Motor Rehabilitation for People with Cognitive Impairments ...deim.urv.cat/~francesc.serratosa/2010_09_07_Joardi_Alsina_MEIS_Mas… · Motor Rehabilitation for People with Cognitive

Motor Rehabilitation for People withCognitive Impairments: The Case of

the Wii Balance

Master’s ThesisMaster en Enginyeria Informatica i de la Seguretat

Universitat Rovira i Virgili

Author:Jordi Alsina

Directors:Agustı Solanas

Francesc Serratosa

02-09-2010

Contents

1 Introduction 6

1.1 Objectives . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 8

1.2 Organization . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 9

2 State of the art 10

2.1 Technology in rehabilitation systems . . . . . . . . . . . . . . . . . . . . . . 10

2.1.1 Horse riding and artificial saddles . . . . . . . . . . . . . . . . . . . . 10

2.1.2 Common computer peripherals . . . . . . . . . . . . . . . . . . . . . 13

2.1.3 Video games . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 24

2.1.4 Virtual Reality . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 25

2.2 The Wii in rehabilitation systems . . . . . . . . . . . . . . . . . . . . . . . . 34

2.2.1 Used for therapy . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 35

2.2.2 Center of pressure measurement . . . . . . . . . . . . . . . . . . . . 42

3 Our method 49

3.1 Design . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 52

3.1.1 Structure . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 53

3.1.2 Test phases . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 55

3.2 Results . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 57

3.2.1 Participants . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 57

3.2.2 Tests . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 57

4 Analysis of the biometric detection system 63

1

Jordi Alsina Pujol - Motor Rehabilitation for People with Cognitive Impairments

5 Conclusions 68

2

Resum

La fisioterapia es una branca de les ciencies de la salut que s’utilitza per ajudar lespersones amb deteriorament cognitiu a millorar el moviment i les habilitats motrius.Normalment, un fisioterapeuta determina el grau de limitacions fısiques de la personaafectada, i decideix quines son les tecniques de fisioterapia mes beneficioses per alpacient amb paralisi cerebral. Aquestes tecniques ajuden als pacients a millorar leshabilitats motrius que utilitzen els musculs grans del cos, com els dels bracos i les cames.Aquest tipus de fisioterapia pot ajudar a millorar l’equilibri i el moviment d’un pacientamb paralisi cerebral. No obstant aixo, els pacients son reticents a fer aquests exercicisperque pensen que son difıcils i avorrits.

Existeix la necessitat d’incrementar la motivacio dels nens amb paralisi cerebral perrealitzar aquests exercicis. L’us de sistemes basats en Tecnologies de la Informacio i laComunicacio (TIC) poden convertir la monotona terapia en una activitat divertida per al’usuari.

En aquesta tesi presentem un resum de les principals tecniques de rehabilitacio,basades en les TIC, per millorar la rehabilitacio motora en persones amb deterioramentcognitiu. A mes, ens centrem en els estudis que utilitzen la Wii Balance Board com ainstrument de rehabilitacio o com una eina per mesurar l’equilibri del pacient. D’altrabanda, presentem la nostra proposta. Consisteix en una aplicacio, basada en la WiiBalance Board, que permet als fisioterapeutes analitzar els trastorns dels pacients enl’equilibri i aplicar exercicis per millorar els problemes motrius.

El nostre software ha estat provat a l’Associacio Provincial de Paralisi Cerebral deTarragona (APPC) [8] pels terapeutes del centre i ha aconseguit resultats esperancadors.

Una caracterıstica important d’aquesta aplicacio, es la capacitat de reconeixer al’usuari que esta a la Wii Balance Board gracies a un sistema de deteccio biometric.

3

Resumen

La fisioterapia es una rama de las ciencias de la salud que se utiliza para ayudar a laspersonas con deterioro cognitivo a mejorar el movimiento y las habilidades motrices.Normalmente, un fisioterapeuta determina el grado de limitaciones fısicas de la personaafectada, y decide cuales son las tecnicas de fisioterapia mas beneficiosas para elpaciente con paralisis cerebral. Estas tecnicas ayudan a los pacientes a mejorar lashabilidades motrices que utilizan los musculos grandes del cuerpo, como los de losbrazos y las piernas. Este tipo de fisioterapia puede ayudar a mejorar el equilibrio yel movimiento de un paciente con paralisis cerebral. Sin embargo, los pacientes sonreacios a hacer estos ejercicios porque piensan que son difıciles y aburridos.

Existe la necesidad de incrementar la motivacion de los ninos con paralisis cerebralpara realizar estos ejercicios. El uso de sistemas basados en Tecnologıas de laInformacion y la Comunicacion (TIC) pueden convertir la monotona terapia en unaactividad divertida para el usuario.

En esta tesis presentamos un resumen de las principales tecnicas de rehabilitacion,basadas en las TIC, para mejorar la rehabilitacion motriz en personas con deteriorocognitivo. Ademas, nos centramos en los estudios que utilizan la Wii Balance Boardcomo instrumento de rehabilitacion o como una herramienta para medir el equilibrio delpaciente. Por otra parte, presentamos nuestra propuesta. Consiste en una aplicacion,basada en la Wii Balance Board, que permite a los fisioterapeutas analizar los trastornosde los pacientes en el equilibrio y aplicar ejercicios para mejorar los problemas motrices.

Nuestro software ha sido probado en la Asociacion Provincial de Paralisis Cerebralde Tarragona (APPC) [8] por los terapeutas del centro y ha logrado resultadosesperanzadores.

Una caracterıstica importante de esta aplicacion, es la capacidad de reconocer alusuario que esta en la Wii Balance Board gracias a un sistema de deteccion biometrico.

4

Abstract

Physiotherapy is a clinical health science used to help people with cognitive impairmentsimprove movement and motor skills. Usually, a physiotherapist determines the degree ofphysical limitations of the affected individual, and decides what physiotherapy techniqueswill be most beneficial to the cerebral palsied patient. These techniques helps patients’ toimprove motor skills that utilize the large muscles in the body, such as those in the armsand legs. This kind of physiotherapy can help improve a cerebral palsy patient’s balanceand movement. However, patients are reluctant to do these exercises because they feelthat they are difficult and boring.

There is the need to increase cerebral palsied children motivation to do theseexercises. The use of systems based on Information and Communication Technologies(ICT) may convert the monotonous therapy in a fun activity for the user.

In this master thesis we present an overview of the main rehabilitation techniques,based on ICT’s, to improve motor rehabilitation in people with cognitive impairments.Furthermore, we focus on studies that utilize the Wii Balance Board as a rehabilitationtool or as a body balance measurement tool. Moreover, we present our proposal. Itconsists on an application, based on the Wii Balance Board, that allow physiotherapiststo analyze the balance disorders of patients and apply exercises to improve the motorproblems.

Our software was tested in the Associacio Provincial de Paralisi Cerebral deTarragona (APPC) [8] by the therapists of the center and has achieved encouragingresults.

An important feature of this application, is the capability to recognize the user who ison the balance board though a biometric detection system.

5

Chapter 1

Introduction

The cerebral palsy is a non-progressive condition, not a disease, caused by damage tothe brain, usually occurring before, during or shortly following birth. It produces a groupof permanent disorders of the development of movement and posture, causing activitylimitation. The motor disorders of cerebral palsy are often accompanied by epilepsy,by secondary musculoskeletal problems, and by disturbances of sensation, perception,cognition, communication, and behavior.

Impairment of motor functioning is the major symptom of cerebral palsy. In general,children with cerebral palsy suffers limited mobility and abnormal (sometimes involuntary)movements. Specifically, such motor impairment can affect locomotion, communication(especially speech) and manual ability. Moreover, many people with cerebral palsyexperience disturbances in posture and balance, which contribute to overall motordysfunction. Children with cerebral palsy tend to have abnormal sitting and standingposture, and demonstrates poor balance compared to healthy individuals of the samestage of development. For example, some children with cerebral palsy have troubleadjusting their posture and sitting independently. Most postural and balance problemsarise from muscle abnormalities. This, induces abnormal gait such as scissor walking1

and toe walking2.

Physical therapy (also physiotherapy), is a health profession that assesses andprovides treatment to individuals with cerebral palsy to improve balance and movement.Since cerebral palsy is a physical and movement disorder that impairs the brain’s abilityto properly control muscle movement, physiotherapy helps cerebral palsy patients gainmobility.

Physiotherapy begins in early infancy and continues throughout adolescence. Theprimary purpose is to facilitate normal neuromotor development. With the help of correctpositioning, appropriate stimulation and intensive exercise the therapist tries to achieve

1Scissor walking. A walking abnormality. It consists of legs flexed slightly at the hips and knees, givingthe appearance of crouching, with the knees and thighs hitting or crossing in a scissors-like movement.

2Toe walking. Walking abnormality consisting of walk on his or her toes without putting much weight onthe heel or any other part of the foot.

6


stability and good mobility in the child.

Monotony of the physiotherapy exercices, is a disadvantage since patients may findtherapy boring or tedious and they can get discouraged. That is why there is the need todo more entertaining therapy and stimulate children to practice this therapy.

Video games are an attractive and direct platform to solve this disadvantage. Theyare based on the interaction between a person and a machine. When playing, we followa series of rules to solve a problem. The main factor in a game is motivation. Thanks tomotivation, players can be happy when they play.

The brain reacts differently to fun ways of learning. The continuous play of videogames helps children with cerebral palsy use their eyes, hands and brains together as atreatment for brain damage. The real benefit is that, while it is difficult for children andtheir families to follow through with conventional prescribed exercises at home betweentherapy appointments, the games are fun so the children will continue to do them.

The Wii Balance Board [35] (Figure 1.1), part of the popular video game WiiFit,contains four transducers which are used to assess force distribution and the resultantmovements in center of pressure. Originally designed as a video game controller, the WiiBalance Board is predominantly used in combination with a video game console and itsassociated software. Given the capacity for providing instant feedback and the potentialfor enhanced motivation levels, this system is a potential tool to be used in physiotherapywith cerebral palsied children.

Figure 1.1: The Nintendo Wii Balance Board which has a similar shape to a body scale,featuring a flat rectangular design. It is a wireless device that can be powered for up to60 h with four AA batteries, and communicates via Bluetooth with the Nintendo Wii videogame console.

There are few research that focus in the Wii Balance Board to improve the motorrehabilitation of cognitively impaired people, because of the complexity of the technologyrequired and the short time that Nintendo Wii is on sale.

Most of the studies that utilize the Wii Balance Board employ the commercialNintendo games to develop therapy with children. Despite this, the results obtained areencouraging. Cerebral palsied children enjoy during therapy because they are playing

7


while they are improving their motor rehabilitation.

