Advanced VR Calibration for Upper Limb Rehabilitation: Making

Immersive Environments Accessible

Vanesa Herrera

1 a

, Ana Reyes-Guzm

2 b

, David Vallejo

1 c

, Jos

e Castro-Schez

1 d

Dorothy N. Monekosso

3 e

, Gonz

alez-Morcillo Carlos

1 f

and Javier Albusac

1 g

School of Computer Science, University of Castilla-La Mancha, Ciudad Real, Spain

Hospital Nacional de Parapl

ejicos de Toledo, Toledo, Spain

Department of Computer Science, University of Durham, Durham, County Durham, GB, U.K.

Keywords:

Upper Limb Rehabilitation, Virtual Reality, Dynamic Environment Adaptation, Free-Hand Calibration,

Immersive Gamiﬁcation.

Abstract:

The creation of accessible spaces is essential for patients with motor injuries to conduct therapy safely and

effectively. Disruptive technologies such as Virtual Reality (VR) are increasingly being used as a complement

to traditional therapy, with excellent results. VR allows, among other things, the realistic recreation of physical

spaces, so much so that it is relatively easy to run the risk of transferring physical barriers into the virtual

space. This article proposes an innovative method of calibration in virtual environments that assesses the

motor limitations of patients with cervical spinal cord injuries, doing so individually for each upper limb. The

result is the dynamic adaptation of virtual environments to make them accessible and safe for rehabilitative

therapy practices. This method has been integrated into the Rehab-Immersive platform, which hosts a series

of serious games aimed at rehabilitating upper limbs, using immersive gamiﬁcation techniques.

1 INTRODUCTION

In recent years, the use of Virtual Reality (VR) has

spread to a variety of ﬁelds, including physical and

psychological rehabilitation (Zhang et al., 2020; Cha

et al., 2021). The application of VR in rehabilitation

process has shown important beneﬁts such as an im-

provement in mobility, balance and cognitive function

(Park et al., 2020), as well as reducing the percep-

tion of pain in patients (Goudman et al., 2022; Mal-

lari et al., 2019). Furthermore, the use of VR devices

in rehabilitation programs improves treatment adher-

ence and increases motivation (Dias et al., 2019), both

important qualities for a successful rehabilitation pro-

cess (Teo et al., 2022).

In particular, the technological advances that im-

https://orcid.org/0000-0002-6187-4794

https://orcid.org/0000-0003-2905-2405

https://orcid.org/0000-0002-6001-7192

https://orcid.org/0000-0002-0201-7653

https://orcid.org/0000-0001-7322-5911

https://orcid.org/0000-0002-8568-9542

https://orcid.org/0000-0003-1889-3065

mersive VR devices have undergone in recent years

have led to a new ﬁeld in upper limb rehabilitation.

Wearable devices, such as head-mounted displays

(HMDs), provide users with immersive experiences

in virtual environments through a viewer and con-

trollers. More recently, this type of system has begun

to offer hand tracking through the existing inside-out

cameras in the headset (Khundam and N

oel, 2021).

As a result, existing applications are being adapted or

new ones are being developed. These applications al-

low the user to interact with the virtual world using

his or her own hands, without the need for controllers

or joysticks. This new form of interaction with vir-

tual environments makes it more intuitive, natural and

simple. In line with this, recent studies demonstrate

the preference for this mechanism of interaction over

controllers (Juan et al., 2023). Moreover, it opens

new opportunities for people with reduced mobility in

their upper limbs who, until now, have been unable to

use these types of devices due to the need for ﬁne mo-

tor skills to hold the controller and press the buttons.

This technological progress has generated new possi-

bilities in the development of applications for upper

limb rehabilitation. In this context, it is essential for

378

Herrera, V., Reyes-Guzmán, A., Vallejo, D., Castro-Schez, J., Monekosso, D., Carlos, G. and Albusac, J.

Advanced VR Calibration for Upper Limb Rehabilitation: Making Immersive Environments Accessible.

DOI: 10.5220/0012624600003690

Paper published under CC license (CC BY-NC-ND 4.0)

In Proceedings of the 26th International Conference on Enterprise Information Systems (ICEIS 2024) - Volume 2, pages 378-389

ISBN: 978-989-758-692-7; ISSN: 2184-4992

patients to use their hands naturally, replicating tradi-

tional rehabilitation movements and grips.

Although hands-free interaction and accurate

hand tracking is an important advance in the acces-

sibility of immersive environments for people with

motor impairments, it is still not sufﬁcient for the

proper development of rehabilitation therapies. The

challenge of developing VR environments that are

accessible and adapted to the different needs of end

users and the speciﬁc characteristics of each applica-

tion. VR makes it possible to simulate worlds and

situations that are difﬁcult to reproduce in real envi-

ronments, which is a signiﬁcant advantage, especially

given the diversity of environments and the mobility

constraints of certain users. This capability highlights

the need for virtual environments to be accessible and

adaptable to both user needs and application-speciﬁc

requirements.

For users with special needs or disabilities, the

creation of adapted, accessible, and inclusive VR en-

vironments is crucial (Soomal et al., 2020; Creed

et al., 2023). This involves tailoring the type of in-

teraction and the placement of interactive elements

to suit each user’s abilities. This is particularly im-

portant for patients with cervical spinal cord injuries

(cSCI), who require exercises to be performed while

seated, considering their upper limb mobility and

gripping capabilities. Proper positioning of interac-

tive elements in the VR environment is key to ensur-

ing safety, prevent overexertion or unwanted compen-

satory movement. These compensatory movements,

which are movement patterns adopted to overcome

motor restrictions (Luo et al., 2023), while facilitating

task performance, can sometimes lead to additional

problems that negatively affect the rehabilitation pro-

cess (see Figure 1).

