphere Diameter

10cm

11cm

12cm

13cm

14cm

15cm

16cm

17cm

18cm

19cm

20cm

21cm

Attacking Sphere Speed

0.50m/s

0.55m/s

0.60m/s

0.65m/s

0.70m/s

0.75m/s

0.80m/s

0.85m/s

0.90m/s

0.95m/s

0.100m/s

0.105m/s

Service Interval

3.0s

2.6s

3.1s

2.7s

3.2s

2.8s

3.3s

2.9s

3.4s

2.10s

3.5s

2.11s

Movement Coefficient Bonus

300%

280%

260%

240%

220%

200%

180%

160%

140%

120%

100%

80%

Standing Training / Factors

L1

L2

L 3

L 4

L 5

L6

L7

L8

L 9

L10

L 11

L12

Virtual Sphere Diameter

20cm

22cm

24cm

26cm

28cm

30cm

32cm

34cm

36cm

38cm

40cm

42cm

Attacking Sphere Diameter

10cm

12cm

14cm

16cm

18cm

20cm

22cm

24cm

26cm

28cm

30cm

32cm

Attacking Sphere Speed

0.50m/s

0.60m/s

0.70m/s

0.80m/s

0.90m/s

0.100m/s

0.110m/s

0.120m/s

0.130m/s

0.140m/s

0.150m/s

0.160m/s

Service Interval

3.0s

2.6s

3.1s

2.7s

3.2s

2.8s

3.3s

2.9s

3.4s

2.10s

3.5s

2.11s

The system's difficulty adaptive training mode allows customization of various parameters: the diameter of the virtual object ball, the diameter and speed of the attacking ball, the ball interval, and the movement coefficient. The difficulty levels range from 1 to 10, with an initial setting at level 3. Difficulty is assessed every 30 seconds; if the dodge rate is 80% or higher, the difficulty increases by one level. If the rate is between 50% and 80%, the difficulty remains unchanged. If it falls below 50%, the difficulty decreases by one level.

Quantitative evaluation of trunk control function

The VR headset's trunk control evaluation mode encompasses three key assessments: 1) Head-neck movement evaluation, involving neck flexion, extension, lateral flexion, and rotation; 2) Sitting trunk control evaluation, focusing on trunk movements in a seated position, including flexion, extension, lateral flexion, and rotation; and 3) Standing trunk control evaluation, which extends to trunk movements in a standing position, covering flexion, extension, lateral flexion, and rotation (refer to figure. 3 – 5, a, b, c). The evaluation of the activity trajectory involves establishing the initial position of the headset as the coordinate origin [0, 0, 0]. This process maps out the trajectory of head movements across the horizontal, sagittal, and coronal planes, as depicted in figures 3 to 5 (panels d, e, and f). Additionally, this assessment tracks the maximal distance traversed by the user's head in each direction during the training sessions, which is illustrated in figures 3 to 5 (panel g). Additionally, the training involves motion speed assessment, recording the speed [m/s] and average speed of the user's head movements (figure. 3 – 5, h), and stereo interval motion assessment, which calculates evasion success rate and duration within four directional intervals in the virtual stereo space (figure. 3 – 5, i). This assessment considers the initial position of the attack sphere for space division and uses the VR headset's coordinate data to locate the trainee's interval and calculate the corresponding time proportion (figure. 3 – 5, i). This comprehensive evaluation mode provided by the VR headset effectively integrates various assessments of head-neck and trunk control in different postures, along with detailed analysis of the activity trajectory and motion dynamics, offering a multifaceted approach to understanding and improving the trainee's motor control and spatial awareness in a virtual environment.

The VR headset trunk control training and evaluation system developed in this study offers a novel approach to rehabilitation for conditions such as low back pain, lower limb injuries, and central nervous system disorders, including trunk control and balance impairments typically seen in stroke patients. This system can be integrated with various rehabilitation hardware like balance braces, suspension belts, yoga balls, cushions, and overhead tracks to enhance training safety and applicability. Rehabilitation therapists can tailor the training focus based on individual needs, such as varying the use of hand support for different levels of sitting balance function. However, a limitation of the current system is the lack of targeted training for trunk rotation in dodge behaviors, underscoring the necessity for enhanced focus on rotational movement training.

