Engineering and experimental evaluation of a smart wireless glove for gamified upper limb rehabilitation in Parkinson's disease

Abstract

Background

Upper extremity rehabilitation is critical for individuals with neurological disorders such as Parkinson's disease, where motor impairments significantly affect daily functionality.

Objective

This study presents the design and prototyping of a novel rehabilitation glove aimed at improving hand and wrist motor recovery through gamified therapy.

Methods

The proposed glove features wireless communication via a Wi-Fi network, adaptability to various hand sizes, and an integrated strength training mechanism using resistance bands and metal hooks. The glove consists of a lightweight forearm frame, a palm component, and a textile glove with embedded mechanical connectors, all designed based on anthropometric data. 3D printing techniques were employed using PLA and flexible TPU materials to create a modular, low-cost, and comfortable structure. The system interfaces with an interactive game, allowing users to control an avatar (bee) through hand movements, promoting motivation and active engagement.

Results

The prototype effectively addresses key limitations of previous rehabilitation systems, including mobility, comfort, adaptability, and functionality.

Conclusion

The proposed rehabilitation glove offers a promising solution for home-based neurorehabilitation.

Keywords

Hand rehabilitation wireless gamified therapy smart glove Parkinson's disease

1 Introduction

Advancements in rehabilitation technologies have led to the development of innovative assistive devices that enhance motor function recovery in individuals with neurological and musculoskeletal disorders. Upper extremity rehabilitation plays a crucial role in improving independence and quality of life for individuals with motor impairments, particularly those recovering from stroke, spinal cord injuries, and neuromuscular disorders. Research have shown that early, intensive rehabilitation interventions play a critical role in enhancing motor function and minimizing long-term disability.^1,2 The effectiveness of VR-based interactive rehabilitation techniques has been widely studied and indicates that interactive or VR-based rehabilitation can improve patient motivation, adherence to therapy, and functional outcomes compared to traditional rehabilitation methods.^3–6 These approaches leverage real-time feedback and gamification elements to create engaging rehabilitation experiences that stimulate neuroplasticity and motor learning.^7–10 In addition to improving the quality of rehabilitation, such technologies can benefit clinicians by enhancing patient engagement, providing precise data tracking, allowing the simulation of various environments and potentially alleviating workloads.¹¹

Rehabilitation of fine motor skills, especially of the wrist, hand, and fingers, is essential for restoring daily functional abilities. However, the development of effective rehabilitation systems is complicated by the hand's complex biomechanics and faces various challenges, such as mechanical complexity, sensorimotor integration, or personalization.^12–15 These challenges become even more complex in neurodegenerative diseases such as Parkinson's disease (PD), where progressive symptoms—including tremors, rigidity, and bradykinesia—cause significant impairments in fine motor control, ultimately reducing patients’ independence in daily activities, like performing household tasks,¹⁶ using a touchscreen, pulling on socks, dialing a phone, gripping function, etc.^17,18 One of the primary challenges in upper extremity rehabilitation for PD is the variability in motor performance due to medication cycles and disease progression.¹⁹ Existing rehabilitation devices often fail to cope with the dynamic and fluctuating nature of PD symptoms. These rehabilitation systems generally lack the capability to dynamically adjust the difficulty of therapeutic exercises to meet individual patient needs. Additionally, they often rely on specialized hardware that is impractical for home use due to its large size, high cost, and operational complexity, making them primarily suitable for clinical settings. Socially assistive robotic (SAR) systems have been introduced for specific applications, such as seated stretching exercises and tabletop rehabilitation games. However, their effectiveness is constrained by the absence of direct physical interaction with users.^20,21

Recent studies have highlighted the effectiveness of motor imagery-based rehabilitation gloves, which utilize VR technology to amplify weak movement signals and enhance motor control.^22–26 Such systems employ high-precision pressure sensors, control units, and VR interfaces to provide real-time feedback and improve patient engagement.^27,28 Additionally, various commercially and non-commercially available gamified rehabilitation systems, such as ‘SY-HR06E’ (Figure 1(a)), ‘Rapael’ (Figure 1(b)), ‘Hand of hope’ (Figure 1(c)), and ‘RehaBEElitation Serious Game’ (Figure 1(d)) have demonstrated the potential benefits of integrating biomechanical analysis, multi-sensory stimulation, and adaptive training modules into rehabilitation protocols.

Figure 1.

Rehabilitation system analogues: (a) hand rehabilitation robot ‘sy-HR06E’.²⁹ (b) Hand mobility rehabilitation system ‘RAPAEL’.³⁰ (c) Hand mobility rehabilitation system ‘HAND OF HOPE’.³¹ (d) ‘RehaBEElitation Serious Game’ two different iterations.^16,32

Despite advancements, very often existing rehabilitation glove designs exhibit limitations such as: (1) restricted adaptability for different hand sizes, therefore limiting widespread usability and customization,²⁸ (2) lack of wireless functionality, which limits mobility and convenience for home-based therapy,^29,31 (3) heavy construction, which reduces comfort and prolonged use feasibility,³¹ (4) cannot perform active or passive rehabilitation^30,33 and (5) cannot train or track movements in all planes.^29,31,33 In order to eliminate these limitations, various improvements can be made. First of all, it would be possible to make a main control device with a microcontroller attached to the wrist or forearm, which would collect information about the position of the hand and send it to a smart phone or other device via wireless network. This would allow such a glove to be freely used without connecting it to other parts of the system and send the necessary instructions for the operation of the glove. Second, when designing the system, the design and material of the glove and the main frame can be changed. It would make sense to use a very elastic, but also sufficiently well-fitting material to make the glove so that it could be adapted to large and small hands, and to make the frame more universal. Such solutions would also help to lighten the weight of the system itself. Third, resistance factors integrated into the glove structure would be suitable for strength training. On the front of the palm, just above the knuckles, hooks can be designed to hold resistance bands, and at the other end, they are attached to the fingers. Bending the fingers creates resistance, and thus, by adjusting the length and stiffness of the bands, the strength of the hand muscles is being trained. This feature would not only help with strength training but also facilitate rehabilitation process for people who are training their flexion function but find it difficult to perform the extension movement on their own: it helps them perform the extension movement.