That is why, we encouraged to develop a customized software that allow children withcerebral palsy to improve their body balance while their are playing on the Wii BalanceBoard. Moreover, this innovative application may help physiotherapists in their work andallow them to check the evolution of their patients.

1.1 Objectives

In this thesis, we will present an overview of the main rehabilitation techniques basedon Information and Communication Technologies (ICT) for people with cerebral palsy.Specially, we will focus in systems which use the Wii Balance Board as a rehabilitationtool or as a body balance measurement tool.

Additionally, we will design, implement and deploy a system, based on the Wii BalanceBoard [35], to improve the motor rehabilitation of cognitively impaired people. It willconsist of measure the patients’ postural balance while they are playing with the WiiBalance Board. Thus, the physiotherapist can verify the disturbances that have thepatients in balance.

This system is designed to people who have trouble adjusting their posture, sittingindependently and maintain their body balance. It means that most of them could notstay balanced for a long time on the Wii Balance Board. Is for this reason that, theapplication should be capable to measure the child’s body balance despite the time thatthe child remains on the platform.

The tests will be distinguished according to its typology (standing, kneeling or sitting).For each test, the physiotherapist will annotate the support tools used in accordance withthe typology.

During a test, the system will have an area where the patients would interact with thesystem. This area will consist of a real-time graph with the child’s center of gravity. As theuser moves on the Wii, the real-time graph’s output will be updated. The physiotherapistdecides in each test if the patient must see this graph or not.

This application should allow the physiotherapists verify the patients’ evolution. Toaccomplish this, the physiotherapist may consult the statistical results of each performedtest.

To facilitate the task of the physiotherapist, a detection system of the biometric datawill be designed. This method just should utilize the measures transmitted by the WiiBalance Board determine which user corresponds the test that has been performed. Thistest mode, will be optional and will be the physiotherapist who decides enable it.

8


1.2 Organization

This document presents a study of the motor rehabilitation for people with cognitiveimpairments, focusing in the use of the Nintendo Wii Balance Board as a therapy tool.

That is the first chapter in which we introduce the subject of the cerebral palsy.Moreover, we explain the motivations to carry out this study and the objectives to beachieved.

In the second chapter, we present an overview of the state of the art in what refersto motor rehabilitation techniques based on ICT’s for cognitively impaired people. First ofall, we explain rehabilitation methods that are not related with the Nintendo Wii console.This includes both projects which develop customized software and/or specific hardware,and projects which use common hardware and commercial software.

In chapter three, we propose our method, based on the Wii Balance Board, to improvethe motor rehabilitation of cognitively impaired people. In this chapter, we detail thedesign of this application and, then, we describe the experimental tests performed, witha total of three cerebral palsied children, and the results obtained.

Afterwards, there is an exclusively chapter dedicated to the detection system ofbiometric data. This chapter explains the method used to identify users when they are onthe Wii Balance Board. A trial with six healthy users was performed and described in thischapter.

Finally, in chapter 5, we detail the conclusions that we have extracted along this study.

9

Chapter 2

State of the art

In this chapter, we will summarize the state of the art of technologies applied to therehabilitation of children with cerebral palsy. To do a more appropiate classification forthis study, we have separated the proposals into two sections.

In the first section, we explain the technology used in cerebral palsy rehabilitation thatis not related with the Nintendo Wii console. This includes both projects which developcustom-written software and/or specific hardware, and projects which use commonhardware and commercial software.

In the second section of this chapter, we focus on studies, which use the NintendoWii console or some of its components (i.e. Wiimote [33], Wii Balance Board [35]). Someof these studies use the Nintendo Wii exclusively for therapy, while other investigationsutilize the Wii Balance Board as a centre-of-pressure detector too onlyl.

As it can be observed, there are several proposals which use new technologies toimprove the therapy with cerebral palsy patients. However, very little research has beenconducted with the Nintendo Wii because it was launched on September 2006.

Furthermore, several studies with Nintendo Wii and other video consoles apply thecommercial videogames to therapy. Customized software has more advantages overcommercial software because it can be more powerful tool in the fields concerningpeople with disabilities, but this is rarely proposed by researches because of the relatedtechnology complexity.

2.1 Technology in rehabilitation systems

2.1.1 Horse riding and artificial saddles

Horse riding therapy (hippotherapy) is a form of physical, occupational and speechtherapy in which a therapist uses the characteristic movements of a horse to provide

10


carefully graded motor and sensory input. For individuals with mental or emotionaldisabilities, the unique relationship formed with the horse can lead to increasedconfidence, patience and self-esteem. The sense of independence found in horsebackbenefits all who ride. The therapeutic qualities of horseback riding are recognized bymany medical professionals.

Hippotherapy is an ideal practice to improve body balance and is beneficial for childrenwith cerebral palsy to maintaining a stable erect posture. However, this practice is oftennot available to most patients due to the limited access to horses, the relatively high costand the dangers of horseback riding.

In 1998, Michal Kuczynski and Karina Slonka studied the use of a microprocessor-driven artificial saddle to simulate horse riding movement [1]. This research had thepurpose of finding out whether the artificial saddle ride has any instantaneous influenceon postural stability. To assess the body balance they used stabilography techniques,specifically the excursion of center-of-pressure (COP), because of relative low cost andsimplicity of procedures. Also, a parametric frequency-based method was applied asa complementary technique to evaluate the time-varying activity of muscles around theankle joint which prevents subjects from falling and may be regarded as a control variablein maintaining upright stance. Moreover, subjective opinions of parents and therapistscould be useful to reinforce the results. Two groups of children were selected to undergothis study. A total of 25 cerebral palsy children range 3-10 years formed the first group. Allof them could independently walk and maintain an upright posture. Moreover, 33 healthychildren composed the second group. This group was selected to have an appropriatereference basis. The patients ride on the saddle for about 20 minutes each session twicea week for 3 months.

The microprocessor-controlled the artificial saddle BABS1 (Brunel Active BalanceSaddle) movements were 3D and mimic the horse’s walk (Figure 2.1).

The postural stability of patients was tested in two occasions (in the first and lastsession) to evaluate the final progress of the artificial saddle. Each test was performedtwo times, before and just after the ride. The test consists in standing on a forceplatform for 20 seconds to record the COP in anteroposterior (AP) and mediolateral (ML)directions. On the basis of COP signals in AP and ML plane were computed: range,standard deviation mean speed and mean radius.

The comparison of mean speed, standard deviation and range of AP and ML plane inboth groups at first session show highly significant differences.

Healthy children exhibit higher values of all metrics in AP than ML plane while CPchildren only display the same relationship for mean speed only. Standard deviation andrange measures are larger in the ML plane.

The parametric frequency values in the ML plane between both groups are muchlower while the respective results in the AP plane in patients’ group are higher than controlgroup.

1BABS. http://www.brunel.ac.uk/about/acad/bib/researchareas/rehabilitationengg/babs

11


Figure 2.1: Artificial saddle BABS

The original parameters of COP in cerebral palsy group decreased considerably in APdirection between first and last sessions. However, parametric frequency changes wereless significant. Moreover, changes in the ML plane had much the same values thaninitial session in traditional metrics but had a noteworthy decline in parametric frequencytest.

Their conclusions are that short-term effect of repeatedly artificial saddle riding is notlost over time and patients are developing gradually progress in the ability to maintainstable stance.

The AP frequency alone remains constant with regard to long-term effect. However,the ML frequency doesn’t change its values during a single session. Nevertheless, it hasa considerably effect with repeatedly administered therapy and decreases even below therespective frequency in healthy children.

In 2008, Fernandes Lorrain C., Chitra Jeba, Metgud Deepa and Khatri S.M.investigate the postural control in children with spastic diplegia2 when they carried out theartificial horse riding therapy (AHR) [2]. To determine the effectiveness of AHR therapy, asample of 30 children, with spastic diplegia participated. Their ages were between 5 and12 years old.

The patients were randomly separated into two groups of fifteen children (group Aand group B). The test was performed 3 times in a week for 6 weeks where each group

2Spastic diplegia. A form of several palsy that is a neuromuscular contition of hypertonia and spasticity inthe muscles of the lower extremities, usually those of the legs, hips and pelvis.

12


exercised different therapy.

Group A used artificial horse riding and the conventional therapy while group B onlyused the conventional therapy.

To measure the outcomes of this test, two methods were used:

• The first one, Gross Motor Functional Measure (GMFM) is a method that uses thecommon equipment in a physiotherapy gym (e. g. mat, bench, toys) to quantifyingchange in the gross motor abilities of children with cerebral palsy.

• The second one, the Pediatric Balance Scale (PBS), a modification of Berg’sBalance Scale3, was developed for school-age children with mild to moderate motorimpairments. It is used to evaluate standing balance during functional activities andwas also designed to require minimal use of specialized equipment.

Paired t-test applied was significant (p < 0.0001) for both groups for GMFM and PBS.Unpaired t-test was significant for group receiving AHR (p = 0.0494) as compared tocontrol group for GMFM. In PBS, an improvement was also noted in group that practicesAHR. Anyway, this is not statistically a significant improvement.

The researchers concluded that artificial saddle riding has a positive effect on posturalcontrol in children with cerebral palsy. Postural control is an important aspect to improvebecause it is the foundation of gross motor activities. Thus, they suggest that artificialsaddle riding is a viable treatment strategy for therapist as a means to improve posturalcontrol in spastic diplegic people.

2.1.2 Common computer peripherals

An input device, such as a keyboard or mouse, is any peripheral that permits the computeruser to communicate with the computer. Some input devices such as joysticks, areoften used by people who play computer games; the touch panel, which senses theplacement of a user’s finger and can be used to execute commands or access files; andthe microphone, used to input sounds such as the human voice that can run computercommands in conjunction with voice recognition software. There is a variety of inputdevices in a common computer.

Several input devices are used by people with multiple disabilities to improve theirlevels of responsiveness, their quality of life, the interaction, and their postural controland body balance.

Common input devices, used as rehabilitation devices, had a lot of advantages whithrespect to specific rehabilitation systems:

3Berg Balance Scale (BBS). Consists of 14 items that are scored on a scale from 0 to 4. Items includemobility tasks such as transfers, standing unsupported, turning 360 degrees, etc. The literature points toconflicting reports on the usefulness of BBS as a predictor of falls. Overall BBS is described as havingmoderate-good specificity but low sensitivity in predicting falls.

13


• Low cost: Usually, specific devices for rehabilitation are very expensive due itslimited use and home users cannot afford this.

• Easy to obtain: You can buy these input devices in any computer store or shoppingcenter while specific hardware is difficult to obtain in stores and often only you canbuy it through the Internet.