One solution to address these challenges lies in the

ability of VR systems themselves to capture and store

relevant data. This includes data related to the kine-

matics of the body parts undergoing therapy, as well

as the position of elements in the virtual scene and

historical tracking of the patient’s progress. In addi-

tion, the capture and storage of kinematic data is es-

sential in the rehabilitation process. This practice pro-

vides healthcare professionals with a more objective

and quantitative view of the patient’s progress. This is

exempliﬁed in the study by Onitsuka et al. (Onitsuka

et al., 2023), which highlights the signiﬁcance of such

data in monitoring and enhancing rehabilitation out-

comes. This information is useful for adjusting and

optimising rehabilitation programmes, ensuring their

maximum effectiveness for the speciﬁc needs of each

patient. Furthermore, the capture and analysis of this

data not only facilitates the identiﬁcation of the effec-

tiveness of the exercises but also allows for the detec-

tion of possible unwanted compensatory movements.

In this context, leveraging VR capabilities, a

practical solution for enhancing the accessibility and

adaptation of virtual rehabilitation environments in-

volves a calibration process before exercise execu-

tion. This calibration should focus on the speciﬁc mo-

tor conditions of the patient, taking into account their

particular characteristics and limitations. This assess-

ment allows for precise modiﬁcations to the position-

ing of the interactive elements within the virtual en-

vironment. Such an approach ensures that the virtual

environment is not only accessible but also highly per-

sonalized, adapting to both the physical capabilities of

the patient and their speciﬁc stage in the rehabilitation

process.

On this basis, the article presents an advanced au-

tomatic calibration system adapted to upper limb re-

habilitation in patients with cSCI. This system takes

into account the physical characteristics of the patient

and the speciﬁc requirements of the exercises to be

performed. Thanks to this pre-calibration, it is pos-

sible to adapt the position of the elements in the vir-

tual environment according to the speciﬁc character-

istics of the exercise. For this reason, the range of

movement is taken into account separately for each of

the upper limbs, since the exercises can be performed

with one or two hands. In addition to the optimal po-

sition, it is taken into account that the patients perform

the exercises in a seated position, thus avoiding possi-

ble undesirable compensatory movements. It should

be noted that the calibration is based on hand posi-

tions, without the need for external elements such as

joysticks or controllers. As a result, patients perform

the exercises in a more natural and ergonomic way,

potentially improving the effectiveness of rehabilita-

tion.

The calibration system, aimed at improving up-

per limb rehabilitation of cSCI patients, has been

integrated into the Rehab-Immersive platform (Her-

rera et al., 2023). This integration facilitates design,

implementation and testing. The Rehab-Immersive

platform is currently developing a series of serious

games. These games are adapted to speciﬁc thera-

pies for the rehabilitation of upper limbs of ICSc pa-

tients. The collaboration with the Hospital Nacional

de Parapl

ejicos de Toledo has been key in this de-

velopment. This hospital is recognised as a refer-

ence centre for patients with SCI. The main objective

of Rehab-Immersive is to create an environment that

favours effective rehabilitation

The rest of the paper is structured as follows. Sec-

tion 2 reviews related works. Section 3 provides con-

text and background information. Section 4 describes

Advanced VR Calibration for Upper Limb Rehabilitation: Making Immersive Environments Accessible

379

Figure 1: On the left, a patient performs compensatory movements, with trunk inclination and grasping the wheelchair with

the left hand to achieve balance. On the right, due to a correct distribution of virtual elements, compensatory movements are

not necessary.

the automatic calibration system. Section 5 presents

the evaluation and the results obtained. Section 6 dis-

cusses the limitations and future work. Finally, the

paper concludes with Section 7, which presents the

conclusions.

2 RELATED WORKS

2.1 VR Rehabilitation for Upper Limbs

Several studies have focused on the rehabilitation of

upper limbs using VR techniques. These works can

be classiﬁed according to the degree of immersion and

the main way of interaction. Depending on the degree

of immersion, there are non-immersive and immer-

sive applications and, the use of devices such as con-

trollers or the user’s own hands (hand tracking) can

be considered.

Non-immersive and hand tracking upper limb VR

rehabilitation studies include the work of Shahmoradi

et al. (Shahmoradi et al., 2021), who used the Kinect

sensor to rehabilitate the upper limb in stroke patients

through a set of games. The results of the study indi-

cated improvements in participants’ range of motion

in terms of horizontal shoulder abduction and adduc-

tion, elbow ﬂexion, and wrist supination and ﬂexion.

The study by Reyes-Guzm

an et al. (de Los Reyes-

Guzm

an et al., 2021) also uses non-immersive VR

and hand tracking by means of a Leap Motion Con-

troller (LMC), and focused on SCI patients. The con-

clusions of the study indicate that the LMC is suitable

for a given sample of cervical patients for rehabilita-

tion purposes.

Regarding studies with immersive VR and the use

of controllers, the following can be found the study

by Lim et al. (Lim et al., 2020) which investigated

the use of VR in combination with conventional re-

habilitation for upper limb rehabilitation in patients

with SCI. The conclusion of the study is that VR train-

ing for upper limb function after SCI may be an ac-

ceptable adjunctive rehabilitation method without sig-

niﬁcant adverse effects. With this degree of immer-

sion and type of interaction, the study by Everard et

al. developed an immersive VR version of the Box

and Block Test (BBT-VR) to assess manual dexterity

in stroke patients and healthy participants. The aim

was to evaluate the concurrent validity of the BBT-

VR, highlighting the need for regular follow-up and

kinematic assessments to objectively measure motor

recovery after stroke. The study by Lee et al. demon-

strates that a fully immersive VR rehabilitation pro-

gramme with an HMD and controllers is feasible for

upper limb rehabilitation in chronic stroke patients,

without serious adverse effects.