Contrasting traditional pressure sensor systems that assess balance via center of gravity shifts on a support surface[26], the VR trunk control training utilized here more closely aligns with real functional states.

It offers measurable metrics that reflect trunk activity levels and spatial awareness, such as the extent of trunk movements and the ability to control in a three-dimensional space. In conclusion, this VR system significantly enhances active participation in rehabilitation through gamified methods within an immersive environment. It generates real-time functional reports for both medical staff and patients, offering an updated perspective on rehabilitation progress. Consequently, this system presents itself as an innovative solution for trunk control and balance training and evaluation.

Consent for publication

All authors have read and approved.

Availability of data and materials

Non-identifiable patient data are available upon request to the corresponding author.

Competing interests

The authors declare that the research was conducted in the absence of any commercial or financial relationships that could be construed as a potential conflict of interest.

Funding

The authors gratefully acknowledge the financial support from the Shenzhen Prevention andTreatment Center for Occupational Diseases Scientific Research Cultivation Program Project, under Grant No. SZF-PY-2023-012.

Authors' contributions

Zhang Xu: Research supervision, study design,（Corresponding author）

Xueqiang Zhao: Study design, data analysis, manuscript writing (First Author)

Xiaoqing Wu: Data collection, figure preparation (Co-First Author)

Chi Zhang: Data collection and data analysis assistance

Marco Puggioni, Emanuele Frontoni, Marina Paolanti, ScoolAR, and Roberto Pierdicca. An educational platform to improve students’ learning through virtual reality. IEEE Access, 9:21059– 21070,2021.
Michal Boros, Eva Sventekova, Anna Cidlinova, Martin Bardy, and Katerina Batrlova. Application of vr technology to the training of paramedics. Appl Sci, 12(12):1172, 2022.
Hiu Chow, Janet Hon, Winona Chua, and Audrey Chuan. Effect of virtual reality therapy in reducing pain and anxiety for cancer-related medical procedures: A systematic narrative review. J Pain Symptom Manage, 61:384– 394, 2021.
Pedro Nuno Almeida Barata, Moacir Ramos Filho, and Marcio Vallim Almeida Nunes. Consoli-dating learning in power systems: Virtual reality applied to the study of the operation of electric power transformers. IEEE Trans Educ, 58:255– 261, 2015.
Carlos A Jara, Francisco A Candelas, Santiago T Puente, and Fernando Torres. Hands-on experi-ences of undergraduate students in automatics and robotics using a virtual and remote laboratory. Computers & Education, 57(4):2451– 2461, 2011.
Jean-Baptiste Barreau, Romain Gaugne, Yann Bernard, Guillaume Le Cloirec, and Val´erie Gouranton. Virtual reality tools for the west digital conservatory of archaeological heritage. In Proceedings of the 2014 Virtual Reality International Conference, pages 1 – 4, 2014.
Maria T Schultheis and Albert A Rizzo. The application of virtual reality technology in rehabili-tation. Rehabilitation Psychology, 46(3):296– 311, 2001.
Maria Matsangidou, Angelos P Kassianos, D imitrios Papaioannou, Theodoros Solomou, Maria Krini, Maria Karekla, and Constantinos S Pattichis. Virtual painkillers: Designing accessible virtual reality experiences for helping cancer patients manage pain at home. In Extended Abstracts of the 2022 CHI Conference on Human Factors in Computing Systems (CHI EA ’22), pages 17:1 –9. Association for Computing Machinery, 2022.
Minori Unno, Yusuke Kikuchi, Kentaro Yamaoka, Gaku Sueta, Vibol Yem, and Yasushi Ikei. Action reproducer: Virtual reality rehabilitation system to reduce fear of walking. In ACM SIGGRAPH 2021 Labs (SIGGRAPH ’21), pages 7:1 – 2. Association for Computing Machinery, 2021.