• Good technical support: It is very easy to obtain technical support for common inputdevices such as mice or webcams.

• Easily updated with the newest technology: e.g., 2.4G wireless or Bluetooth mouse.

Mice

Since 2008, Ching-Hsiang Shih and Ching-Tien Shih are researching in the useof computer technology to assist people with multiple disabilities to improve theirenvironmental control stimulation, express their personal needs, interact with others andcommunicate. Many commercially available products which possess the characteristicsof a sensor can be turned into special switches suitable for people with disabilities. Anexample is the common mouse that everyone uses for many computer input tasks.

In September 2008, Ching-Hsiang Shih and Ching-Tien Shih assessed the combina-tion of multiple mice aid with two persons with multiple disabilities [3]. The aim of thiswork was to determine whether the multi-mice implementation could enhance the point-ing performance of people with developmental and physical disabilities.

Two children participate in this study. The first one, Wu was a 10 years old boy whohad limited physical control. He was unable to maintain independent sitting stability andneeded an adjustable positioning chair to help maintain sitting balance.

The other one, Luo was an 11 years old girl with a wide range of intellectual disability.She could maintain an independent sitting balance but she needed the assistance of acaregiver to walk. Moreover, she had poor finger function, but could still control her wrist.

Both participants could not use a normal mouse because of the poor handcoordination.

In this study, the boy employed three mice (Figure 2.2). The function of the right handmouse was only to transfer the right and left movement. In left hand, a trackball was usedto transfer his left thumb movement into up-to-down cursor movements. Finally, a thirdmouse with only button function enabled was placed under his left toe.

The girl was provided with two mice (Figure 2.3). The first one, with only themovement function enabled, was placed on her right hand. Using her left wrist, thesecond mouse was used for clicking.

Two different modes of software were developed:

• Practice mode: designed to provide repeated pointing exercises for participants.

14


1 2 3

Figure 2.2: The three mice employed by Wu. The first mouse was placed under his righthand, rotated clockwise 90 degrees. The second mouse was a trackball, placed near tohis left thumb. The third mouse was placed under his left toe to press the mouse button.

1 2

Figure 2.3: The two mice used by Luo. The first mouse was operated by her right hand,rotated 90 degrees clockwise. The second mouse was placed under her left wrist forclicking.

15


• Test mode: designed to analyze the participants’ performance a certain period oftime.

Both software modes consist in a circle with eight circular destination targets (Figure2.4). In practice mode, the participants had to move the mouse cursor from the center ofthe circumference to the visible target, and then click to complete a hit. Then, the cursoris moved to the center again, the clicked target is removed and the next target appears.

Figure 2.4: The flow diagram of the computer mouse training and test software.

The test mode was run under the same conditions, except for the targets thatappeared randomly. This mode lasted 5 minutes and the number of hits were recorded.

The experiment comprised three phases:

• Baseline, in which the test mode was performed at least three times to collectparticipants’ data.

• Intervention, that consisted in 20 minutes of pointing practice, 5 rest minutes and 5minutes of test mode.

• Maintenance, performed one week after the second phase.

The patients participated directly in the test mode.

A total of three baseline sessions, 11 intervention sessions of 30 minutes and threemaintenance sessions were performed.

Both participants had no successful mouse pointing during the baseline phase but, inthe intervention phase a slow increase of mouse pointing speed was observed. Finally,

16


in the maintenance phase the boy (Wu) retained the skills that he acquired during theintervention phase and the girl (Luo) improved her performance.

Figure 2.5: The mouse pointing speeds of the two participants after the implementationof multi-mice. The curve indicates that both two participants improved their pointingefficiency, and maintained their acquired skills, during the maintenance phase.

In November 2009, Ching-Hsiang Shih and Ching-Tien Shih decided to improve theirsoftware adding more mouse functions [4].

The architecture of this system was composed by a mouse that is responsible for thecursor movement (horizontal and vertical) and four externally connected switches linkedto the computer through a special-switch USB device via a USB interface. Switchesprovide click, double-click and scroll functions.

The software developed (MTD-Software) (Figure 2.6) contains a new mouse driverthat replaces the standard mouse driver, determines the mouse-holding orientation (0o,90o, 180o, 270o), mouse operation direction (horizontal, vertical), and the four switches’functions.

Three not disable people using standard mice and three people with disabilities usingthis system participated in the same test program of thev first study to evaluate thefeasibility of the system. The average successfu mousel pointing times of not disablepeople within 5 minutes was 210 and was adopted as 100%. The three participants withdisabilities successful pointing times within 5 minutes were 65, 45, 55, respectively, with

17


Figure 2.6: The MTD-software.

an overall mean of 55. This means their overall efficiency were 26%.

A disadvantage of this system was that switches needed additional USB hardwareto link to the computer. With this work, Ching-Hsiang Shih and Ching-Tien Shihdemonstrated that both participants rapidly improved their pointing efficiency and retainedtheir acquisition. Moreover, they could operate some ordinary study and game softwarewith multi-mice after the experiment. Furthermore, to solve the disabled people pointingproblems they only need mice and trackballs that have the advantages of being low cost,easy to obtain and have good technical support.

In July 2009, Ching-Hsiang Shih, Ching-Tien Shih, Kun-Tsan Lin and Ming-ShanChiang redesigned Shih’s mouse driver to change a mouse wheel into a precisethumb/finger poke detector to match with the unique characteristics of people withmultiple disabilities and minimal motor behavior [5]. The standard mouse driver wasreplaced by a new version of Shih’s mouse driver that cancelled movement and buttonfunctions of mouse and only transmitted the thumb poke signal (wheel rotation).

The architecture of that system was composed by a minicomputer used as a controlsystem and two mice installed with Shih’s revised mouse driver. Two people, Shen andHua, with wide multiple disabilities and minimal motor behavior but thumb poke abilityparticipated in this study. They reacted to familiar sounds, songs, vibrotactile stimulationand visual stimuli by alerting and following the visual stimuli or turning/widening theireyes.

Both participants initially received an ABAB sequence, in which A representedbaseline and B intervention phases. After this sequence, a post-intervention check wasconducted and three to five sessions a day were occurred. Sessions were conductedat home and lasted 10 minutes. In baseline phases, the mouse wheel and the controlsystem were available, but any stimulation was produced by the control system. Theparticipants only receive vocal promptings from the research assistant to ask them topoke the mouse wheel.

Intervention phases were like baseline phases, but the control system produced atotal of six patients’ favorite stimulation.

Both participants received 12 and 9 sessions in each baseline phase and 48 and63 sessions in each intervention phase, respectively. Before the post-intervention check,

18


they continued to receive a total of eighteen sessions carried out for 2 months comparableto those occurring during the intervention phase.

The first participant (Figure 2.7) had a mean of about 5 independent responses persession in first baseline phase. During the first intervention phase, the mean increasedto about 20 responses per session. This mean frequency dropped to 6.67 during thesecond baseline phase. During the second intervention phase, the mean frequencyeventually increased. Finally, the intervention frequencies were largely maintained atthe post-intervention check.

Figure 2.7: Shen’s data. Data points represent mean frequencies of Shen’s targetresponses per session (independent of prompting) over blocks of three sessions. Onlythe final points of a phase can represent a block of two sessions.

During the first baseline session, the second participant (Figure 2.8) had a meanof about 7 independent responses per session. This mean increased to about 22.6responses per session during the first intervention phase. The mean frequency droppedto 10 during the second baseline phase to be fully restored and eventually increasedduring the last intervention phase. At the post-intervention phase, the interventionfrequencies were largely maintained.

The authors of this study assert that with the assistance of the Shih’s mouse driver,a standard mouse wheel can be used as a high performance thumb/finger poke detectorwith high revolution, low cost, easy to obtain and good technical support. Furthermore,this system can be very important for people with multiple disabilities and minimal motorbehavior to improve their quality of life because it allows them to increase their levelof responding and stimulation control and this can augment the rehabilitation process.Nevertheless, further studies are necessary to focus on other people with multipledisabilities.

Webcam

In March 2004, the ’Asociacion Provincial de Paralisis Cerebral de Tarragona’ (APPC)[8] and CREA Sistemes Informatics [9] signed a cooperation agreement to develop I+D+I

19


Figure 2.8: Hua’s data. Data points represent mean frequencies of Hua’s targetresponses per session (independent of prompting) over blocks of three sessions. Onlythe final points of a phase can represent a block of two sessions.

software, using common and low-cost hardware in the field of IT like webcam, to allowusers with disabilities to improve their environment interaction.

Two programs were developed: ’Facial Mouse’ and ’Webcolor’.

The first one, ’Facial Mouse’ (Figure 2.9) is a handsfree mouse. A webcam detectsthe head movements of the user and turns them into the mouse movement. To click, thereare two methods: click by waiting and click by sound. Furthermore, the program allowsto set up several webcam options like the pointer speed, webcam options, movementstremor and smoothing.

Figure 2.9: ’Facial Mouse’.

The second one, an application called ’Webcolor’ allows the detection in real time ofthe presence or absence of a color mark, locate its position and track it in the capturedimage by a webcam. In addition, the system is also capable of detecting the color ofthe mark automatically. The application can be used as a click or key emulator and as ajoystick emulator.

Sixteen users (11 males and 5 females) from 3 to 35 years old were selected from theAPPC centers. They were selected considering their physical shape and communication

20


and intellectual skills.

For a year, weekly sessions were conducted with those users. The duration of thesessions was based on each user considering their level of attention and fatigue, but amaximum of 45 minutes was preset.

First sessions were used to define an initial contact with the webcam and to familiarizewith the application. These sessions allow to establish initial assessments of user-computer interaction. In next sessions, the application settings and environment wereadjusted in accordance to the user. Several utilities and commercial software like theInternet or Office packet were used by the patients. Moreover, weekly meetings wereheld to pooling observations and improve the application and sessions.

After the test, the researchers could divide the users into three groups:

• Users familiarized with the computer with severe physical disabilities and highcognitive level that cannot use the conventional computer mouse. These userslearned the use of the handsfree mouse to enhance their computer access. Someusers of this group also learned the use of Webcolor as a joystick emulator.

• Users familiarized with the computer with severe physical disabilities and lowcognitive level. The goal for this group was to learn the use of Webcolor as a clickor key emulator. To train them with this application they played with action-reactionsoftware like Toca-Toca4 or specially developed software.

• Users non-familiarized with the computer with severe physical disabilities andseveral cognitive levels. The learning process for these users was slower than forother groups due to their lack of experience with computers. However, the resultsof some users of this group were spectacular.