To address these challenges, this study focuses on the prototyping of a novel rehabilitation glove. Unlike existing rehabilitation gloves, our prototype integrates a main control device with a microcontroller attached to the forearm that collects information about the position of the hand and sends it to the smart display via Wi-Fi network. This is for addressing the limitation of wireless use. Beyond that, the glove introduces a unique combination of wireless motion capture and resistance-inducing elements (bands and hooks) for active strength training, which has been largely overlooked in previous analogues. Another important improvement is the interactive rehabilitation functionality. The system is designed to provide real-time feedback and gamified exercises via the smart display, encouraging patient engagement and motivation. Importantly, the interactive rehabilitation games are specifically adapted to the needs of PD patients, considering their motor limitations such as tremor, slowness of movement, and reduced grip strength. Finally, the construction of the glove is optimized to achieve a lighter weight compared to other analogues and to ensure adaptability to a variety of hand sizes. Taken together, these features represent a step beyond the state-of-the-art, as the proposed glove integrates wireless monitoring, strength training, and Parkinson's-specific interactive rehabilitation in a single device.

2 Methods

2.1 System architecture

The proposed gamified rehabilitation system integrates a wearable sensing device with wireless data transmission and an interactive virtual rehabilitation environment. The system is composed of two main modules: (1) the wearable sensing module, consisting of a palm segment and a forearm segment embedded with sensors, a microcontroller, and wireless communication components; and (2) the host computer running the rehabilitation software, which processes motion data, translates it into control commands, and provides real-time feedback through gamified exercises.

Figure 2 presents the structural diagram of the gamified rehabilitation system, which integrates the system's operation and its hardware components.

Figure 2.

Structural diagram of the gamified rehabilitation system.

2.2 Design rationale

The part of the gamified rehabilitation system being developed is a hand device. In this system, a patient with PD controls a bee in virtual reality³⁴ with the bee's movements corresponding to the patient's real-world hand movements. To enable wireless interaction and accurately reflect hand movements in the game, a system of electronic components is required. These components must be strategically placed on the patient's body to allow continuous monitoring, recording, analysis, and storage of movement data.

The most convenient and effective way to achieve this is to create a structure that conforms to human anatomy and anthropometry, ensuring the electronic components are securely and accurately positioned. Since this project focuses on upper limb rehabilitation, particular attention is given to hand movements. The device consists of two main interconnected parts: a rigid frame covering the forearm and hand, and a textile glove worn underneath the frame. The frame was designed first due to its structural complexity, while the glove was subsequently adapted to fit the completed frame, as it requires mechanical integration with the frame.

During the development of design alternatives, special focus was placed on the connection between the hand and forearm, particularly around the wrist area. This connection is crucial to avoid restricting wrist joint mobility and to ensure that the forearm section remains as universal as possible, accommodating different users.

Additionally, the design needed to fulfill several functional requirements: first, the proper positioning of electronic components—these must be firmly, precisely, and conveniently attached to ensure an effective rehabilitation process. Second, mechanical functionality must be ensured—specifically, the incorporation of resistance elements (hooks and resistance bands). Therefore, the device must be securely attached to the patient's body, with all elements not only tailored to human anthropometry and anatomy but also mechanically interconnected to maintain structural integrity without breaking or deforming under stress.

SolidWorks was used throughout the design process to develop components and explore alternative design concepts, thanks to its extensive parametric modeling capabilities, ease of arranging component relationships, load simulations, and other advanced design tools.

2.3 Anthropometric calculations

After the frame variant design was carried out and the most functionally appropriate structure was determined, the geometry of the device of the required size was created, and its dimensions were determined. For this, calculations were made based on anthropometric values.³⁵

The height of the person for the calculations was chosen from 150 to 200 cm according to the average height of adults and the maximum and minimum limits, to make the device suitable for as many adults as possible. PD most often affects people from 50 to 60 years old, and older people are usually shorter in height. However, height was not the most important variable, since some components of the device, such as resistance rubber bands and flexible plates, can be of several different sizes. Therefore, it is possible to make the device suitable for many people by applying various combinations of these parts. Calculations were performed for the following parts, with a maximum patient height of 200 cm and a minimum of 150 cm: palm, forearm, and flexible plates.

It is important that the frame covers only part of the hand – it does not reach the fingers and is positioned only on the outer part of the palm and ends at the wrist joint. The approximate proportions of the hand are as follows: the length of the palm is 40–50% of the entire length of the hand, while the length of the fingers (from the base of the fingers to the tips) is about 50–60% of the entire length of the hand.³⁶ Thus, the required maximum and minimum lengths of the palm and part of the fingers were calculated by taking the averages between the proportions – the length of the palm is about 45%, and the length of the fingers is about 55% of the length of the hand. The hand also contains finger joints, the movement of which should not be limited by the palm component of the frame. Therefore, the geometry of the palm component should not completely cover the entire palm, as well as the wrist joint, so as not to limit the movement of the wrist.

When designing the forearm component, the length and circumference of the forearm were considered. The forearm component is intended to accommodate the electronic elements with the greatest mass, including the main control unit and the battery. This decision is based on the need to minimize the load on the distal part of the limb, which is particularly important for users with limited muscle strength. Concentrating the heaviest components on the forearm reduces the inertial load on the hand and enhances overall wearability and comfort. The length and circumference of the forearm part were selected considering the average anthropometric and electronic component sizes.

Conversely, the hand component is designed to be as lightweight and unobtrusive as possible. This component houses only a necessary sensor unit and hooks for attaching thumb resistance rubber.

The most important part of the frame design is the selection of the connection between the forearm and the palm components. The wrist is a joint with 3 degrees of freedom and performs flexion/extension, supination/pronation, and ulnar/radial deviation. To maintain these movements while wearing the device on the hand, a structure is required that does not restrict the movements. For this purpose, a connection between the palm and the forearm of special geometry, construction, and material was designed, which will consist of plates of several different sizes. So, for example, having six plates of different lengths, it is possible to compensate for the differences in length between the palm and forearm segments and ensure that the connection coincides with the center of the wrist joint.

Another part of this system is a glove that will be placed under the frame and to which hooks will be attached to tighten the resistance rubber bands between the phalanges of the fingers, thus forming a kinematic chain. The hooks were designed with a special groove so that the rubber bands do not slip and stay in the same position. 6 different lengths (2 different lengths for each phalanx) of the rubber bands were designed for each finger to be able to change the amount of resistance force (for example, choose shorter bands to make the tasks more difficult) and to adapt to different lengths of patient's fingers.