In the ﬁeld of upper limb rehabilitation using im-

mersive VR systems and direct interaction with the

patient’s hands, is the study by AlMousa et al. (Al-

Mousa et al., 2020). This study presents the develop-

ment of a game designed for stroke patients requiring

upper limb rehabilitation at home. The game uses the

Oculus Rift HMD and the LMC hand tracker for a

fully immersive virtual reality experience.

In the latter group, the combination of immer-

sive virtual reality (VR) and hand tracking has proven

to be most effective for upper limb rehabilitation.

ICEIS 2024 - 26th International Conference on Enterprise Information Systems

380

This approach provides a more natural interaction,

which is crucial for patients with motor limitations

(AlMousa et al., 2020; Juan et al., 2023). In ad-

dition, non-immersive systems can limit depth per-

ception due to their design for ﬂat monitors, which

negatively affects the user experience (Everard et al.,

2022).

2.2 Adaptation of VR Environments for

Rehabilitation

Adaptation in VR environments is essential to ensure

accessibility and effectiveness in rehabilitation. Sev-

eral studies have highlighted the accessibility chal-

lenges associated with these technologies, particu-

larly for people with physical disabilities. Mott et

al. (Mott et al., 2019) identify ﬁve key elements for

the accessibility of VR content: accessibility of inter-

action, accessibility of devices, inclusive representa-

tions and diversity of applications. In addition, in a

separate study conducted in 2020, (Mott et al., 2020)

identiﬁed seven speciﬁc barriers for people with re-

duced mobility, one of which related to the prepara-

tion of VR peripherals and the deﬁnition of VR game

boundaries.

The need to adapt virtual environments to the ca-

pabilities of rehabilitation patients is a topic widely

supported in the scientiﬁc literature. One example is

the study by Lagos Rodriguez et al. (Lagos Rodr

ıguez

et al., 2022) which highlights the importance of per-

sonalisation in VR environments, focusing on the spe-

ciﬁc needs of each patient. On the other hand, the

study by Carrington et al. (Pei et al., 2023), allows

wheelchair users to assess the accessibility of unfa-

miliar places remotely by exploring detailed digital

replicas. The study introduces a VR technique called

‘Embodied Exploration’, which allows wheelchair

users to explore high-ﬁdelity digital replicas of phys-

ical environments. However, a limitation of the study

is the need for users to provide accurate measure-

ments of their biometrics.

3 BACKGROUND

Focused on the need to create patient-centred virtual

environments for upper limb rehabilitation in patients

with cSCI, the Rehab-Immersive (Herrera et al., 2023)

project arises, coordinated with the Hospital Nacional

de Parapl

ejicos (Toledo, Spain). In this project, a set

of serious games is being developed to improve the

mobility and manipulative skills of patients. Serious

games in the context of Rehab-Immersive take advan-

tage of the beneﬁts of VR, such as motivation and re-

inforcement of the patient’s commitment, as well as

accurate monitoring and evaluation of the evolution

of the treatment by the specialist.

Through various analyses and tests conducted

with patients and specialists, a key element has been

identiﬁed in VR environments for the rehabilitation of

cSCI patients: the signiﬁcance of three-dimensional

(3D) positioning of elements within the space that pa-

tients interact with.

In order to solve this problem, a manual calibra-

tion of the elements with which the patient interacts

was implemented in two ways: with predeﬁned posi-

tions (closer and further away) in the three axes, and

through manual positioning of the object in 3D space.

However, moving barriers were found in the man-

ual calibration. First, depending on the patient’s mo-

bility, it can be a tedious process. This is the case

when the elements with which the patient must in-

teract are positioned too far apart, leading to a sub-

optimal interaction location. Additionally, complex-

ity increases when there are numerous elements that

require individual conﬁguration. Second, there are

patients who, given their conditions, cannot calibrate

the elements well due to grip difﬁculties. In the case

of predeﬁned positions, these may not be adequately

adapted to the optimum. Thirdly, this calibration has

to be done for each element with which interactive

element and for each serious set, which is a slow pro-

cess.

Particularly in patients with limited or no func-

tional grip, this approach presents a signiﬁcant prob-

lem, making it necessary for the specialist to intervene

by positioning the item. However, the specialist’s

view of the 3D space is not through the VR HMD, but

on a screen, which leads to a loss of depth sensation

and can lead to errors in calibration.

4 CALIBRATION SYSTEM

In response to the challenges mentioned above, the

calibration system proposed in this paper is tailored to

the speciﬁc needs of cSCI patients. This system is de-

signed to accommodate several factors to accomplish

effective rehabilitation. To achieve this, both the cali-

bration system and the BBT-VR have been developed

using Unity

and the Meta XR Interaction SDK

. First

of all, it takes into account the patient’s posture and

limited mobility, since the exercises are performed sit-

ting and without moving. This consideration includes

https://unity.com/

https://developer.oculus.com/documentation/unity/

unity-isdk-interaction-sdk-overview/

Advanced VR Calibration for Upper Limb Rehabilitation: Making Immersive Environments Accessible

381

(a) (b) (c)

Figure 2: Movements to be reproduced for the calibration of the right and left arm in the three planes: (a) transverse plane,

(b) sagittal plane, and (c) frontal plane.

the height of the items and the position of the patient’s

legs.

The system also provides individual calibration

for each upper limb, adjusting the virtual environment

to respect the limitations and capabilities of each side

of the body.

In addition, the system is applicable to different

games and interactive elements. It incorporates cali-

bration data to adapt the interactive elements depend-

ing on the type of exercise, whether monomanual or

bimanual, and whether it involves gripping or not.