In general, the Webcolor application as a click or key emulator is useful for peoplewho have severe physical and cognitive disabilities. The joystick emulator mode is moreappropriate for users whose physical and cognitive disabilities are lower. Moreover, the’Facial Mouse’ application is suitable for people with severe physical disabilities and maybe encouraging for them to think in a possible future career.

Webcam and microphone

The SATI5 project [6] [7] is an interactive music system which aim is to encourage peoplewith several mental disabilities to interact. It was developed by Cesar Mauri from CREASistemes Informatics [9], Mabel Garcıa from APPC [8] and Joan Bages, sound artist fromCICM [10].

The system consists of two software modules:4Toca-Toca. Cause-effect software developed by the ’Subdireccio General de Tecnologies de la

Informacio del Departament d’Educacio de la Generalitat de Catalunya’5SATI. Sistema Audiovisual Terapeutico Interactivo.

21


Figure 2.10: Gesture capture system.

• Gesture capture system: An application captures and processes real-time videosignal to extract parameters such as direction of movement, the position of a colormark or amount of activity (Figure 2.10).

• Audiovisual motor : This application generates image and sound through gesturecapture system’s data. It also incorporates sound effects such as reverb and echoto treat the patient’s voice through a microphone. (Figure 2.11).

The MIDI6 protocol was used to communicate both application. To carry out the trial,the users were divided into three groups according to their capabilities:

• Group A: Users with enough intellectual level to understand the activity’s perfor-mance.

• Group B: Users with less intellectual level than group A, but enough to provide anyresponse in sessions.

• Group C: Users with low intellectual level. It is very difficult to observe any responsein sessions, but is not impossible.

Sessions lasted 30 minutes and consisted of 6 working phases, four with voice andtwo with movement. The order of working phases swapped to prevent influences, fatigueor other factors. Every working phase lasted 3 minutes. After the session, an interviewwas conducted with the therapist.

The sessions were recorded to be analyzed later. So they could study the immediateeffects and long-term effects of therapy. Anyway, the long-term effects could not beassociated strictly with this study because it depended on other factors.

6MIDI. Musical Instrument Digital Interface

22


Figure 2.11: Audiovisual motor.

The therapist was responsible for analyzing the videos after the sessions. Somevariables to consider were: user’s attention (dispersed, focused...), the duration, if userwas voluntary or induced, connectivity with the task, user’s satisfaction and disinhibition.

After a year working in this project the authors assert that it is a very useful tool. In theteaching field it promotes the work in areas such as attention, perception, communicativeintention, relaxation or muscle control. In the therapeutic field, this system helps usersto enhance their expressiveness, imagination, and carry out their own desires of playinga musical instrument. Finally, in the recreational area, nowadays it is very difficult to usetherapeutic tools for people with several impairments enabling them to learn, play andhave fun at the same time.

Furthermore, the researchers propose the development of tools to improve thecharacterization of the participants, establish methodologies for analyzing the effectsof the system and relate them to specific user profile, involve other participants in theprocess of analysis (e.g. parents or carers) or to develop automated techniques ofanalysis and creating more autonomous systems and easier to use by professionals.

23


2.1.3 Video games

A video game is a computer program specifically created to entertain. It is based onthe interaction between a person and a machine. These programs recreate virtualenvironments where players can control mainly characters or any other element toaccomplish one or several goals by observing a set of rules.

Video games as learning tools have some advantages:

• Scholar success: Pupils that have used video games have increased their readingcomprehension capability.

• Cognitive abilities: Pupils train these abilities using environments based ondiscovery and creativity.

• Motivation: Games are an encouragement mechanism for children; they make thelearning process easier and increase attendance considerably.

• Attention and Concentration: Attention and concentration are increased to solveconcrete problems due to pupils’ nature towards games.

Video games are an attractive and direct platform to approach children. Theyprovide interesting human-computer interaction methods to enrich the learning processin special education. Moreover, games help to improve social relationships, to raisethe communication level and to ease the assimilation of new concepts that improve thelearning process.

However, the use of video games in special education has important problems:

• Educational video games often are not designed for people with special needs. Fewvideo games with educational contents may be used in this context.

• The already existing video games for special education are mainly didactic unitswhich have lost the essence and attributes of games.

• The devices where these didactic games are implemented on are just simple PCs.They often do not raise children’s interest.

In 2005, Jose Luis Gonzalez, Marcelino J. Cabrera, and Francisco L. Gutierrezdeveloped an application called Sc@ut [14] (Figure 2.12). The goals of this researchproject were:

• The development of the communication skills of people with special needs.

• Enhancing personal autonomy: it must be portable and compact.

• Low-Cost.

24


• Easy to learn and an intuitive interface.

• It should facilitate the work of teachers and parents.

• It must allow to analyse the student’s performance.

This project was an alternative and augmentative communicator (AAC), developed bythe research group GEDES of the University of Granada using PDA devices as hardwareplatforms.

Figure 2.12: Sc@ut communicator.

The therapist could create templates that represent the elements that children needto communicate, depending on the scenario. This allowed the child to express hiswishes by selecting the desired components and listening sounds associated with thesecomponents, like a game. Therapists, watching this navigation, could understand whatthe child wants.

The main features of this application are that templates can be customized basedon user and it assists children integration through its communication. For fifteen days,therapists collected information on children’s behavior.

After analyzing the results, they found that behind every behavioural problem therewas an intention to communicate something. For example, a child sits at his desk whilewatching the teacher plays with another child. Immediately, the child begins to pinch theteacher. The teacher understands that the child’s behavior is a warning (the teacher) tostop playing with other children and play with him.

Through Sc@ut application, the child can communicate by clicking the ’I want to play’template.

2.1.4 Virtual Reality

Virtual reality (VR) is a term that applies to computer-simulated environments that cansimulate places in the real world, as well as in imaginary worlds. Most current virtualreality environments are primarily visual experiences, displayed either on a computer

25


screen or through special stereoscopic displays, but some simulations include additionalsensory information, such as sound through speakers or headphones. Some advancedsystems responds to the user’s movements to give the sense of being immersed in avirtual environment.

Users can interact with a virtual environment or a virtual artifact either through the useof standard input devices such as a keyboard and mouse, or through multimodal devicessuch as a glove, the Polhemus Tracker7, and omnidirectional treadmills.

Virtual reality therapy has the capability of creating a virtual rehabilitation scene wherethe intensity of practice and sensory feedback can be systematically manipulated toprovide the most appropriate, individualized, play-based motor retraining in children [20]or adults with neurological impairments [21]. The use of VR in children with cerebral palsymay enhance postural control and provide them with a sense of mastery and sense ofcontrol over their actions.

The Denise T. Reid study hypothesis [11] was whether the sense of competence orself-efficacy among children with cerebral palsy would improve after engagement in avirtual play intervention program. In the beginning of this clinical trial, several problemswere found using VR with children with disabilities because VR systems tend to restrictmovement, are heavy to the user, cause motion sickness, have a limited field of view, andare not very comfortable. Also, the use of a peripheral like a joystick diminished sense ofimmersion in VR.

To solve these problems the authors used the 1996 patented Mandala GestureXtreme (GX) technology developed by Vivid Group Inc. A video camera located above thescreen was used as a capturing and tracking device to put the user inside VR experiencesand a selected set of VR games were provided. Each game required arm movements ormid-line control through using lateral flexion and trunk rotation movements.

This system offers an alternative to other forms of VR, because users do not have towear Head Mounted Displays (HMD), data gloves, or other devices that tether them tothe computer.

Three children diagnosed with cerebral palsy between the ages of eight and twelveyears participated in this study. Two sessions a week for four weeks were provided. Eachintervention session was 90 minutes and it was split into six sessions of 15 minutes werethe patient played the game that he selected.

The main outcome measure used in this clinical trial was the Canadian OccupationalPerformance Measure8 (COPM) because assesses a person’s level of self efficacy.

Patients were asked to identify the activities that they wanted to do, needed to do, orwere expected to do, and they were asked to think about those activities that required the

7Polhemus Tracker. Magnetic tracking device that allows the capturing of motion of a human body in sixdegrees of freedom.

8Canadian Occupational Performance Measure (COPM). It is an individualized, client-centred measuredesigned for use by occupational therapists to detect change in a client’s self-perception of occupationalperformance over time.

26


Figure 2.13: Experimental set up shows a child interacting with ’Orbosity’ application.

use of their arms. Furthermore, children were asked to discuss a typical day in order toidentify and break down tasks in order to focus on aspects of the task that were difficultfor the child. The COPM’s occupational performance classification scheme showed that44% were self-care tasks, while 42% were leisure tasks, and 14% were productivity tasks.

Additionally, two test periods were defined along the study. The first one was beforeVR sessions and, the second one was after sessions. Comparing the results of them,children enhanced they performance. This means they rated their satisfaction withperformance. The children suggest the VR intervention was a pleasurable experienceand was motivating for them.

The authors assert that the use of VR provide CP children with an opportunity tointeract with virtual play activities that are enjoyable and non-threatening. Furthermore,these exercises allow children to exercise control over their actions which enhancemotivation and feelings of self-efficacy.

In February 2004, Christine Brumback, Liubo Borrisov and Jeffrey Galusha from NewYork University and Annette Dilorio from the Pediatric Center at the New York FoundlingHospital developed IntelliVision, an interactive multi-sensory environment intended forcognitively impaired children in a hospital settings [27].

The project consisted of a large screen display mounted over the bed or wheelchair.The screen displayed video and sound content generated by a computer. The patient’smovements where captured by a video camera and they induce changes in the imageson the screen and sound.

Three different content modules were available. The change of visual effects variedbased on which module were selected.

27


Figure 2.14: IntelliVision system.

Tactile stimulation were provided through a tactile module positioned at different pointson the bed or wheelchair. This device vibrates in response to bass frequencies in theaudio. The therapist could select and adjust the module and its additional settings (motionsensitivity, color palettes, music genre and overall volume) based on a patients particularcondition through a simple screen interface.

A patient with cerebral palsy who had cortical visual impairment and sensory deficitsparticipated in the trial. To capture her attention, she required different visual stimuli thana typical environment provides. The IntelliVision system were placed over the patient’sbed. The therapist selected the MusicMoves program, which shows the patient imagewith a vertical bar scrolling across the screen. Any movement produced musical notes,which appeared on screen. When the bar moved across the notes location on the screenwere played.

Two phases were carried out in this test:

• The first one, was an observation phase. A variety of video footage was shown,including nature scenes, urban imagery and more abstract visuals. Therapistobserved the children’s responses and rated them based on whether it captured orfocused their attention, as well as the overall effect on their behavior. Nature scenesseems to be more calming and focusing, so the rate of change in the content wasmore influential in capturing attention than the content itself.