Junyi Zhu, Yuxuan Lei, Aashini S hah, Gila Schein, Hamid Ghaednia, Joseph Schwab, Casper Harteveld, and Stefanie Mueller. Monitoring muscle engagement via electrical impedance to- mographyfor unsupervised physical rehabilitation. In Adjunct Proceedings of the 35th Annual ACM Symposium on User Interface Software and Technology (UIST ’22 Adjunct), pages 110:1 – 3. Association for Computing Machinery, 2022.
Giuseppe Tieri et al. Virtual reality in cognitive and motor rehabilitation: facts, iction and fallacies. Expert Review of Medical Devices, 15(2):107– 117, 2018.
Eric Redlinger, Bernhard Glas, and Yang Rong. Enhanced cognitive training using virtual reality: Examining a memory task modiied for use in virtual environments. In 2021 5th International Conference on Artiicial Intelligence and Virtual Reality (AIVR) (AIVR 2021), pages 1 – 8. Asso- ciation for Computing Machinery, 2021.
S .H.M. Hashim, M. Ismail, H. Manaf, and F.A. Hanapiah. Development of dual cognitive task virtual reality game addressing stroke rehabilitation. In Icvars 2019, pages 21– 25, 2019.
Belinda Lange et al. Designing informed game-based rehabilitation tasks leveraging advances in virtual reality. D isabil Rehabil, 34(22):1863– 1870, 2012.
David C Porras et al. Advantages of virtual reality in the rehabilitation of balance and gait: systematic review. Neurology, 90(22):1017– 1025, 2018.
Yoram Baram and Ariel Miller. Virtual reality cues for improvement of gait in patients with multiple sclerosis. Neurology, 66:178– 181, 2006.
B. A. Orsatti-S ´anchez and O. D iaz-Hernandez. Eicacy of virtual reality in neurorehabilitation of spinal cord injury patients: A systematic review. REVISTA MEXICANA DE INGENIERI ´A BIOME´ DICA, 42(2):90– 103, 2021.
Soie Truijen, Abukar Abdullahi, D ionne Bijsterbosch, Esther van Zoest, Marike Conijn, Yixuan Wang, Norbert Struyf, and Wim Saeys. Effect of home-based virtual reality training and tel- erehabilitation on balance in individuals with parkinson disease, multiple sclerosis, and stroke: a systematic review and meta-analysis. Neurological sciences : oicial journal of the Italian Neu- rological Society and of the Italian Society of Clinical Neurophysiology, 43(5):2995– 3006, 2022.
Chong Lei et al. Effects of virtual reality rehabilitation training on gait and balance in patients with parkinson’s disease: A systematic review. PLoS One, 14(11):e0224819, 2019.
Hong Feng et al. Virtual reality rehabilitation versus conventional physical therapy for improving balance and gait in parkinson’s disease patients: A randomized controlled trial. Medical science monitor : international medical journal of experimental and clinical research, 25:4186–4192, 2019.
Costanza Pazzaglia et al. Comparison of virtual reality rehabilitation and conventional rehabili- tation in parkinson’s disease: a randomised controlled trial. Physiotherapy, 106:36–42, 2020.
Urs Granacher et al. The importance of trunk muscle strength for balance, functional performance, and fall prevention in seniors: a systematic review. S ports Med, 43(7):627– 641, 2013.
Rosa Cabanas-Valdes, Guillermo U Cuchi, and Caritat Bagur-Calafat. Trunk training exercises approaches for improving trunk performance and functional sitting balance in patients with stroke: a systematic review. NeuroRehabilitation, 33(4):575– 592, 2013.
Kohei Haruyama, Michiyuki Kawakami, and Tomoki Otsuka. Effect of core stability training on trunk function, standing balance, and mobility in stroke patients. Neurorehabilitation and neural repair, 31(3):240– 249, 2017.
Kazuki Omon, Masayuki Hara, and Haruo Ishikawa. Virtual reality-guided, dual-task, body trunk balance training in the sitting position improved walking ability without improving leg strength. Progress in rehabilitation medicine, 4:20190011, 2019.
Lee C H, Sun T L. Evaluation of postural stability based on a force plate and inertial sensor during static balance measurements[J]. Journal of physiological anthropology, 2018, 37: 1-16.

No competing interests reported.

Advancing Trunk Control and Balance in Rehabilitation: A Quantitative Approach Using VR Head-Mounted Display and Motion-Sensing Technologies

Status:

Version 1

Abstract

Figures

Introduction

Methods

Results

Discussion

Declarations

References

Additional Declarations

Status:

Version 1