A key feature of this system is its automation and

customization capabilities. It automates the calibra-

tion process based on patient mobility and capacity

data, facilitating accurate and efﬁcient customization

for each individual.

Thanks to this calibration, a safe VR environ-

ment is provided. Feature that helps prevent exces-

sive efforts and compensatory movements that could

be detrimental to the patient’s health.

Finally, the system improves ease of use, improv-

ing patient autonomy in the use of VR.

In summary, the proposed calibration system aims

to improve the patient’s experience and optimize re-

habilitation results, adjusting to the evolution of the

patient’s motor capacity and ensuring that the exer-

cises are performed safely and effectively.

Additionally, the calibration system will be tested

with one of the games implemented within Rehab-

Immersive, speciﬁcally the BBT-VR. The BBT, is a

manipulative dexterity test in which the patient must

move as many blocks as possible from one compart-

ment to another within a box, all within 60 seconds

(Mathiowetz et al., 1985).

The implemented calibration system is designed

to adapt the virtual workspace to ﬁt the motor skills

of each cSCI patient. This calibration process takes

into account three aspects: i) head position, ii) elbow

ﬂexion and, iii) shoulder movements.

The initial state of the HMD is captured to de-

termine the position and rotation of the head at rest.

On the other hand, elbow ﬂexion, with the hands hip-

width apart, is used to determine the central area to

the patient and the position of the hands in this pose,

which is used for some therapeutic exercises. How-

ever, it is the analysis of the shoulder movements

that plays a central role in this process, following the

principles established by I. A. Kapandji (Kapandji,

1971). In this book, Kapandji details the complex

movements of the shoulder in the sagittal, frontal and

transveral planes, which include ﬂexion and exten-

sion, abduction and adduction, and horizontal ﬂexion

and extension, respectively.

The objective of this calibration phase is to de-

termine the maximum reach that a user can achieve

with each hand in three planes. The kinematics of the

shoulder are critical in this process, as the maximum

reach is achieved with the arm and elbow fully ex-

tended. The use of ellipses, rather than circles, to de-

ﬁne the workspace is adopted because of the potential

mobility restrictions that patients may have. Ellipses

offer greater ﬂexibility to accommodate these limita-

tions. Since users are seated and cannot perform the

full circumduction movement, the workspace is ide-

ally limited to the front and side areas, rather than ex-

tending behind the patient. In an optimal scenario,

these ellipses will approximate to circles. Through

various tests, it has been proven that using ellipses to

ICEIS 2024 - 26th International Conference on Enterprise Information Systems

382

Figure 3: An example of sagittal, frontal, and transverse ellipses resulting from the calibration process. Images taken from

different perspectives and decorative elements have been removed for clearer visualization.

deﬁne the workspace results in greater accuracy than

using circles.

The calibration process has been designed in an

intuitive and detailed way. Each phase of calibration

is accompanied by an instructional video in which an

avatar demonstrates the speciﬁc movements that the

user must replicate. At the end of each instructional

video, the application emits a sound to indicate that

the user must begin to perform the movements or po-

sition themselves in the explained pose.

During the whole process, the user must keep his

back well supported and without twisting his trunk.

This requirement is essential in order to know the de-

gree of mobility of the upper limbs without the inter-

vention or inﬂuence of trunk movement.

In the ﬁrst stage, the patient is instructed to look

forward for a few seconds. This ﬁrst step captures

the position and rotation of the HMD, as its default

position is not always aligned with the zero position

on all three axes.

hmdInitialPosition = (x

hmd

hmdInitalRotation = (α

hmd

,β

hmd

,γ

hmd

where (α

hmd

,β

hmd

,γ

hmd

) are the Euler angles.

The second stage is for the patient to bend his

elbows 90 degrees while keeping his arms aligned

across the hip, making sure his hands are above the

knees. This position helps determine the central area

and ideal position for many virtual reality games.

An example of the importance of these positions is

the BBT-VR. This calibration ensures that the box is

placed in front of the patient at the correct height,

aligning with the elbows bent at 90 degrees.

centerPalmR = (x

centerPalmL = (x

During the third step, the calibration process is

further divided into two new phases. First, a detailed

calibration of the upper limbs movements in three di-

mensional planes is performed.

In this new ﬁrst phase, the avatar asks the user to

fully extend the arm in the transversal plane and move

the arm from one end of the X-axis to the other, keep-

ing the arm above the knees (see Figure 2.a). This

stage is designed to calibrate the range of motion in

the horizontal plane and the patient’s ability to reach

laterally, thus capturing the range of possible move-

ments in this plane.

The second part of the calibration relates to the

frontal and sagittal planes. As in the previous stage,

the patient is asked to fully extend the arm. After the

avatar has shown the necessary movements, the cali-

bration begins. These movements consist of extend-

ing the arm from one end of the X-axis to the other

and moving it along the Y-axis (height) and Z-axis

(depth) (see Figure 2.b and c). The trajectory of these

movements simulates the drawing of a dome in the

air. This process is designed to assess and calibrate

the patient’s ability to perform both vertical and depth

movements. Figure 3 shows an example of ellipses

formed in the sagittal, frontal, and transverse planes,

outlining the patient’s movement scope. Next to the

ellipses, less ﬁrm traces, generated by the user during

the calibration process, are visible.

Simultaneously, and transparently to the user, the

system creates two lists of points during the calibra-

tion process. The ﬁrst list, denoted as (P

), collects

the points traced by the patient’s arm in the X Z plane.

This list encapsulates all the trajectories performed in

the XZ plane.