• The second phase incorporated the patient’s image projected onto the screen andtheir movement would trigger colorful abstract visual effects to the image. They alsotested a variety of music and sounds but again found that the change in the contentwas more important. Vibration stimuli focuses her attention.

In 2005, Sung H You, Sung Ho Jang, Yun-Hee Kim, Yong-Hyun Kwon, IreneBarrow and Mark Hallett investigated the effects of virtual reality therapy on corticalreorganization and associated motor function in a patient with cerebral palsy [22]. Thesystem, IREX9 VR therapy, consisted of a television monitor, a video camera to capture

9IREX. Interactive Rehabilitation EXercise.

28


and track movement to immerse the patient inside the virtual reality scene, cyber gloves,virtual objects and a large screen.

The patient, an eight years old boy had hemiparetic10 cerebral palsy and encelphalo-malacia11 on the left temporo-parietal lobe. He was unable to control his head until 6 to 7months or to stand until 12 months. He performs all functional reaching or grasping tasksusing the left hand.

Five times a week for 4 weeks, the boy played a total of three games:

• The bird-ball : The VR scene was set on an idyllic pastoral hill where small ballsflew towards the child from different directions. The user was only allowed to usehis affected right hand to burst them. Then, the balls transformed into birds and flewaway. This exercise was designed to facilitate development of coordinated reachingand touching motor skills.

• Conveyor : A virtual scene which has multiple applications for exercising real-lifefunctional tasks such as reaching, grasping, holding, and lifting an object. Thisexercise also involves total body exercise, such as bending, twisting, jumping,weight shifting, and stepping.

• Soccer exercise: This virtual scene, simulates a soccer goalkeeper attempting toblock balls from entering the net. The child was only allowed to use the involvedhand to block the balls.

A pretest was implemented before the fourth week virtual reality intervention, followedby the post-test. They included motor function tests and functional MRI (fMRI12).

Motor function tests consisted of:

• The sixth item from subtest 5 of the BOTMP13 [23]. Were used to measure upperlimb coordination.

• The modified PMAL14 [24] [25] questionnaire was used to determine the amount ofuse and quality of movement of the child’s affected upper limb during activities ofdaily living.

• The upper limb subtest of the FMA15 [26] was used to examine sensation, range ofmotion, reflexes, synergy, muscle strength and movement speed.

10Hemiparetic cerebral palsy. Defines an uncontrollable shaking that affects the limbs on one side of thebody and impairs normal movement.

11Encephalomalacia. Referred to a localized softening of the brain substance, due to hemorrhage orinflammation.

12fMRI. Functional Magnetic Resonance Imaging.13BOTMP. Bruininks-Oseretsky Test of Motor Proficiency.14Modified PMAL. Modified Pediatric Motor Activity Log.15FMA. Fugl-Meyer Assessment.

29


In functional MRI, elbow flexion-extension movements imaging were performed on amagnetic resonance scanner. Statistical parametric maps were obtained and the regionsof interest (areas that have been reported to have neuroplastic recovery potential) weredrawn.

In pretest motor functional analysis, the child had no functional use of the affectedhand, which was evident in the PMAL test score. After virtual reality therapy, the BOTMPitem score improved from 1 to 5. The modified PMAL showed that the use of theaffected limb improved from 0 to 3, suggesting increased amount of use and qualitymovement of the affected hand during funcional motor skills. FMA showed improvementin active movement control, reflex activity, and coordination in the upper extremity motorperformance.

In movements of the unaffected elbow, the cortical activations in the regions ofinterest were primarily contralateral, which was similar to normal activation to begin with.Essentially, virtual reality therapy did not influence any meaningful change in the regionsof interest.

The authors conclude that virtual reality therapy produced measurable neuroplasticchanges associated closely with enhancement of age-appropriate motor skills in theaffected limb. The child was able to perform spontaneous reaching, self-feeding, anddressing, which were not possible before the intervention. However, further investigationsare needed to reduce the cost of virtual reality therapy and to compare whetherthe effectiveness and related neuroplastic changes after VR therapy are unique orcomparable with those of other neurorehabilitations.

In 2006, William Li, Tom Chau, Sophie Lam-Damji and Darcy Fehlings concernedin PlayStation 2 (PS2) to develop and evaluate a low-cost virtual reality therapy systemthat elicits practice of neuromotor movements, accommodates a wide range of abilities,provides a simple way for clinicians to track usage, and could be used at home [12].Specifically, the authors focused on the accessory called the ’EyeToy’ (Sony ComputerEntertainment America Foster City, CA, USA). This accessory is a color digital cameradevice, similar to a webcam. This device uses computer vision and gesture recognition toprocess images taken by the camera. Players see their mirror image on the screen, andtheir physical movements are inputs to the games. This allows players to interact withgames using motion, color detection and also sound, through its built-in microphone.

The authors of this study selected two games from the PS2 to perform the test:

• The first one is ’Secret Agent’, where a player looks for toys that appear on thescreen.

• The second one is called ’Mr. Chef’. In this case, the player looks for various fooditems.

To play the games, participants sit on a chair. To control that they sit on the chairsuccessfully, two pushbutton switches (Figure 2.15) were incorporated in the chair. Onepushbutton was situated underneath the seat of the chair on the child’s non-involved side

30


and is held with the non-involved hand. The other pushbutton was situated on the back ofthe chair and must be pressed by sitting upright. These switches encourage movementand extension of the hemiplegic upper extremity in the games. To ensure that they arepressed, the buttons activate an infrared transmitter that controls the TV display. If theyare released, the TV turns off.

Figure 2.15: System setup. The participant sits approximately six feet away from thescreen and looks for virtual objects visible on the TV.

A biometric system was incorporated to activate the system. The user should scan hisor her fingerprint with a fingerprint reader connected to a PC. Then, computer softwarelogs the amount of time that the pushbuttons switches are held.

Five children with hemiplegic cerebral palsy (4 males and 1 female), aged 6 to 9 yearsparticipated in the test. Their levels of fine motor difficulties as indicated by their scoresin the House classification system was between 2 to 5 and the Quality of Upper ExtremitySkills Test (QUEST) [13] were 34.1 to 70.5. In the study, participants played both gameswith a caregiver and occupational therapist present. The sessions were videotaped, andthe number, type, and quality (greater or less than 50% range) of neuromotor movementswere counted by the occupational therapist via video review. After sessions, surveys wereadministered to child and caregiver to evaluate their satisfaction (Figure 2.16).

Figure 2.17 displays the average rates of each neuromotor movement observed forthe five participants. The average playing time was 18.9 minutes and there were a positivecorrelation between the participant’s rate of movements and their total QUEST score.

Figure 2.18 show some statistics about movements and pushbuttons. The largespread in pushbutton compliance occurred because some participants were sliding outof the chair during the games and releasing the back pushbutton.

Few distal movements involving the wrist, finger and thumb were seen because ofplaying with games involve gross reaching motions.

Caregivers commented in their questions that the system was ’great to make [child]

31


Figure 2.16: Average responses to selected questions from child/caregiver question-naires.

work’ and ’a way to make therapy fun’. Their concerns included the speed and difficultyof certain game tasks.

In 2008, Meredith R. Golomb and Monica Barkat-Masih from Indiana UniversitySchool of Medicine and Brian Rabin, Moustafa Abdelbaky, Meghan Huber, and GrigoreBurdea from the Rutgers Telerehabilitation Institute collaborated for over a year on aclinical pilot study of in-home hand telerehabilitation [17]. The system developed (Figure2.19) consists of a 5DT, 5 Ultra sensing glove fitted to the plegic hand, a 26 inch high-definition TV, a keyboard, mouse, a PlayStation3 game console and DSL/Cable routerto access to the Internet. Custom games were programmed in Java3D and the systemswere connected to Riley Hospital and Rutgers Institute.

A total of three patients with hemiplegic cerebral palsy participated in this trial. Duringthe study none of them received other forms of therapy or medical interventions that couldaffect hand function.

Before installing the systems in the subjects’ homes, they practice donning regularwinter gloves. Two of the three children achieved the ability to don gloves withoutassistance.

Subjects were asked to exercise their plegic hand with the system for 30 minutes aday, five days a week. The subjects exercised their plegic hand wearing the sensingglove to play the games. The difficulty level was customized to the subject. Each subjectopened and closed his/her hand and flexed/extended his/her thumb as well as he/shecould at the beginning of an exercise session. Then subjects had to repeatedly reachtheir own maximum movement to earn points in games.

32


Figure 2.17: The average rate of targeted neuromotor movements. Error bars representthe maximum and minimum rates among the five participants.

Figure 2.18: Summary statistics of test sessions.

The time of the exercises and the changes in finger range of motion were remotelymonitored. One of the participants stayed in the study for six months. The other twosubjects remained a total of eleven months.

Subjects were evaluated at the beginning of the study and after 3 months ofintervention using several standardized occupational therapy instruments: grip strengthusing a dynamometer16, pinch strength using a pinchometer17, and general hand functionusing the Jebsen and Bruininks-Oseretsky tests of hand function [18] [19]. Subjectsalso received evaluations of forearm bone health using a DEXA scan18 and peripheralquantitative CT scan19. In addition, changes of range of motion measured by glovesensors were collected remotely throughout the study. One patient received a third set ofoccupational therapy and forearm bone health evaluations at 10 months.

In-home telerehabilitation increases the accessibility to rehabilitation, and has thepotential to become inexpensive with large scale production. The research team found

16Dynamometer. A device for measuring force, moment of force, or power17Pinchometer. A type of dynamometer that measures the strength of a finger pinch.18DEXA scan. Typically used to diagnose and follow osteoporosis.19CT scan. A technique that uses X-rays to obtain radiographic.

33


Figure 2.19: Home telerehabilitation system.

that regular scheduled videoconferences and additional phone and email meetings wereinvaluable and catalyzed timely problem solving. Despite this, the research team believesthat home telerehabilitation is the future of rehabilitation.

2.2 The Wii in rehabilitation systems

Video games enhance hand-eye coordination and muscle movement, the things thatchildren with cerebral palsy work in for hours in physical and occupational therapy. Videogames are more common activities for children. The brain reacts differently to fun waysof learning. The continuous play of video games helps children with cerebral palsy usetheir eyes, hands and brains together as a treatment for brain damage. The real benefitis that, while it is difficult for children and their families to follow through with conventionalprescribed exercises at home between therapy appointments, the games are fun so thechildren will continue to do them. In addition, they can play them with their friends.

Children with hemiplegic cerebral palsy predominantly use one side. Video gamesencourage them to use both hands. Games have been developed for Nintendo Wii andfor personal computer use that require the child to hold down a button with the dominanthand while using the less preferred hand to manipulate the controls.