2.4 Material selection and manufacturing techniques

The hand device of the gamified rehabilitation system must be made from lightweight yet durable materials. For the forearm, palm parts, pin, hooks, boxes, and covers, PLA plastic was chosen due to its strength, skin safety, smooth finish without additional processing, easy cleaning, and biodegradability.^37,38 PLA's deformation temperature is 56 °C, and its elastic modulus is 2–4 GPa.³⁹

Flexible plates connecting the forearm and palm parts are made of FLEX CONDUCTIVE plastic (flexible thermoplastic polyurethane (TPU)), offering flexibility (87% elongation at break) and a deformation temperature of 55 °C.⁴⁰

For the fastening strips, a 15 mm wide Velcro tape is selected, ensuring flexibility and adaptability to different hand sizes. And for the resistance bands, 1 mm diameter rubber was chosen for its elasticity, light weight, and ability to return to its original shape, which is essential for continuous hand and finger movement during rehabilitation. Band lengths were based on average phalange lengths, with the option to adjust resistance by varying band properties and lengths.

The electronic components are housed in PLA boxes and connected by insulated wires to ensure safety. The textile gloves are made from a spandex and polyester blend, chosen for their durability, elasticity, breathability, and comfort. The frame hooks are 3D printed from PLA plastic for uniformity, while the glove hooks will be made of lightweight, corrosion-resistant 304 stainless steel for enhanced strength and durability. A summary of the selected materials is presented in Table 1.

Table 1.
Selected materials for the structural parts of the device.

Hand device structural element Material

Forearm, palm parts, pin, hooks, boxes, covers PLA plastic

Fastening strips Velcro tape, self-adhesive and with hooks, attached to an elastic rubber band

Resistance bands Rubber covered with textile

Gloves Polyester and spandex, covered with Lycra material

Flexible plates FLEX CONDUCTIVE plastic (flexible TPU)

Glove hooks 304 stainless steel

Hand device structural element	Material
Forearm, palm parts, pin, hooks, boxes, covers	PLA plastic
Fastening strips	Velcro tape, self-adhesive and with hooks, attached to an elastic rubber band
Resistance bands	Rubber covered with textile
Gloves	Polyester and spandex, covered with Lycra material
Flexible plates	FLEX CONDUCTIVE plastic (flexible TPU)
Glove hooks	304 stainless steel

The textile parts and glove hooks were sewn or glued. 3D printing was used to manufacture the forearm, palm parts, pin, hooks, boxes, and their covers, with a Prusa printer selected for this task. 3D printing was also chosen to produce flexible plates, but the Zortrax M300 Plus printer was selected due to its specific characteristics. The main parameters of the selected printers are provided in Table 2.

Table 2.

Printing parameters of the 3D printers.

Parameter	Parameter setting of ‘Prusa’	Parameter setting of ‘Zortrax M3000 plus’
Filament diameter	1.75 mm	0.40 mm
Layer thickness	0.20 mm	0.20 mm
Quality	High	High
Fill density	15%	90%
Nozzle temperature	215 °C	230 °C
Bed temperature	60 °C	50 °C
Material	PLA	FLEX CONDUCTIVE

2.5 System control architecture

The gamified rehabilitation system incorporated in the RehaBEElitation rehabilitation game, developed by an interdisciplinary team using Unity 3D.³⁴ In the game, the user is required to control the movements of a bee within a 3D environment. The tasks have been specifically designed to elicit movements that are commonly assessed in individuals with PD using the Unified Parkinson's Disease Rating Scale (UPDRS). The targeted movements include hand opening and closing, flexion and extension, adduction and abduction, forearm supination and pronation, and finger tapping. Interaction with the game is facilitated through a display screen, which presents the virtual environment to the user (Figure 3). Moreover, three more games – Minecraft, Geometry Dash, and Vampire Survivors – were used to test the system.

Figure 3.

Stages of the game: (a) first – pollination of flowers; (b) second – feeding of larvae; (c) third – nectar collection; (D) fourth – nectar drying.³⁴

The system palm module integrates the touch sensors and 9-DOF (Degrees of Freedom) Inertial Measurement Unit (IMU) sensors. Conductive thread was sewn onto the textile surface to serve as a touch sensor electrode. These sensors are interfaced with the forearm module, which houses the central microcontroller (ESP8266), with build-in Wi-Fi communication interface. The forearm module is responsible for initial data processing and communication with the host computer. The electronic components are powered by a 3.7 V, 2000 mAh lithium-polymer (Li-Po) battery, also housed in the forearm section. A charging circuit and voltage regulation module (TP4056) are integrated into the PCB. A PCB was designed to hosts the microcontroller, voltage regulator, charging circuitry, and connector interfaces (Figure 4). Flat ribbon cables with modular connectors are used to interconnect the IMUs with the main board, providing both signal lines and power in a compact form factor.

Figure 4.

PCB design of the forearm module.

The ESP8266 firmware (Arduino C++) continuously fuzes 9-DOF IMU data with a Kalman filter and streams orientation quaternions to the host computer for the hand motion interpretation. Quaternions were selected because they encode 3-D rotation without gimbal-lock and are computationally efficient for real-time processing. A companion Python 3 application (PySerial, NumPy) receives the quaternion stream via Wi-Fi, classifies each motion and triggers the corresponding keyboard or mouse events through the pynput library. The designed software supports user-selectable profiles: each profile stores individual calibration offsets, gesture thresholds and key-binding tables (JSON format). Thus, a switching between participants, different games or difficulty levels is available. Additional details regarding the electrical circuitry and control software are provided in the Supplementary Materials.

3 Results

3.1 Design results

First, a set of design alternatives was developed, resulting in three different framework concepts, which are presented in Figure 5.

Figure 5.

Variant design results: (a) first variant: 1 – hand-forearm joint, 2 – hooks, 3 – forearm component; (b) second variant: 1 – hand-forearm joint, 2 – forearm component, 3 – hand component, 4 – finger hooks, 5 – thumb hooks; (c) third variant: 1 – hand-forearm joint, 2 – hand component, 3 – forearm component, 4 – pin, 5 – holes for fastening strap, 6 – finger hooks, 7 – thumb hooks.

In the first concept (Figure 5(a)), the structure was created from a flexible material to avoid restricting hand movements and to ensure a close fit to the patient's hand, while also being easily adjustable for different users through straps or ribbons. However, this design had significant drawbacks: the flexible material made it difficult to securely attach the hooks, leading to potential deformation or breakage, and the hooks would not remain in a stable position. Moreover, the shape of the forearm component would only fit one hand (either left or right), limiting the universality of the design. Finally, attaching electronic components would be extremely difficult due to the complex and irregular shape of the forearm structure.