The second list,P

XY/Y Z

, consists of the points gen-

erated during movements in the XY and Y Z planes.

These two lists, together provide a complete rep-

resentation of the patient’s arm movement capabili-

ties in all three axes. This separation into two lists

Advanced VR Calibration for Upper Limb Rehabilitation: Making Immersive Environments Accessible

383

Figure 4: An example of sagittal, frontal, and transverse

ellipses resulting from the calibration process.

allows a more detailed and speciﬁc assessment of the

movement capabilities in different planes, facilitating

a more accurate calibration.

The fourth stage shows the calibration results. Af-

ter completing the calibration, the system draws three

ellipses. Each ellipse corresponds to the range of mo-

bility of the patient’s arm in different planes: XY , XZ,

and Y Z. These ellipses are calculated based on the

trajectory data (P

and P

) collected for each arm,

ensuring they represent the patient’s actual movement

capabilities. The ellipses are deﬁned as follows:

1. The center of each ellipse, denoted as C, is com-

puted as the average position of the trajectory

points within each plane:

C =

∑

i=1

∑

i=1

∑

i=1

for (x

) ∈ P

or P

2. The radii of the ellipses are calculated based on

the maximum distances from the center point to

the points within each plane. For a set of points

{(x

)}, the radii in the X,Y , and Z directions

are calculated as:

= max

)∈P

− x

|,r

= max

)∈P

− y

= max

)∈P

− z

To test the accuracy of the calibration, a virtual

block is displayed at the center of the palm of the

patient’s hand (see Figure 5). The patient can move

their hand freely, with the block remaining attached

to the palm’s center. The color of the block provides

immediate feedback: it appears green if it is within

the deﬁned ellipses, signifying that the hand is within

the calibrated workspace. If the block turns red, it in-

dicates that the hand has moved outside the deﬁned

working area, surpassing the calibrated range of mo-

tion. To determine if a point is inside the calibrated

workspace, deﬁned by ellipses in the XY , XZ, and Y Z

planes, the following mathematical condition is used:

Let O = (o

) represent the position of the

object (e.g., virtual block) in 3D space, and let C =

) be the center of the ellipses with radii r

, and r

in the X ,Y , and Z dimensions, respectively.

The point O is considered to be inside the calibrated

workspace if the following conditions are met for

each plane:

Inside XY-plane:



− c





− c



≤ 1

Inside XZ-plane:



− c





− c



≤ 1

Inside YZ-plane:



− c





− c



≤ 1

Figure 5: Post-calibration environment with a cube attached

to the hand, indicating boundaries in green (within limits) or

red (outside limits).

Finally, the ﬁfth stage involves the storage of the

calibration data. Once the calibration is completed

and veriﬁed for accuracy, the system stores the rele-

vant data in a historical record. This allows for the

calibration to be referenced or adjusted in future ses-

sions, ensuring a tailored and efﬁcient VR experience

for each patient.

ICEIS 2024 - 26th International Conference on Enterprise Information Systems

384

The stored historical data, used for game cali-

bration, also presents another value for medical spe-

cialists. By regularly repeating the calibration pro-

cess over time, changes in a patient’s mobility can

be closely monitored. This enables a more personal-

ized rehabilitation approach, tailored speciﬁcally ac-

cording to the evolving mobility data of each patient,

thereby optimizing the rehabilitation process.

4.1 Application of Calibration Data in

BBT-VR

In the speciﬁc case of the BBT-VR (see Figure 6),

which is a monomanual exercise, the precise position-

ing of the virtual box is essential. The calibrated data

is used to ensure that the box is placed in a comfort-

able position in front of the user, without requiring

excessive arm extension.

Additionally, the height of the virtual box is care-

fully adjusted to be suitable for the patient’s stature. It

is positioned to avoid being too high or too low, which

might interfere with the patient’s legs, considering the

seated position during the exercise.

Furthermore, the box is centered relative to the pa-

tient’s midline, ensuring that the exercises are sym-

metric and balanced, regardless of which hand is be-

ing used.

To facilitate these adjustments in BBT-VR, an op-

tion has been added to the conﬁguration menu that

allows users to choose between automatic or man-

ual calibration. The manual option involves the user

physically grabbing the box and moving it to the ap-

propriate position. Additionally, while the box is ide-

ally positioned in front of the user, the option for lat-

eral positioning has been introduced. This lateral po-

sition will be to the right if the right hand is selected

as dominant, and to the left if the left hand is chosen.

Depending on the selected conﬁguration, the

BBT-VR begins by placing the box either manually

or automatically, and in the selected position, either

central or lateral.

5 EVALUATION AND RESULTS

The evaluation was conducted with a diverse group

of participants, including both men and women, aged

between 20 and 43 years. All participants selected did

not have any mobility issues in their upper limbs. The

study was divided into two sessions: the ﬁrst involved

6 participants, while the second included 9.

Initially, participants received a comprehensive

brieﬁng about the test session’s procedures. The ﬁrst

part of the session focused on testing an automatic

Figure 6: BBT implementation in VR.

calibration system (see Figure 7). Participants were

required to follow instructions provided by an avatar,

mimicking the indicated movements. To correctly

perform the calibration, participants were asked to

keep their backs against the backrest and ensure all

movements were made without twisting or shifting

their torso. After calibration, they were asked to ver-

ify if the deﬁned area appropriately adapted to the ac-

cessible area for them, without the need for excessive

movements or trunk torsions.

Subsequently, the participants were introduced to

a BBT-VR application. In this part, they tested the

automatic calibration, both in central and lateral posi-

tioning, for the right and left hands.