34


Figure 2.20: Sample game screens used in home telerehabilitation.

2.2.1 Used for therapy

In 2007, Elaine Pearson and Chris Bailey from University of Teesside (United Kingdom)researched about the accessibility of the game consoles and control devices, theinteraction afforded with the games and the social aspect of using the Nintendo Wii indisabled students’ education [30].

The most distinguishing feature of the Nintendo Wii is the wireless controller (calledWii Remote [33]). This controller makes motion sensitivity more intuitive due its sensorable to detect motion and rotation in three dimensions.

In this study, the authors aimed to identify the types of game and elements withinthose games that provide support to disabled students in achieving the stated goalsof the games. Additionally, they explored the physical requirements of games usingthe Nintendo Wii and which people with physical disabilities are able to carry out theseactions.

To select the appropriated games they divided them into three groups:

• Simulation of real life activities in a fun and safe setting: games that can support in

35


developing transferable skills in a safe environment (Cooking Mama, Wii Play).

• Online team game: the main feature of these games is the developmentof collaboration, communication and group decision (Mario Strikers: ChargedFootball, Wii Sports).

• Adventure game: this kind of games focus with experiences that players might nototherwise have such as flying and support collaborative two player mode (WingIsland).

To evaluate the accessibility of the input device of the console and the potential of theWii for developing key skills, the researchers based on children playing observation. Theresults of these observations allow them to produce an outline specification for gamesthat can be used in an educational context and the ways that these games might be usedfor therapeutic purposes, for simulation of real life activities in a safe environment, andfor supporting the development of key skills such as collaboration, communication andcoordination.

In 2008, Judith E. Deutsch, Megan Borbely, Jenny Filler, Karen Huhn and PhyllisGuarrera-Bowlby studied the feasibility and outcomes of using a low-cost, commerciallyavailable gaming system (Wii) to augment the rehabilitation of patients with cerebral palsy[28].

No specified software was used in this study. They used the Wii Sports gamessoftware based on several factors:

• This type of system encourages movements that can be performed in both sittingand standing positions.

• There are stock games available in the system that can be selected based onpatient interest and task requirements.

• It provides users with knowledge of performance KP20 and knowledge of resultsKR21.

• They were interested in probing how interaction with other users22 would bereceived by the patient.

An adolescent with spastic diplegic cerebral palsy classified as GMFCS23 levelIII [31] participated in this trial. He had difficulty maintaining task focus, was easilydistractible and used his left upper extremity as his dominant functional arm. Posturein supported standing was described as semicrouched, with hip medial (internal) rotation

20Knowledge of performance. Information about the kinematics of the movement.21Knowledge of results. Information about the outcome of the movement which have been shown to

improve performance and skill in children with cerebral palsy.22Wii sports allow multiple users.23Gross Motor Function Classification System

36


and adduction. The sitting posture included a posteriorly tilted pelvis, laterally flexed trunkwith convexity to the right and flexed posturing of the right upper extremity, and decreasedloading through both lower extremities.

During the treatment the patient was attending a summer program at a school forchildren with developmental disabilities and also continued receiving his physical therapyfor 3 times a week and his occupational therapy for 2 times a week.

A total of 11 training sessions were carried out over 4 weeks. Sessions lasted from60 to 90 minutes and consisted of playing individual games ranged from 5 seconds to 5minutes based on the patient’s availability.

In the first 7 sessions, the patient used the system by himself. In the next 3 sessions,a child who was developing typically played with him. Finally, in the last session, thepatient experienced with 2 and 3 multiplayer games.

The different activities he played were tennis, boxing, bowling, baseball and golf inboth modes (game and training) and in both positions (sitting and standing). For gamesplayed in a sitting position, a therapist took care from behind, occasionally stabilizing thechair. For games played in a standing position, the therapist took care from behind or onthe side.

Activities were stopped or modified when the therapists observed deterioration in thepatient’s physical performance, technique, or postural control resulting from overexertion.

Three main outcome measures were used:

• Visual-perceptual processing was tested (Table 2.1) because it is know to beimpaired in children with cerebral palsy. It was tested using a motor-free perceptualtest (Test of Visual Perceptual Skills, third edition TVP-324) by an occupationaltherapist.

• Postural control was examined (Table ??) using measures of weight distributionand sway collected during static stance with eyes open and with eyes closed on aPosture Scale Analyzer25 (PSA). This system consists of 4 scales that are joinedas a single plate on which the patient stands. A computer connected to the systemcalculates weight distribution, center of pressure, and sway rates. Measurementsfor each condition were taken 3 times for 10 seconds. The validity of the weightmeasurements was established, using a scale, on 25 subjects who were healthy.

• Functional mobility measurements were extracted using retrospective data from thechart review. This method was selected to have repeated measurements relative tothe patient’s therapy goals by the same rater over time.

24TVP-3. It is a nonmotor assessment of visual perception for individuals 4 to 18 years of age. Itcomprises 7 subtests arranged in order of difficulty from least difficult to most difficult. It includes categoriesof perception such as spatial relationships and figure-ground that are particularly pertinent to integratingperception and action.

25Midot Medical Technology, Irus 276, Gan Ner, 19351 Israel

37


Table 2.1: Preintervention and Postintervention test of TVPS-3 scores and WISC-IV [32]Classifications.

Table 2.2: Preintervention and Postintervention Measurements on the Posture ScaleAnalyzer (PSA).

Figure 2.21 illustrates the distribution of time in both positions (standing and sitting)while Figure 2.22 shows the distribution of games for each week.

Figure 2.21: Training distribution by position.

Outcomes were assessed at different times after training. A repeat TVPS-3 wasadministered by the occupational therapist approximately 1 month after the training. Thepostural control measures were retested 1 day after training, and walking was evaluatedby the treating therapist during the training and approximately 3 months after the trainingended.

Visual-perceptual processing improved in all domains except visual memory. Posturalcontrol improved in a variety of measures. Centre of pressure sway decreased by

38


Figure 2.22: Distribution of games for each week.

approximately 60% in both eyes-open and eyes-closed conditions. Functional mobility,measured by the physical therapist, increased during the training and continued toincrease after training. The distance acquired had never been achieved or maintainedbefore training.

With the incorporation of the second player in the 10th session, the therapistsobserved turn taking, strategy sharing, and encouragement. This produced animprovement of patient’s score.

The authors assert that introducing the Wii for this student was feasible. The multiple-player capability of the system facilitated social interaction and unexpected therapeuticbenefits. He had increased his attention and was motivated to play games. Moreover,he exceeded duration on task relative to his therapy sessions. It also helped to hisimprovements in performance as well as his learning.

Another study that uses video games was performed in 2008. Professor Janet Eyrefrom Newcastle University’s Institute of Neuroscience carried out a research projectto encourage children suffers from hemiplegia to play customized computer games, inorder to improve muscle movement, hand-eye coordination and children’s control of theirdisabled arm [29].

Children with hemiplegia find it difficult to use what J. Eyre calls their ’never learntto use arm’. The children find it very hard to move this arm so it gets stiffer and stiffer.Around one in 1.000 children have hemiplegic cerebral palsy, affecting one arm and oneleg on the same side of the body.

The Eyre’s system consists of a laptop and Nintendo Wii controllers. The games wereprogrammed to run on a computer and its design was adapted to suit the patients specificneeds, which are less fast-paced and complex than commercial versions. The graphicswere also quite simple to follow.

An important aspect to consider was that the exercises must be fun and verycompetitive for children so they would not get bored. A total of two games were developedusing funding from the Children’s Foundation charity. These games mean children hadto use two hands cooperatively. One of the games involved exploding balloons, with one

39


hand targeting while the other hand fires the gun.

A total of ten hemiplegic cerebral palsy children participated in this trial. The patientsfamilies, also took part to create a more favorable environment for the children.

During this therapy, the therapists tried to give them an incentive so that they will usethe disabled arm a lot. After three months playing with games, J. Eyre observed a bigimprovement in arm function and important enhance of hand-eye coordination.

Patients get to the stage where, without thinking about it, they began to use theaffected hand to play correctly. As a result, the children began to use their disabledarm more in everyday life. Children sometimes feel stigmatized by therapy but everyoneplays games, and they can play them with their parents or their brothers and sisters.

A year after, in 2009, Jose Luis Gonzalez, Marcelino J. Cabrera, and Francisco L.Gutierrez developed Sc@ut DS [15] (Figure 2.23). This project is an evolution of theprevious project Sc@ut [14].

The researchers observed a series of deficiencies in PDA’s during the use of thiscommunicator:

• Fragility: The touch screen is very delicate, getting scratched or broken easily.

• Autonomy: About 3 hours. The battery requires replacing several times a day, beingthis aspect a great shortcoming for children and caregivers.

• Multimedia: The main memory is limited and consequently also the multimediaobjects and user interfaces which make them not attractive to children.

• Price: High, around 300 euros, being a negative factor for parents and caregivers.

The main goal of this project was to obtain a new attractive platform, less limitedthan Pocket PCs for children and to change the communicator’s philosophy to develop alearning tool using game concepts. To develop this new communicator, they selected theNintendo DS console instead of other alternatives such as PSP (Playstation Portable) orUMPCs (Ultra Mobile PCs), because they think it has the following advantages:

• Feedback: It has two screens. One of them is a touch screen which offers moreinteraction possibilities.

• Multimedia: It’s a game device with great multimedia options (sound, video andgraphics) without apparent limitations of memory.

• Autonomy: Approximately 11 hours.

• Connectivity: Wireless (Wi-Fi), to communicate with other DSs and PCs.

• Price: About 150 euros, much more affordable than a PDA device.

• Durability: It is a device designed to be used by children.

40


• Motivation: Other commercial games can be used as a complement, gift o rewardfor children.

Figure 2.23: Sc@ut DS communicator.

The goal of this project was that children learn how to associate letters, words orsentences to concepts. These concepts could be pictograms or videos that represent anaction, with the correspondent oral pronunciation [16].

A hero, ’Leoncio, The Lion’ was the mediator in the learning process and the one whocarried out the actions following the pictograms’ path. He was very expressive and helpsto obtain a positive action assimilation. In all games, the players follow a series of rules tosolve a problem individually, or collaboratively, against other players or the machine. Themain factor in a game is motivation. Thanks to motivation, players can be happy whenthey play.

Along with the correct stimuli, erroneous stimuli were introduced where the child mustselect to obtain the correct result. If the child chooses the right option, a positive feedbackis shown (animation or video). But if the child fails, the right concept is remarked.

Finally, the game’s results were sent to the teacher as a complete report using thewireless connection of Nintenfo DS. This report could be used afterwards to gatherinformation of several aspects of the learning process of children.