Considering these shortcomings, a second design iteration was developed (Figure 5(b)), with a primary focus on improving the hand-forearm joint. This version was intended to be constructed from rigid materials, presenting challenges in maintaining hand mobility and achieving a good fit. To address the issue of limited wrist movement, a flexible plate was integrated, inserted into the forearm component from one side and the hand component from the other. This solution allowed for wrist flexion and extension. However, movements such as ulnar/radial deviation and supination/pronation remained restricted. As shown in the figure, hooks for strength training were included: four rows of hooks for the fingers were attached to the hand component, and thumb hooks were added to both sides of the forearm component. This configuration allowed the device to be adapted for both left and right hands. Unlike the first concept, the hooks were attached to a stable structure. Wearing this structure like a bracelet eliminated the need for additional straps or bands.

In the third design iteration (Figure 5(c)), improvements were made to enhance wrist joint mobility. While the connection between the flexible plate and the hand component remained unchanged, the connection to the forearm component was upgraded with the addition of a pin, enabling additional movements—ulnar/radial deviation and supination/pronation—that were not possible in the previous concept. Furthermore, the area of the forearm component covering the hand was reduced. Compared to the first concept, where flexible materials made fitting easier, this adjustment was necessary because the third design used rigid, non-deformable materials. With the reduced coverage compared to the second concept, and the inclusion of fastening straps (visible as holes in Figure 5), this design could accommodate a wider range of hand sizes. The hooks for strength training were similarly arranged: four rows of hooks for the fingers were attached to the hand part, and thumb hooks were located on both sides of the forearm component, maintaining adaptability for both hands.

Although the third iteration was the most promising, some issues remained. After 3D printing and testing on a real hand, it was found that the hooks were too small and the spacing between them insufficient. The flexible plate connecting the hand and forearm components was also slightly too short, making it unsuitable for individuals with longer hands. Additionally, despite reducing the forearm component's width, there was still excess material, limiting its adaptability to different hand sizes. It was also desired that the forearm component gradually widens from the wrist toward the elbow to better match the natural shape of the arm. Due to these shortcomings, further design improvements were necessary to refine the hand device, meet the established criteria, and eliminate the identified issues.

In the second stage, the design of the device was further improved, and the appropriate geometry was created by calculating anthropometrics. After performing anthropometric calculations, the hand part was designed slightly shorter than the calculated value, i.e., 70 mm, in order to avoid restricting finger joint mobility and to ensure greater comfort and adaptability for different users. Also, according to the average values of the maximum and minimum lengths of the hand segments presented in the literature, the lengths of the hand and fingers were selected: the length of the hand is 189 mm, and the length of the fingers is 104 mm (Table 3).

Table 3.

Lengths of the hand and its segments.

Segment	Min length (at 150 cm) (mm)	Max length (at 200 cm) (mm)	Selected length (mm)
Hand	162.0	216.0	189.0
Palm	72.9	97.2	70.0
Fingers	89.1	118.8	104.0

The width of the palm part (Figure 6(a)) is chosen to be similar to the length, because the palm part will need space for the hooks of the elements that cause mechanical resistance and the box (Figure 6(b)) that houses the IMU sensor.

Figure 6.

Palm geometry: (a) palm dimensions in millimetres; (b) palm with hooks and space for box.

The maximum forearm components that can be manufactured are 290 and 217.5 mm long for people 200 cm and 150 cm tall, respectively, as these were the calculated forearm lengths (Table 4).

Table 4.

Forearm lengths.

Segment	Calculated average radius of the forearm (mm)	Hook attachment area radius (mm)	Radius of the distal end of the forearm bracelet (mm)
Forearm	42	30	40

The maximum dimensions of the electronic component box, which is placed on the forearm, are 60 × 40 × 20 mm. The length required for attaching the hooks is 30 mm. Thus, to implement all the necessary conditions, the length of this component is chosen, which is equal to 100 mm (Figure 7). This length was chosen so that the structure would be as small and lightweight as possible.

Figure 7.

Forearm part lengths.

Typically, the circumference of a man's forearm is between 251 and 305 mm. Meanwhile, for women, it is between 236 and 267 mm.⁴¹ The average circumference is then 265 mm. Since the component being manufactured does not cover the entire forearm, but only part of it, the circumference was approximated as circular and converted into an average radius for design calculations. For this purpose, the cross-section of the forearm is simplified and assumed to be circular. Thus, the average radius for the design of the forearm part should be about 42 mm. The geometry of this component will consist of two parts. The first is the hook attachment area (Figure 8, area 1). The hooks can be attached to either side. Its shape is approximately one-third of a circle, to suit people of the most diverse build, but also to support the hooks intended for mechanical tension. The radius of this part is smaller – 30 mm, because the circumference of the forearm at the wrist is much smaller.

Figure 8.

Forearm part geometry: 1 – hook attachment area, 2 – forearm strap.

The second part is the forearm bracelet (Figure 8, area 2). However, considering that the length of the forearm part is only 100 mm and does not reach the thickest part of the forearm, a shape that widens from the hook attachment area has been chosen for the forearm bracelet, the radius of which at the distal end will be 40 mm (Table 5). In addition, the second part is approximately a quarter circle in shape, smaller in circumference than the beginning of the part.

Table 5.

Radius of the forearm parts.

Segment	Min length (at 150 cm) (mm)	Max length (at 200 cm) (mm)	Selected length (mm)
Forearm	290.0	217.5	100.0

Later, a hand-forearm joint was designed. It consists of two interconnected plates (Figure 9. 1, 2). They are connected to each other by a specially designed pin (Figure 9. 3). This joint design allows for all necessary movements. In addition, it also firmly maintains the integral frame structure and allows for easy adjustment of plates of different lengths.

Figure 9.

Hand-forearm joint: 1 and 2 – flexible plates, 3 – pin.

Also, at the hook attachment points on both sides of the forearm, symmetrical grooves are made for hook attachment — plates with hooks for the thumb (Figure 10). The required part is placed on only one required side at a time, depending on which hand is being rehabilitated.

Figure 10.

Attaching the thumb resistance bands hooks: (a) forearm part; (b) joining the parts.

The glove was made approximately according to the hand, palm, and finger lengths obtained above (Figure 11).

Figure 11.