Finally, participants completed a questionnaire to

gather their impressions and feedback on the experi-

ence. The questionnaire, which was explained to each

Advanced VR Calibration for Upper Limb Rehabilitation: Making Immersive Environments Accessible

385

participant before completion, comprised 10 ques-

tions. The initial set collected basic information, such

as the participant’s assigned identiﬁer, age, height,

and hand dominance. Question 4 asked about previ-

ous experience with immersive virtual reality (HMD)

devices, asking: ’have you used immersive virtual re-

ality HMD devices before?’, with possible answers:

’Never (N)’, ’Occasionally (O)’ or ’Frequently (F)’.

The following questions examined their experiences

with the calibration system and the VR application,

using a Likert scale for responses:

• Q6: did you ﬁnd the automatic calibration system

complex? Rate on a scale of 1 to 5, where 1 is not

complex at all and 5 is very complex.

• Q7: after calibration, do you feel that the work-

ing area deﬁned by the three ellipses adequately

matches the range of motion limits of your upper

limb? Please rate on a scale of 1 to 5, where 1

means ’not at all’ and 5 means ’completely’.

• Q8: is the position of the box in the BBT-VR

application, with automatic calibration in normal

mode (centered to the user), correct for you?

• Q9: is the position of the box in the BBT-VR ap-

plication, with automatic calibration in right lat-

eral mode, correct for you?

• Q10: is the position of the box in the BBT-VR ap-

plication, with automatic calibration in left lateral

mode, correct for you?

For questions Q8, Q9, and Q10, participants rated

the appropriateness of the box’s position on a Likert

scale from 1 to 5, where 1 indicates the lowest score

(least appropriate) and 5 indicates the highest score

(most appropriate).

Also, to facilitate observation and veriﬁcation of

the participants’ movements and actions during the

test, an external screen was used. This screen mir-

rored in real-time what the participants were seeing

through the Meta Quest 3.

Figure 7: Moments captured during evaluation, while par-

ticipants test the calibration method.

5.1 Results

The results obtained in the study were analyzed from

two complementary perspectives. The ﬁrst, a quanti-

tative approach, focused on technical considerations

related to the elliptical volume deﬁned by ellipses in

the XZ, XY , and ZX planes. For the central position

of the box, it was veriﬁed that the center of the box

was located within the elliptical volume, in relation to

both the right and left sides of the upper limbs. For

the lateral positions of the box, it was conﬁrmed that

the center of the box was within the control area de-

ﬁned by the corresponding elliptical volume (right or

left).

The second perspective, with a qualitative ap-

proach, focused on the patient’s viewpoint and per-

ception. This subjective evaluation sought to under-

stand the experiences and impressions of the partic-

ipants regarding the calibration process and their in-

teraction with the virtual environment. Analyzing the

results of the VR experience questionnaire, we identi-

ﬁed several signiﬁcant trends that reﬂect participants’

perceptions of the VR calibration system and its sub-

sequent use in the BBT-VR application.

The data presented in Table 1 includes the re-

sponses to the questionnaire. To complete this

overview, it is worth mentioning that for question Q5,

all participants reported being right-handed. The ease

of use of the calibration system is highlighted, as ev-

idenced by the low average score of 1.07 in question

Q6, with a standard deviation of only 0.26. This indi-

cates that most participants found the system straight-

forward, with very little variation in their responses,

reinforcing the notion of an intuitive and easy-to-

manage design.

Regarding the accuracy of the working area (Q7),

the high average scores of 4.27, with a standard devi-

ation of 1.03, suggest that participants generally per-

ceived that the deﬁned working area aligned well with

the movement limits of their upper limbs.

As for the appropriateness of the box’s position in

the BBT-VR application (Q8, Q9, Q10), the predomi-

nantly positive responses are reﬂected in the high av-

erage scores (4.73 for Q8, 4.33 for Q9 and Q10), with

relatively low standard deviations (0.46 for Q8, 0.62

for Q9, and 0.49 for Q10). These ﬁgures indicate a

generally favorable perception of the box calibration

in all modes. However, it is interesting to note the

trend of slightly higher scores for the box’s central

position (normal mode) compared to the lateral posi-

tions. This could suggest a perceived greater accuracy

or comfort in the central setup, although the variations

are not substantial.

To conclude, the study provides a comprehensive

ICEIS 2024 - 26th International Conference on Enterprise Information Systems

386

understanding of the VR upper limb calibration sys-

tem and its integration into the BBT-VR, using both

quantitative and qualitative approaches. From a tech-

nical perspective, the precise placement of the BBT-

VR box within the elliptical volume was validated in

both central and lateral positions, ensuring the cor-

rect execution of the exercises without excessive ef-

fort or compensatory movements. From a user per-

spective, the ease of use of the system and the posi-

tive perception of the accuracy of the workspace and

box placement in the application were highlighted.

These results underline the technical effectiveness of

the system and its positive reception by users, high-

lighting its potential in precision-oriented VR appli-

cations while ensuring user comfort.

Table 1: Table summarizing the responses collected from

the questionnaire on automatic calibration. The columns,

from left to right, represent the participant ID, age, height,

and responses to questions related to the automatic calibra-

tion system and the BBT-VR.