An improvement of the system to overcome the small size of the touch screen, wasto use a Wii remote control (Wiimote [33]) as a new interaction mechanism (Figure2.24). This control detects the user’s movements and position in 3D space similarly to athree dimensional mouse. With this control, the games were adapted to be used with atelevision or PC screen.

After several tests with children, the authors concluded that videogames in specialeducational offer very interesting results:

• Better spatial, temporary and hand-view coordination ability.

• Better concentration, motivation, attention and reasoning capability.

41


Figure 2.24: Alternative interaction in Sc@ut DS using Wiimote.

• Better assurance in the learning process. Children can repeat a task until theydominate it.

• Better assimilation of strategies and consequences in determinate actions.

• Children are happy playing and learning. This improves the social relationships withother children and the environment.

2.2.2 Center of pressure measurement

Many people with cerebral palsy experience disturbances in posture and balance, whichcontribute to overall motor dysfunction. The cerebral palsied tend to have abnormal sittingand standing posture, and demonstrate poor balance compared to healthy individualsof the same stage of development. For example, some children with cerebral palsyhave trouble adjusting their posture and sitting independently. Both postural aberrationsand poor balance contribute to difficulties with ambulation. Most postural and balanceproblems arise from muscle abnormalities, but some forms of physical therapy appear tohelp with postural control and balance in the cerebral palsy patient.

Standing balance is an important skill that can be significantly impaired in many healthconditions. That is why there are a lot of assessment protocols to evaluate the impairedstanding balance.

Laboratory-based assessment have identified important outcome measures using thecenter of pressure (COP) recorded from a force platform. Such system provides usefulinformation, however they are not a feasible assessment in a clinical setting due to theircost, one difficulty to setup and cumbersomeness to transport.

Subjective assessment tools, such as Berg Balance Scale [34] (a test developed todetermine the ability of the elderly to keep their balance), are commonly used because itdoes not require specialized equipment and are more clinically applicable. Unfortunately,these protocols suffer from limitations such as limited precision to detect small changes

42


in performance. In addition, previous research indicates that the relationship betweenscores on subjective tests and measures of centre of pressure displacement is onlymoderate.

This highlights the need to create a portable, inexpensive balance assessment systemthat has widespread availability. The Wii Balance Board (WBB) [35], satisfies all thesecriteria. The WBB possesses similar characteristics to a laboratory-grade force platformin that it contains four transducers which are used to assess the forces distribution andthe resultant movements in centre of pressure.

Given the capacity for providing instant feedback and the potential for enhancedmotivation levels, this system has already been integrated into the rehabilitation programsof neurological patients with balance problems.

In addition to its use as a biofeedback and gaming tool, the WBB could potentiallybe used by clinicians to collect and analyse laboratory-grade balance data using thetechniques and outcome measures most specific to the patients population of interest.The WBB is a small fraction of the cost of a laboratory-grade force platform, it is mass-marketed and portable, and consequently it has the potential to become a key componentof a clinician’s testing battery if it can be shown to produce reliable results.

Ross A. Clark, Adam L. Bryant, Yonghao Pua, Paul McCrory, Kim Bennell and MichaelHunt aimed to compare centre of pressure data collected on a WBB with that of alaboratory-grade force platform [36]. A total of thirty young (10 males and 20 females)between 18 to 29 years without any limb pathology participated in this trial. A test wasperformed on two occasions, completed within 2 weeks and at least 24 hours apart.Each test consisted in a three successful series of four different standing balance taskson a laboratory-grade force platform26 calibrated in accordance with the manufacturer’srecommendations and on a WBB located beside the force platform. The WBB wasinterfaced with a laptop using customized software27, and was calibrated by placing avariety of known loads at different positions on the WBB.

The balance standing tasks were:

• Single limb standing (on the dominant limb) with eyes closed.

• Single limb standing with eyes open.

• Double limb standing with eyes closed and feet together.

• Double limb standing with eyes open and feet a comfortable distance apart.

Single limb trials lasted 10 seconds while double limb trials lasted 30 seconds. Theparticipants keep their hands placed on their lips and remained as still as possible.

26AMTI Model OR6-5, Watertown, MA, U.S.A.27Labview 8.5 National Instruments, Austin, TX, U.S.A.

43


Both devices were sampled at 40 Hz and filtered using an eighth order Butterworthfilter with a lowpass cut-off frequency of 12 Hz due to the high frequency noisecontamination observed during a preliminary inspection.

The outcome measure used in this trial was the total COP path length. Based on themedian of the three repetitions, a single value for each of the outcome measures wasobtained for each of the four tasks, device (force platform or WBB), and test occasion(day 1 or day 2). The test was performed on the data of the 30 participants in all testsexcept the single limb eyes closed trial, which was limited to the data of 28 participantsbecause two participants were unable to successfully complete three trials.

To evaluate the differences between the two devices a Bland-Altman plot28 for theCOP was realized for each protocol. Further, intraclass correlation coefficients (ICC),standard error of measurement (SEM) and minimum detectable change (MDC) valueswere calculated to assess reliability and concurrent validity between the WBB and theforce platform. All values were reasonably high for both devices, with the WBB MDCvalues higher than the force platform values in three of the four trials.

Table 2.3: Reliability and concurrent validity analysis of COP path length (cm) measuresduring each of the four standing balance trials. FP: force plate; WBB: Wii BalanceBoard; COP: center of pressure; AP: anteroposterior; ML: mediolateral; SD: standarddeviation; CI: confidence interval; ICC: intraclass correlation coefficient; Diff: difference;SEM: standard error of the measurement; MDC: minimum detectable change, expressedas a percentage of the Day 1 mean value.

Both devices showed excellent COP test for the single limb with eyes open, singlelimb with eyes closed and double limb with eyes closed and feet together tests (Table2.3). Only the double limb standing with eyes open and feet apart (ICC = 0.66) obtained

28Bland-Altman plot. http://en.wikipedia.org/wiki/Bland-Altman plot

44


an ICC value smaller than 0.75, considered the threshold of excellent range.

There can be differences comparing the results between the two devices. This is dueto the device specific factors such as the precision and sensitivity of the sensors, but theconsistency of the within-device results indicates that this would not have an effect oncomparisons performed on patients using the same device.

However, the WBB has some limitations that prevents it from being a directreplacement for a force platform in activities that require rapid or high force movementssuch as jumping or running. Also, a limitation of more importance in the assessmentof balance is the inability to assess force in the horizontal axes, which are importantcomponents of the standard COP equations. This limitation would have a significantimpact on the results of balance tests. The authors concluded that the WBB is aportable, inexpensive and valid system that could provide numerous benefits in clinicalfield because it provides supplementary balance information that is not discernible usingvisual assessment alone. It possesses excellent reliability for COP assessment andconcurrent validity with a laboratory-grade force platform.

A standing posture detector was designed in 2009. After several studies focusedon redesigning the mouse driver to enable people with multiple disabilities to controlenvironmental stimulation, Ching-Hsiang Shih, Ching-Tien Shih and Ming-Shan Chiangfocused on body swing [37]. In this case, they studied the changing standing posture(CSP) to control environmental stimulation. Therefore, a Wii Balance Board can be usedas a precise device to detect any CSP.

Two patients participated in the study:

• The first one, Lin was an 11 years old boy. He was rated in the severe intellectualdisability range. He was diagnosed as having congenital cerebropathy withmoderate spastic quadriplegia. He had greater mobility on his left side and onlycould stand independently for less than 7 minutes, and in abnormal posture. Duringthe test, he received both physical and occupational therapy as part of his daily life.

• The second one, Su was a 9 years old boy. He was also diagnosed with moderatespastic quadriplegia. However, he was rated in the moderate intellectual disabilityrange. Despite this, he was able to stand independently for greater than 5 minutes,but exhibiting poor posture.

Figure 2.25 shows the configuration of the study. The WBB detected the changes ofstanding posture of patients and transmitted them via Bluetooth. A Bluetooth dongle wasconnected to a minicomputer to establish a connection with the WBB. A CSP detectionprogram (CSPDP29) was developed to transfer the pressure values of the WBB sensorsinto a sequence data of CSP. The minicomputer was connected to TV though cables tobroadcast the participant’s favorite videos.

It was necessary to set a critical value for the participant’s CSP amount, because anyslight but undesired CSP could be detected by the WBB. The control system CSPDP

29CSPDP. Change of Standing Posture Detection Program

45


Figure 2.25: Scheme of the configuration of this study.

would check the CSP amount, and start a 6 seconds stimulation period contingent on thetarget response (except in baseline phases) and to record the response, once the CSPamount exceeded the critical value. The critical value can be adjusted according to theparticipants’ personal body conditions. The stimulation period involved the activation ofthe favorite stimulus which consisted of various types of cartoon videos offered by hisparents.

Both participants received an ABAB sequence, in which A represented baseline andB represented intervention phases. The baseline phases included 18 and 15 sessions,respectively. At the beginning of the sessions, participants received promptings froma therapist to ask to swing their bodies from left to right in order to change theirstanding posture. The Wii Balance Board and the control system were available, butany stimulation was produced by the control system.

Intervention phases consisted of 66 and 84 sessions, respectively. Proceduralconditions were as during the baseline phase, but the control system produced 6 secondsof patients’ favorite stimulation.

From three to five sessions of about 5 min a day were carried out. The sessionswere conducted at home and responses were recorded automatically through the minicomputer.

Figure 2.26 shows the first participant results. He had a mean of about 6 independent

46


responses per session in the first baseline phase. This mean increased to about 19.23responses per session during the first intervention phase. During the second baselinephase, the mean frequency dropped to 10 to be fully restored and eventually increasedduring the second intervention phase.

Figure 2.26: Lin’s data.

During the first baseline phase, the second participant results (Figure 2.27) had amean of about 7.17 independent responses per session. In the first intervention phase,this mean increased to about 25 responses per session. The mean frequency droppedto 11.8 during the second baseline phase to be fully restored and eventually increasedduring the second intervention phase.

Figure 2.27: Su’s data.

The difference between the baseline and the intervention responding frequency wassignificant (p < 0.01) on the Kolmogorov-Smirnov test30 for both participants.

The WBB suitable to detect simple behavioral acts such as standing posture changeenabled the two boys to improve their levels of responding and stimulation control. Thisachievement could be very important to improve patients’ quality of life. Moreover, theWBB is a commercial product that has many advantages, such as low cost, easy toobtain, good technical support, and easily updated with the newest technology.