Glove and its components: (a) glove hooks; (b) resistance rubber bands.

A hand device for a gamified rehabilitation system for Parkinson's patients has been designed. During virtual reality rehabilitation sessions, patients will wear this device, which consists of a palm part (Figure 12.1), a forearm part (2), and two flexible connecting plates (3) that enable unrestricted wrist mobility. The forearm component is essential, as it houses most of the electronic components, thereby minimizing the load on the hand and reducing interference with natural movement. The flexible plates are interconnected using a pin joint (4), ensuring stability while maintaining flexibility. For strength training purposes, hooks (5) are mounted on the palm part, designed for attachment to rubber bands (6) used in therapeutic exercises. The palm part is secured to the patient's hand with Velcro fastening straps (7), while thumb hooks (8) are affixed to the forearm part to accommodate the unique anthropometry of the thumb. The forearm part is likewise secured to the patient's arm using adjustable Velcro straps (9). A box with a lid will be attached to the forearm component to house the electronics components (10). The palm electronics box with a lid (11) will contain an IMU sensor. The hand unit will be placed on the hand with a fabric glove (12). A special sheath (13) is used to neatly connect and isolate the wires.

Figure 12.

Device parts: 1 – palm part, 2 – forearm part, 3 – flexible plates, 4 – pin, 5 – palm hooks, 6 – glove hooks, 7 – palm fastening straps, 8 – forearm hooks, 9 – forearm fastening strips, 10 – forearm electronics box, 11 – palm electronics box, 12 – glove, 13 – insulating sheath with wires inside.

Finally, the 3D printing of the device parts and their connection into a common system was performed. The printing time for all components except flexible plates was 6 h 14 min, and the amount of material used was 75.85 g (25.43 m). The printing time for one flexible plate was 18 min, and the amount of material used was 3g (1.23 m). The fabricated prototype device is shown in Figure 13.

Figure 13.

Device prototype.

3.2 Structural simulation results

To assess whether the designed device can withstand the expected mechanical loads, simulation tests were performed using SolidWorks software. The loads for these tests were taken from scientific literature.^42,43 A summary of the structural simulation results is given in Table 6.

Table 6.
Summary of structural simulation results.

Component Applied force (N) Maximum stress (MPa) Material Yield strength (MPa) Plastic deformation

Palm hook (tension) 58.6 6.80 PLA 60 No

Forearm hook (tension) 108.6 21.29 PLA 60 No

Hook slider (tension, forearm) 108.6 19.02 PLA 60 No

Hook slider (pull-out, forearm) 108.6 43.23 PLA 60 No

Flexible plate (tension, pin from palm) 100 20.68 Flex. TPU 25 No

Palm part (tension, flexible plate) 100 2.65 PLA 60 No

Forearm part (tension, flexible plate) 100 4.70 PLA 60 No

Pin (shear force simulation) 50/−50 25.60 PLA 60 No

Component	Applied force (N)	Maximum stress (MPa)	Material	Yield strength (MPa)	Plastic deformation
Palm hook (tension)	58.6	6.80	PLA	60	No
Forearm hook (tension)	108.6	21.29	PLA	60	No
Hook slider (tension, forearm)	108.6	19.02	PLA	60	No
Hook slider (pull-out, forearm)	108.6	43.23	PLA	60	No
Flexible plate (tension, pin from palm)	100	20.68	Flex. TPU	25	No
Palm part (tension, flexible plate)	100	2.65	PLA	60	No
Forearm part (tension, flexible plate)	100	4.70	PLA	60	No
Pin (shear force simulation)	50/−50	25.60	PLA	60	No

The first simulation focused on the tensile strength of the hooks. Both the palm and forearm hooks were subjected to forces generated by resistance bands attached to them. The forearm hook was tested under a tensile load of 108.6 N, yielding a maximum stress of 21.29 MPa. This value remains within the safe working limits for PLA. When the hook slider on the forearm was subjected to a tensile force of 108.6 N, the highest stress observed was 19.02 MPa. The simulation for the pull-out force on the forearm hook slider applied a load of 108.6 N, with the maximum resulting stress reaching 43.23 MPa. A tensile force of 100 N applied to the flexible plate, manufactured from flexible TPU plastic, resulted in a maximum stress of 20.68 MPa. Simulation of the palm component, when tensioned through the flexible plate with a 100 N force, revealed a maximum von Mises stress of 2.65 MPa. Applying a tensile force of 100 N to the forearm component via the flexible plate resulted in a maximum stress of 4.70 MPa. A shear force of ±50 N was applied to the pin, resulting in a maximum von Mises stress of 25.60 MPa. Overall, the simulation confirms components can withstand the applied force without plastic deformation (Figure 14).

Figure 14.

Simulation results: (a) palm hook; (b) forearm hook; (c) hook slider; (d) flexible plate; (e) palm part; (f) forearm part; (g) pin.

3.3 System Performance and Limitations

The prototype glove was evaluated through preliminary functional no medical intervention testing with four healthy student volunteers. During testing, all major wrist movements were performed and hand positioning was defined using a 3D Cartesian coordinate system from IMU-derived quaternions, expressed as pitch (flexion/extension), roll (radial/ulnar deviation), and yaw (pronation/supination). The goal was to assess how accurately the system could capture these movements and link them to game control actions. Movements were selected based on the commands required for game interaction, and later, game profiles were created for each gesture's activation thresholds.

Two observation channels were used simultaneously: a Python program displaying filtered and processed data (see Figure 15), and the Arduino Serial Monitor, which showed raw, unprocessed angular IMU data directly from the ESP8266.

Figure 15.

Python program displaying filtered and processed data.

Experiments showed that angular changes smaller than 5° were too small for our system and were usually ignored – such minor movements are often filtered out. However, larger tilts were detected accurately and reliably. It was observed that, due to the limited range of wrist motion, unintended angular components would sometimes occur during a movement. For example, when rotating the hand sideways, the hand would slightly tilt downward, resulting in two different movements being recorded simultaneously. This issue was resolved by setting appropriate threshold limits and incorporating dead-zone functions. The user slowly performs the movement in all three planes, and the angular data (pitch, roll, yaw) is monitored on a computer.

In this testing phase, particular emphasis was placed on evaluating the accuracy of pitch and roll motion interpretation and their linkage to in-game actions. Angular changes between 10° and 15° were found to be optimal for precise control, with smaller fluctuations filtered out to prevent unintended actions, resulting in smooth and intuitive user experiences. For the RehaBEElitation serious game, the system enabled successful completion of the first level through wrist-based control, demonstrating its flexibility, even when interfaced via a modified Python-based emulator.