ID Age Ht Q4 Q6 Q7 Q8 Q9 Q10

1 21 178 N 1 5 4 3 5

2 27 160 F 1 4 5 4 4

3 24 174 F 1 4 5 5 5

4 21 191 N 2 5 5 4 4

5 21 165 N 1 4 5 5 4

6 21 180 O 1 5 5 5 5

7 40 166 F 1 4 5 4 5

8 42 172 F 1 4 4 5 4

9 38 179 O 1 5 5 4 4

10 30 165 O 1 4 5 5 4

11 37 166 F 1 5 4 4 4

12 41 178 N 1 5 5 4 4

13 29 175 O 1 4 5 4 4

14 27 173 F 1 1 5 5 5

15 43 177 F 1 5 4 4 4

6 LIMITATIONS AND FUTURE

WORK

Although the initial results of the calibration system

are promising, further validation is needed to conﬁrm

its effectiveness. A larger study with a larger number

of participants, including people with cSCI, is needed

to conﬁrm the preliminary results. Previous tests were

conducted exclusively with healthy individuals, as the

primary objective of this initial phase was to estab-

lish the system’s safety before extending the testing

to patients in subsequent phases. Additionally, storing

other variables like the maximum degree of wrist ﬂex-

ion/abduction and arm rotation should be considered

to enhance the calibration accuracy and applicability.

Following the study, there is a need to automate

the implementation of the calibration system for each

set. In the current iteration, speciﬁc values of the cal-

ibration data were manually selected and applied to

correctly position elements in 3D space for the test

set. The aim for future development is to automate

this aspect so that the system can classify each set ac-

cording to a set of parameters and automatically apply

the calibration values in an efﬁcient manner.

This will require a thorough study and ranking

of the relevant parameters. These parameters could

include factors such as whether the exercise is one-

handed or bimanual, involves lateral trunk displace-

ment or not, and requires centred positioning or posi-

tioning adapted to the range of motion of each arm. It

will also be important to consider the plane of move-

ment, whether vertical or horizontal. This second

phase of implementation of the VR upper limb reha-

bilitation calibration system aims to facilitate its im-

plementation in any game in a simple way.

7 CONCLUSIONS

This article presents an innovative calibration method

for immersive spaces used in upper limb rehabilita-

tion. The method particularly accounts for the mo-

tor limitations of patients with cervical spinal cord in-

juries who require rehabilitative therapy to regain mo-

bility. A key novelty of the method is its considera-

tion of the limitations of each limb individually, form-

ing three ovals in the sagittal, frontal, and transverse

planes. The set of generated ovals, along with the

proposed methods, are used to reconﬁgure the virtual

environment in which the patient exercises, ensuring

that all objects they interact with are within reach

based on the detected limitations. In other words, the

environment adapts to become an accessible, suitable,

and safe space for therapy.

The correct distribution of virtual components

with which the user interacts, taking their limitations

into account, also prevents more signiﬁcant problems

such as body compensations that could endanger the

integrity of healthy parts.

Finally, the different calibrations performed by the

same patient at different times, along with the record-

ing of kinematics during exercise, provide therapists

with a valuable tool for objectively measuring a pa-

tient’s progress. In particular, comparing calibrations

allows for the measurement of improvements in mo-

bility.

Advanced VR Calibration for Upper Limb Rehabilitation: Making Immersive Environments Accessible

387

ACKNOWLEDGEMENTS

This work has been founded by the Spanish Ministry

of Science, Innovation and Universities under the Re-

search Project: Platform for Upper Extremity Reha-

bilitation based on Immersive Virtual Reality (Rehab-

Immersive), PID2020-117361RB-C21 and PID2020-

117361RB-C22.

REFERENCES

AlMousa, M., Al-Khalifa, H. S., and AlSobayel, H. (2020).

Move-IT: A Virtual Reality Game for Upper Limb

Stroke Rehabilitation Patients. In Computers Helping

People with Special Needs: 17th International Con-

ference, ICCHP 2020, Lecco, Italy, September 9–11,

2020, Proceedings, Part I, pages 184–195, Berlin,

Heidelberg. Springer-Verlag.

Cha, K., Wang, J., Li, Y., Shen, L., Chen, Z., and Long,

J. (2021). A novel upper-limb tracking system in a

virtual environment for stroke rehabilitation. Journal

of NeuroEngineering and Rehabilitation, 18(1):166.

Creed, C., Al-Kalbani, M., Theil, A., Sarcar, S., and

Williams, I. (2023). Inclusive AR/VR: accessibility

barriers for immersive technologies. Universal Access

in the Information Society.

de Los Reyes-Guzm

an, A., Lozano-Berrio, V., Alvarez-

Rodr

ıguez, M., L

opez-Dolado, E., Ceruelo-Abajo, S.,

Talavera-D

ıaz, F., and Gil-Agudo, A. (2021). Rehab-

Hand: Oriented-tasks serious games for upper limb

rehabilitation by using Leap Motion Controller and

target population in spinal cord injury. NeuroReha-

bilitation, 48(3):365–373.

Dias, P., Silva, R., Amorim, P., La

ıns, J., Roque, E.,

Ser

odio, I., Pereira, F., and Santos, B. S. (2019). Using

Virtual Reality to Increase Motivation in Poststroke

Rehabilitation. IEEE Computer Graphics and Appli-

cations, 39(1):64–70. Conference Name: IEEE Com-

puter Graphics and Applications.

Everard, G., Otmane-Tolba, Y., Rosselli, Z., Pellissier, T.,

Ajana, K., Dehem, S., Auvinet, E., Edwards, M. G.,

Lebleu, J., and Lejeune, T. (2022). Concurrent valid-

ity of an immersive virtual reality version of the Box

and Block Test to assess manual dexterity among pa-

tients with stroke. Journal of NeuroEngineering and

Rehabilitation, 19(1):7.

Goudman, L., Jansen, J., Billot, M., Vets, N., Smedt, A. D.,

Roulaud, M., Rigoard, P., and Moens, M. (2022). Vir-

tual Reality Applications in Chronic Pain Manage-

ment: Systematic Review and Meta-analysis. JMIR

Serious Games, 10(2):e34402.