More dynamic scenarios, such as in Vampire Survivors game, highlighted the system's limitations, with occasional delays in response to rapid gestures, indicating a need for further optimization of gesture thresholds for fast-paced applications. Additionally, the pinch gesture, defined by thumb and index finger contact, was used as an activation input and achieved a detection success rate of approximately 90%, with most failures occurring during simultaneous angular movements. Importantly, the contact sensor remained reliable and comfortable over extended use, with no false activations or degradation in sensitivity observed. Overall, the system provided accurate and user-friendly control in 2D and 3D gaming scenarios.

One of the fundamental requirements for the system was the ability to respond to user movements in real time. To assess this, response time measurements and data transmission rate analysis were conducted. Users performed rapid gestures (e.g., sudden increases in pitch), while reaction times were determined by recording the display at 120 frames per second and analyzing the footage frame-by-frame. The results demonstrated response times of approximately 70–90 ms for pitch actions in Minecraft, 80–100 ms for pinch gestures triggering jumps in Geometry Dash, and about 100 ms for yaw-based actions in Vampire Survivors. It was observed that, in some cases, gesture filtering required a longer action duration for recognition, particularly in fast-paced scenarios. Additionally, the Python GUI consistently showed a stable data stream of around 50 Hz, with no dropped or delayed packets recorded during testing. The main results and limitations of the system are summarized in Table 7.

Table 7.

Summary of main results and system limitations.

Tested parameter	Description	Results and limitations
Wrist movements detection	Flexion/extension, radial/ulnar deviation (roll), pronation/supination (yaw)	Angular changes <5° ignored; 10–15° optimal for control
Motion-to-action mapping	Pitch/roll/yaw mapped to in-game actions (e.g., movement, direction)	Gestures successfully linked to game controls; intuitive user experience in 2D and 3D games
Gesture thresholds	Minimum angle for detection and action activation	10–15° threshold prevents false triggers;
Response time (ms)	Time from movement to in-game action	RehaBEElitation 60–80 ms, Minecraft: 70–90 ms, Geometry Dash: 80–100 ms, Vampire Survivors: 100 ms
Dynamic scenario performance	Testing in fast-paced game Vampire Survivors	Occasional delays in rapid gestures
Pinch gesture detection	Contact between thumb and index finger as an activation gesture	The detection success rate is approximately 90%, with failures primarily occurring during simultaneous angular movement. No false activations or sensor degradation observed
Sensor reliability and comfort	Performance and comfort during extended use	Sensors remained accurate and comfortable during >1 h of use; no decrease in sensitivity or false positives
Data transmission rate	Stability and frequency of data sent to host device	Stable 50 Hz data stream; no dropped or delayed packets observed

4 Discussion

Using iterative design, we optimized the device's functionality and usability. The third proposed variant offers the best combination of mechanical functionality, adaptability, and wearability. Unlike the previous designs, this iteration allowed for a full range of wrist movements—including ulnar/radial deviation and supination/pronation—thanks to its improved joint mechanism and pin connection. The refined design also reduced unnecessary material, making it more suitable for a wider range of hand sizes and increasing user comfort without sacrificing device stability.

One of the critical design goals in the development of the prototype was the reduction of device weight, as excessive mass—particularly when concentrated distally—can negatively affect comfort, endurance, and usability during rehabilitation exercises. The prototype has a total weight of 227.6 g, of which 60.3 g is attributed to the palm–hand segment, while the remaining 167.3 g is located on the forearm (including a 40 g battery). When compared to other available solutions, such as SaeboGlove (∼450 g), Hand of Hope (700–800 g), and RAPAEL (132 g at the wrist), the proposed device demonstrates a comparatively low mass at the hand level. The RehaBEElitation Serious Game device reports a palm–hand unit weight of approximately 60 g without a battery. In that system, the battery is mounted on the palm–hand segment, which increases distal loading. A key advantage of the proposed system is therefore not only the absolute reduction of mass but also its intentional distribution. By shifting the majority of the device weight to the forearm rather than the wrist or hand, the prototype minimizes distal loading, which is particularly important for individuals with impaired motor function.

3D printing played a crucial role in the development of the rehabilitation glove prototype. The use of additive manufacturing allowed for rapid prototyping and iteration of complex, customized structures based on anthropometric data. Materials such as PLA and flexible TPU were selected for their strength, flexibility, and biocompatibility, making them ideal for wearable medical devices. 3D printing enabled the creation of lightweight, modular components like the forearm frame, palm supports, and sensor housings, which were essential for ensuring both functionality and user comfort. Additionally, this manufacturing technique significantly reduced production costs and lead times compared to traditional methods, facilitating quicker transitions from design to physical testing. The ability to customize designs easily also supports future scalability and personalization for individual users, enhancing the device's adaptability and therapeutic effectiveness.

The proposed design integrates strength training elements, modular adjustability, and wireless communication in a lightweight, low-cost structure. Compared to commercial analogues, it prioritizes user comfort, universal fit, and home-based usability, addressing common limitations such as lack of adaptability, restricted mobility, and operational complexity. The proposed system's gamified and wireless features encourage patient engagement and make it more accessible for independent, home-based therapy.

While the design shows considerable promise, several limitations remain. The prototype has not yet undergone comprehensive testing with its target population—individuals with PD—which limits conclusions about clinical effectiveness. Additional concerns include potential material durability issues under prolonged use and the need for further optimization of the hook and resistance band sizes to accommodate all users. Moreover, all simulations confirmed that the maximum stress experienced by each component did not exceed the yield strength of the respective materials, indicating no plastic deformation is expected under the tested loading conditions.

The summarized results in Table 7 demonstrate that the proposed system effectively captures and translates key hand movements into reliable in-game actions, with optimal detection thresholds established for pitch and roll angles. Real-time response times ranging from 70 to 100 ms across various game scenarios indicate that the system is suitable for interactive applications, while the pinch gesture showed high detection reliability with minimal false activations. Nevertheless, performance in fast-paced environments revealed limitations, including occasional delays in rapid gesture recognition and decreased accuracy during simultaneous gestures. Overall, these findings validate the usability and technical feasibility of the glove-based interface, while also highlighting areas for future improvement to enhance responsiveness and robustness in dynamic use cases.

The modular, wireless, and gamified nature of the glove suggests broader applications in rehabilitation for other neurological or musculoskeletal disorders, and its adaptability and low cost may support its use in remote or resource-limited settings.

5 Conclusions

This study introduced a lightweight, wireless rehabilitation glove integrating motion capture, resistance training, and gamified therapy. Experimental testing confirmed accurate detection of major wrist movements with 70–100 ms response times, 90% pinch gesture reliability, and stable 50 Hz data transmission. Structural simulations showed that all components withstand expected loads without plastic deformation. With a total weight of 227.6 g—only 60.3 g on the hand—the glove reduces distal loading compared to existing devices, improving comfort and usability. Future work should include testing with larger cohorts to further assess system accuracy, usability, and clinical applicability across different rehabilitation contexts, including but not limited to PD.

The design of the rehabilitation glove prioritizes ergonomic fit, lightweight construction, and modular integration to accommodate users with reduced muscular strength. The structure is divided into two main components: the palm and the forearm. The palm component is designed to be as compact and unobtrusive as possible, housing only essential sensors to minimize mass and preserve hand mobility. In contrast, the forearm component is engineered to support the heavier electronic elements, including the battery and main control unit, thereby shifting the load away from the hand and improving overall wearability. Dimensions and geometry were guided by average anthropometric data to ensure a secure and comfortable fit, while internal mounting features and protective enclosures were incorporated to safely house the electronic components without interfering with natural joint movement. This mechanical-electronic integration provides a stable, user-friendly platform suitable for rehabilitation tasks requiring precision, repeatability, and extended use.

Supplemental Material

sj-7z-1-thc-10.1177_09287329261438652 - Supplemental material for Engineering and experimental evaluation of a smart wireless glove for gamified upper limb rehabilitation in Parkinson's disease

Supplemental material, sj-7z-1-thc-10.1177_09287329261438652 for Engineering and experimental evaluation of a smart wireless glove for gamified upper limb rehabilitation in Parkinson's disease by Kotryna Vaišnoraitė, Justinas Višinskas, Adriano de O Andrade, Luanne C Mendes, Camille M Alves, Kristina Daunoravičienė, Vytautas Abromavičius and Jurgita Žižienė in Technology and Health Care

Footnotes

Aknowlegments

The present work was carried out with the support of the National Council for Scientific and Technological Development (CNPq - 442150/2023-7, 131534/2024- 6 and 405365/2023-3), Coordination for the Improvement of Higher Education Personnel (CAPES - Program CAPES/COFECUB 88881.370894/2019-01), and the Foundation for Research Support of the State of Minas Gerais (FAPEMIG). A.O.A. (302942/2022-0) is a fellow of CNPq, Brazil.

Ethical considerations

This study involved self-experimentation conducted exclusively by the authors. All procedures were noninvasive, posed minimal risk, and were performed in accordance with the principles of the Declaration of Helsinki. No ethical approval was required.

Consent to participate

Not applicable

Consent for publication

Not applicable

Author contribution statement

KV, JV: Conceptualization, Methodology, Investigation, Visualization, Writing – original draft. AdOA, LCM, CMA: Conceptualization, Writing – review & editing. KD, VA, JZ: Conceptualization, Data curation, Supervision, Project administration, Resources, Writing – review & editing.

Funding statement

The authors received no financial support for the research, authorship, and/or publication of this article.

Declaration of conflicting interest

The authors declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.

Data availability statement

The data supporting the findings of this study are available from the corresponding author upon reasonable request.

Supplemental material

Supplemental material for this article is available online.

ORCID iDs

Kotryna Vaišnoraitė

Adriano de O Andrade

Camille M Alves

Kristina Daunoravičienė

Vytautas Abromavičius

Jurgita Žižienė

References

Frazzitta

Maestri

Bertotti

, et al. Intensive rehabilitation treatment in early Parkinson’s disease: a randomized pilot study with a 2-year follow-up. Neurorehabil Neural Repair 2015; 29: 123–131.

Fan

Sun

, et al. Early intensive rehabilitation for patients with traumatic brain injury: a prospective pilot trial. World Neurosurg 2020; 137: e183–e188.

Laver

Lange

George

, et al. Virtual reality for stroke rehabilitation. Cochrane Database Syst Rev 2017; 11: CD008349.

Iosa

Aydin

Candelise

, et al. The michelangelo effect: art improves the performance in a virtual reality task developed for upper limb neurorehabilitation. Front Psychol 2021; 11: 611956.

Rojo

Santos-Paz

JÁ

Sánchez-Picot

, et al. Farmday: a gamified virtual reality neurorehabilitation application for upper limb based on activities of daily living. Appl Sci 2022; 12: 7068.

Lülsdorff

Junker

Studer

, et al. Neurorehabilitation of the upper extremity – immersive virtual reality vs. Electromechanically assisted training. A comparative study. Front Neurol 2023; 14: 1290637.

Ali

Kumaran

Unni

, et al. Effectiveness of an intensive, functional, and gamified rehabilitation program on upper limb function in people with stroke (EnteRtain): a multicenter randomized clinical trial. Neurorehabil Neural Repair 2024; 38: 243–256.

Pillai

Sunny

MSH

Shahria

, et al. Gamification of upper limb rehabilitation in mixed-reality environment. Appl Sci 2022; 12: 12260.

Levin

Demers

. Motor learning in neurological rehabilitation. Disabil Rehabil 2021; 43: 3445–3453.

10.

Huang

Chiang

Yeh

, et al. Effects of virtual reality-based motor control training on inflammation, oxidative stress, neuroplasticity and upper limb motor function in patients with chronic stroke: a randomized controlled trial. BMC Neurol 2022; 22: 1–14.

11.

Vaezipour

Aldridge

Koenig

, et al. ‘It’s really exciting to think where it could go’: a mixed-method investigation of clinician acceptance, barriers and enablers of virtual reality technology in communication rehabilitation. Disabil Rehabil 2022; 44: 3946–3958.

12.

Gherman

Zima

Vaida

, et al. Robotic systems for hand rehabilitation—past, present and future. Technologies 2025; 13: 37.

13.

Weerasoory

WAKC

Kumasaru

URE

Nipun

HMW

, et al. A robotic hand for rehabilitation of wrist and fingers. In: 2023 8th International Conference on Information Technology Research (ICITR). IEEE, pp. 1–6.

14.