Herrera, V., Vallejo, D., Castro-Schez, J. J., Monekosso,

D. N., de los Reyes, A., Glez-Morcillo, C., and Al-

busac, J. (2023). Rehab-Immersive: A framework to

support the development of virtual reality applications

in upper limb rehabilitation. SoftwareX, 23:101412.

Juan, M.-C., Elexpuru, J., Dias, P., Santos, B. S., and

Amorim, P. (2023). Immersive virtual reality for upper

limb rehabilitation: comparing hand and controller in-

teraction. Virtual Reality, 27(2):1157–1171.

Kapandji, I. (1971). The Physiology of the Joints, Volume I,

Upper Limb, volume 50. Churchill Livingstone.

Khundam, C. and N

oel, F. (2021). A Study of Physical

Fitness and Enjoyment on Virtual Running for Ex-

ergames. International Journal of Computer Games

Technology, 2021:1–16.

Lagos Rodr

ıguez, M., Garc

ıa, A. G., Loureiro, J. P., and

Garc

ıa, T. P. (2022). Personalized Virtual Reality En-

vironments for Intervention with People with Disabil-

ity. Electronics, 11(10):1586. Number: 10 Publisher:

Multidisciplinary Digital Publishing Institute.

Lim, D. Y., Hwang, D. M., Cho, K. H., Moon, C. W., and

Ahn, S. Y. (2020). A Fully Immersive Virtual Reality

Method for Upper Limb Rehabilitation in Spinal Cord

Injury. Annals of Rehabilitation Medicine, 44(4):311–

319.

Luo, Z., Lim, A. E.-P., Durairaj, P., Tan, K. K., and Ve-

rawaty, V. (2023). Development of a compensation-

aware virtual rehabilitation system for upper extrem-

ity rehabilitation in community-dwelling older adults

with stroke. Journal of NeuroEngineering and Reha-

bilitation, 20(1):56.

Mallari, B., Spaeth, E. K., Goh, H., and Boyd, B. S. (2019).

Virtual reality as an analgesic for acute and chronic

pain in adults: a systematic review and meta-analysis.

Journal of Pain Research, 12:2053–2085.

Mathiowetz, V., Volland, G., Kashman, N., and Weber, K.

(1985). Adult norms for the box and block test of man-

ual dexterity. The American Journal of Occupational

Therapy, 39(6):386–391.

Mott, M., Cutrell, E., Gonzalez Franco, M., Holz, C., Ofek,

E., Stoakley, R., and Ringel Morris, M. (2019). Acces-

sible by Design: An Opportunity for Virtual Reality.

In 2019 IEEE International Symposium on Mixed and

Augmented Reality Adjunct (ISMAR-Adjunct), pages

451–454.

Mott, M., Tang, J., Kane, S., Cutrell, E., and Ringel Mor-

ris, M. (2020). “I just went into it assuming that I

wouldn’t be able to have the full experience”: Under-

standing the Accessibility of Virtual Reality for Peo-

ple with Limited Mobility. In The 22nd International

ACM SIGACCESS Conference on Computers and Ac-

cessibility, pages 1–13, Virtual Event Greece. ACM.

Onitsuka, K., Kubota, K., Yokoyama, M., Miyazawa, T.,

Hamaguchi, T., Taniguchi, H., Usuki, N., Miyamoto,

S., Okazaki, K., Murata, K., and Kanemura, N.

(2023). Clinical utility of markerless motion capture

for kinematic evaluation of sit-to-stand during 30 s-

CST at one year post total knee arthroplasty: a ret-

rospective study. BMC Musculoskeletal Disorders,

24(1):254.

Park, J.-S., Jung, Y.-J., and Lee, G. (2020). Vir-

tual Reality-Based Cognitive–Motor Rehabilitation

in Older Adults with Mild Cognitive Impairment:

A Randomized Controlled Study on Motivation and

Cognitive Function. Healthcare, 8(3):335. Number:

3 Publisher: Multidisciplinary Digital Publishing In-

stitute.

ICEIS 2024 - 26th International Conference on Enterprise Information Systems

388

Pei, S., Chen, A., Chen, C., Li, F. M., Fozzard, M., Chi, H.-

Y., Weibel, N., Carrington, P., and Zhang, Y. (2023).

Embodied Exploration: Facilitating Remote Accessi-

bility Assessment for Wheelchair Users with Virtual

Reality. In Proceedings of the 25th International ACM

SIGACCESS Conference on Computers and Accessi-

bility, ASSETS ’23, pages 1–17, New York, NY, USA.

Association for Computing Machinery.

Shahmoradi, L., Almasi, S., Ahmadi, H., Bashiri, A., Azadi,

T., Mirbagherie, A., Ansari, N. N., and Honarpishe,

R. (2021). Virtual reality games for rehabilitation of

upper extremities in stroke patients. Journal of Body-

work and Movement Therapies, 26:113–122.

Soomal, H. K., Poyade, M., Rea, P. M., and Paul, L. (2020).

Enabling More Accessible MS Rehabilitation Train-

ing Using Virtual Reality. Advances in Experimental

Medicine and Biology, 1262:95–114.

Teo, J. L., Zheng, Z., and Bird, S. R. (2022). Identifying

the factors affecting ‘patient engagement’ in exercise

rehabilitation. BMC Sports Science, Medicine and Re-

habilitation, 14:18.

Zhang, W., Paudel, D., Shi, R., Liang, J., Liu, J., Zeng, X.,

Zhou, Y., and Zhang, B. (2020). Virtual Reality Ex-

posure Therapy (VRET) for Anxiety Due to Fear of

COVID-19 Infection: A Case Series. Neuropsychi-

atric Disease and Treatment, 16:2669–2675.

Advanced VR Calibration for Upper Limb Rehabilitation: Making Immersive Environments Accessible

389