Ferreira

Menezes

. Gamifying motor rehabilitation therapies: challenges and opportunities of immersive technologies. Information 2020; 11: 88.

15.

Jackson

Abdullah

. Development and testing of a soft exoskeleton robotic hand training device. Sensors 2023; 23: 8395.

16.

Mendes

Marques

Alves

, et al. Multidimensional assessment of individuals with Parkinson’s disease: development and structure validation of a self-assessment questionnaire. Healthcare 2022; 10: 1823.

17.

Aitken

Samotus

Naidu

, et al. Force control issues in upper and lower limbs in Parkinson’s disease and freezing of gait. IEEE Trans Neural Syst Rehabil Eng 2024; 32: 577–586.

18.

Cohen

Kizony

. Patient perspectives on upper-limb daily function in Parkinson’s disease. Mov Disord Clin Pract 2025; 12: 196–202.

19.

Poewe

Seppi

Tanner

, et al. Parkinson disease. Nat Rev Dis Primers 2017; 3: 1–21.

20.

Feingold-Polak

Barzel

Levy-Tzedek

. A robot goes to rehab: a novel gamified system for long-term stroke rehabilitation using a socially assistive robot—methodology and usability testing. J Neuroeng Rehabil 2021; 18: 1–18.

21.

White

Vining Radomski

Finkelstein

, et al. Assistive/socially assistive robotic platform for therapy and recovery: patient perspectives. Int J Telemed Appl 2013; 2013: 948087.

22.

Proulx

Beaulac

David

, et al. Review of the effects of soft robotic gloves for activity-based rehabilitation in individuals with reduced hand function and manual dexterity following a neurological event. J Rehabil Assist Technol Eng 2020; 7: 205566832091813. https://doi.org/101177/2055668320918130

23.

Xie

Meng

, et al. Design of a SMA-based soft composite structure for wearable rehabilitation gloves. Front Neurorobot 2023; 17: 1047493.

24.

Ozioko

Dahiya

. Smart tactile gloves for haptic interaction, communication, and rehabilitation. Adv Intell Syst 2022; 4: 2100091.

25.

Chen

Gong

Wei

, et al. A wearable hand rehabilitation system with soft gloves. IEEE Trans Industr Inform 2021; 17: 943–952.

26.

Wang

Kang

Kristensson

. Supporting Playful Rehabilitation in the Home using Virtual Reality Headsets and Force Feedback Gloves. Proceedings - 2022 IEEE Conference on Virtual Reality and 3D User Interfaces, VR 2022 2022; 504–513.

27.

Qiaoling

Shujing

Hexiang

, et al. Hand rehabilitation exoskeleton control system with multi-feeling synergistic stimulation. CN117426948A, China, https://worldwide.espacenet.com/patent/search/family/089557550/publication/CN117426948A?q=pn%3DCN117426948A (accessed 6 March 2025).

28.

Xiaoyu

Min

Xiaofeng

, et al. Hand function training and assessment rehabilitation glove system based on motor imagery. CN115364327A, China, https://worldwide.espacenet.com/patent/search/family/084065774/publication/CN115364327A?q=pn%3DCN115364327A (accessed 6 March 2025).

29.

Syrebo. Hand Rehabilitation Robot SY-HR06E for Home Use – SIYI INTELLIGENCE, https://www.syrebo.com/rehabilitation-system-for-clinic/hand-rehabilitation-robot-sy-hr06e.html (accessed 6 March 2025).

30.

Neofect: Digital Solutions for Occupational Therapists, https://www.neofect.eu/ (accessed 6 March 2025).

31.

Rehab-Robotics. Hand of Hope, https://www.rehab-robotics.com.hk/hoh/hoh_basic.html (accessed 6 March 2025).

32.

Mendes

Cabral

Alves

, et al. Objective Evaluation of Bradykinesia Using the RehaBEElitation Serious Game. 2022 IEEE Workshop on Complexity in Engineering, COMPENG 2022. Epub ahead of print 2022. DOI: 10.1109/COMPENG50184.2022.9905437.

33.

Saebo. Hand Therapy Rehabilitation Glove for Stroke | SaeboGlove, https://www.saebo.com/products/saeboglove (accessed 6 March 2025).

34.

Mendes

de Sá

AAR

Marques

, et al. RehaBEElitation: the architecture and organization of a serious game to evaluate motor signs in Parkinson’s disease. PeerJ Comput Sci 2023; 9: 1–30.

35.

Rudolfs

Renato

Maurice

. Body segment parameters: a survey of measurement techniques. O&P virtual library. Artif Limbs 1964; 8: 44–66.

36.

Sayadizadeh

Daliri

Rahimi

, et al. Grip and pinch strength prediction models based on hand anthropometric parameters: an analytic cross-sectional study. BMC Musculoskelet Disord 2024; 25: 809.

37.

Hsueh

Lai

Wang

, et al. Effect of printing parameters on the thermal and mechanical properties of 3D-printed PLA and PETG, using fused deposition modeling. Polymers (Basel) 2021; 13: 1758.

38.

Ramírez-Revilla

Camacho-Valencia

Gonzales-Condori

, et al. Evaluation and comparison of the degradability and compressive and tensile properties of 3D printing polymeric materials: PLA, PETG, PC, and ASA. MRS Commun 2023; 13: 55–62.

39.

3D plastics. PLA plastic, https://www.3dplastic.eu/produktas/pla-gf-2/ (2020, accessed 7 April 2025).

40.

3D plastics. Flex conductive plastic, https://www.3dplastic.eu/produktas/flex-conductive/ (2020, accessed 7 April 2025).

41.

Kilani

Abu-Eisheh

. Optimum anthropometric criteria for ideal body composition related fitness. Sultan Qaboos Univ Med J 2010; 10: 74.

42.

János Bretz

Jobbágy

Bretz

. Force measurement of hand and fingers. Biomechanica Hungarica III évfolyam.

43.

Eschweiler

Praster

Quack

, et al. Musculoskeletal modeling of the wrist via a multi body simulation. Life 2022; 12: 581.

Supplementary Material

Please find the following supplemental material available below.

For Open Access articles published under a Creative Commons License, all supplemental material carries the same license as the article it is associated with.

For non-Open Access articles published, all supplemental material carries a non-exclusive license, and permission requests for re-use of supplemental material or any part of supplemental material shall be sent directly to the copyright owner as specified in the copyright notice associated with the article.

0.00 MB

0.02 MB