30Kolmogorov-Smirnov test. http://en.wikipedia.org/wiki/Kolmogorov-Smirnov test

47


Therefore, in addition to the exclusive use of specialized alternative devices, peoplewith disabilities would also benefit from being trained to use very common, cheap andpowerful commercial products, if these commercial products can be modified throughadditional design or modification into assistive devices to match their special needs.

Moreover, software technology has more advantages over hardware modification toreset the functions of commercial products, turning them into a much more powerful tool,especially in the fields concerning people with disabilities. The authors of this studybelieve that further studies are necessary to extend this study applicable to posturecorrection, develop new software technology, and apply more commercial products tofields concerning people with disabilities.

48

Chapter 3

Our method

The use of technology in motor rehabilitation for people with cognitive impairments canpromote their interest in practice the therapy, raise diversity of therapy methods andimprove the efficiency of the treatment.

After the study of the state of the art about technology in rehabilitation systemsfor children with cerebral palsy, we concluded that there is the need to develop moreexclusively software for therapy systems. Most studies use commercial software andvideo games. However, with commercial software the therapy is not as profitable as itcould be.

Nowadays, the therapists still use traditional methods to the therapy. This methodsare increasingly combined with technology to improve the performance of the exercisesand to make it more comfortable for both the patient and the therapist. Moreover,technological systems are able to show the results in form of statistics and to maintain anhistory of every patient automatically.

In March 2010, we had a meeting with the therapists of Associacio Provincial deParalisi Cerebral de Tarragona (APPC) [8] to know what are their needs in terms ofrehabilitation.

Since 2007, they are using innovating systems in therapy with cerebral palsiedchildren through the SATI project [6]. This has resulted in the improvement of patients’quality of life. They were more motivated and therefore, they practiced the exercises forlonger.

However, the physiotherapists told us that they not have any mechanism to detectthe improvements in postural balance of cerebral palsied children. The only way to keeptrack of a patient, is annotating the comments of the therapist and comparing them. Thismethod is not reliable because it depends on the personal opinion of the therapist. Inother words, the same session, seen by different therapists may have different results.That is why we focused on the design of an application able to measure these changesin body balance.

49


The aim for this application is that it can be used in therapeutic and familiarenvironment. It must enable both physiotherapists and familiars of children with cerebralpalsy to use this tool. Therefore, it must be a low-cost system and both the hardware andthe software must be easy to obtain.

Another important aspect of this project is the portability. The system must be easyto move from one place to another and easy to assemble. In addition, this system isdesigned for people with disabilities. For that reason, this software must be easy to use.

Children with cerebral palsy have different levels of disability. There are children whocan be for a long time stand up while others just can be a few seconds stand up in anabnormal posture. The system must be capable to evaluate data regardless the durationof the session.

Additionally, the system should have an option to automatically detect which user isperforming the test, obviously, if the user has previously participated. The physiotherapistwill decide if he wants to enable or disable this option.

The Wii Balance Board [35] is the most appropriate device to form part of this project.It consists of a table capable to calculate the pressure exerted on it. It has four pressuresensors situated at each corner from which enough information is available to be ableto get calibrated readings. The sensors give different pressure values when a user’sstanding posture changes. The user’s center of gravity can be calculated from analyzingthis pressure values transmitted via Bluetooth.

During the meeting, we conduct a preliminary study on the operation of the systemconsidering the use of the Wii Balance Board.

Data collected

The tool is based on the analysis of the antero-posterior and lateral balance distribution.In a test, the system must obtain the following information:

• User’s information: name, date of birth and date of test.

• Typology: standing, kneeling, sitting.

• Positioning reference according to the typology:

Standing: separation of the legs and antero-posterior reference.

Kneeling: position of the knees.

Sitting: how the user sits.

• Support tools according to the typology.

50


Scenario of an assessment test

In the beginning of the test procedure, the physiotherapist helps the user to go up to theWii Balance Board. He aligns the reference marks and holds the user to prevent a fall.The physiotherapist annotates the reference and leaves the user. The system measureshow long the user stays on the Wii Balance Board and stores the balance measures.

Figure 3.1: Scheme of a session.

Figure 3.1 shows the scheme of a session:

• ti : represents preparation time (when the physiotherapist helps the user to go up).In this case the system do not have to take measures.

• e1: mouse click or space bar to start the test.

• te: equilibration time. Until the user finds the balance.

• e2: the physiotherapist indicates the end of te.

• tm: measurement time. The physiotherapist helps the user verbally to maintainhis/her balance.

• e3: end of the test. It is produced when the user needs help, the user gets down orthe physiotherapist decides to end the test.

• ts: noise time. When the user gets down. In this case the system do not have totake measures.

• e4: The Wii Balance Board is empty.

Furthermore, during ti, te and tm the system has to show in real time:

• Graphic center of gravity.

• Numeric center of gravity (antero-posterior and lateral balance).

In te analysis:

• The time of this step.

• Variance (antero-posterior and lateral).

51


In the te phase:

• User’s weight.

• Average (antero-posterior and lateral).

• Average deviation (antero-posterior and lateral).

At the end of the test the system should display a report showing this information andtwo graphics with antero-posterior and lateral balances.

3.1 Design

The configuration of this study is shown in the Figure 3.2. A Wii Balance Board [35]is placed under the child’s feet to detect their body balance. The data captured by thebalance board is transmitted via Bluetooth wirelessly to a laptop with a Bluetooth USBadapter incorporated. This laptop has the customized software installed to convert thedata received in the appropriate statistics.

Figure 3.2: Scheme of the configuration of this study. The Wii Balance Board transmitsthe child’s body balance to the laptop. The therapist can obtain these data in form ofstatistics.

52


3.1.1 Structure

A MySQL1 database has been used to store and query information about users and tests.

The application has been implemented in Java and utilizes the WiiBBToOSC2

software and the OSC protocol3 to capture the data from the Wii Balance Board.

A total of three different sections compose the application (initial window, testwindow and statistics window). Through the buttons at the bottom of the screen, thephysiotherapist can go from one section to another.

Initial window

When the application starts, it appears this window (Figure 3.3). It is exclusively designedfor the therapist or familiar of the cerebral palsied child. It allows to select or create a userand choose an action (view the user’s statistics or start the test).

Figure 3.3: Initial window.

Moreover, the therapist can choose the autodetection mode to detect the userautomatically when he or she performs the test. Obviously, this user should be registeredinto the database.

1MySQL. http://en.wikipedia.org/wiki/MySQL2WiiBBToOsc. Read the information from the sensors of the Wii Balance Board and sends it via the OSC

protocol.3OSC protocol. http://en.wikipedia.org/wiki/Open Sound Control

53


In this window and for each test, the therapist must specify the typology of the testand information about the support tools, if exists.

Test window

The test window (Figure 3.4) is designed for both the therapist and the cerebral palsiedchild. It contains the zone of interaction between the child and the application.

Figure 3.4: Test window.

This window is divided into two parts. In the first one, at left of the screen, the centerof gravity is shown through a real-time graph. This area can be used to capture user’sattention because the movement of the user produces an action-reaction in the movementof the square.

The second one, at right of the screen, contains information of the evolution of thetest and it is designed for therapist and familiars of the user.

Statistics window

This window, exclusively for therapist or familiars of the users, allow them to observeinformation and graphics of a specific test. With a combobox, it is possible to navigate forall of tests carried out by the user selected. The tests are ordered by date.

Statistics window is a very important area because it enables follow the evolution of a

54


user through the statistics of all his tests. Through this window the therapist can check theeffectiveness of the rehabilitation applied to the user and can demonstrate in a graphicalmode the improvement of the patient’s motor rehabilitation.

Figure 3.5: Statistics window.

As shown in Figure 3.5, in the upper left of the window there is the informationabout the user. Below this, appears the date of the test and the data introduced bythe physiotherapist as support tools and some observations. In the bottom left, there is ascattergram that represents the user’s center of gravity. In the upper right of the window,two graphs represent the variance in antero-posterior and lateral axis of the equilibrationphase. Finally, in the bottom right, there are other two graphs that represent the antero-posterior and lateral balance in the measurement phase.

3.1.2 Test phases

When a user carries out a test, he or she performs a total of three different phases(training, equilibration and measurement).

In the statistics window, equilibration and measurement phases are shown separately.Each phase has two different graphs, the first one for antero-posterior axis and the secondone for lateral axis.

55


Training phase

This phase is used to up the user on the Wii Balance Board and it can be also used tofamiliarize the user with the system. Only the numeric and graphic center of gravity inantero-posterior and lateral axis are shown. The data received by the Wii Balance Boardis not stored in this phase.

Equilibration phase

From this phase, the information transmitted by the Wii Balance Board is stored in thedatabase. It starts when the users is on the Wii Balance Board but he is not still rightbalanced.

Right balanced is considered when the user adopts a posture which can maintainalmost motionless. When the therapist observe this pattern of balance goes to the nextstep.

In this phase, the system shows numeric and graphic center of gravity, an stopwatchthat indicates the time of this phase and antero-posterior and lateral variance.

Due to this data must be shown in real-time, the variance is also calculated in real-time. To obtain this variance, we utilize a moving average that is updated with the datareceived since the start of the equilibration phase and so far.

Measurement phase

When the patient achieves to be balanced starts the measurement phase. In this case,the purpose is that the child be motionless or follows the physiotherapist instructions.This phase is also stored in the database and should ends when the user looses theequilibrium, the user falls or the physiotherapist decides to end the test.

In this phase, the numeric and graphic center of gravity are still in operation, but theantero-posterior and lateral variance and the equilibration stopwatch remains with thelast values of equilibration phase. Moreover, the system shows another stopwatch whichindicates the time of this new phase, the weight of the patient and the balance averagefor antero-posterior and lateral axis.

It bears mentioning that the weight of the user is calculated using all values taken inthe measurement phase. This means that we calculate the average of the weight in thisphase. With this method, if the user moves on the Wii and distorts the values, these havenot enough influenced on the calculated weight.

When the physiotherapist presses the button to finalize the test, the stopwatch ofmeasurement phase, the weight, and the antero-posterior and lateral balance averagesstop. The real-time graph also stops and the square indicates the average of the centerof gravity. Pressing the button again, the test is stored in the database.

56


3.2 Results

In this section, we explain all the experimental tests realized in the Associacio Provincialde Paralisi Cerebral de Tarragona (APPC) [8], a brief description of each patient thatparticipated in the study, and finally, all the results obtained using the Wii Balance Boardand the developed software.

3.2.1 Participants

The selection of the children to perform the experimental tests for this study was carriedout by the physiotherapists of the Associacio Provincial de Paralisi Cerebral de Tarragona(APPC) [8]. They based on the patients’ capabilities to select them.

A total of three users were selected to participate in this study. The first one, Fatima,is a 18 years old girl. Her weight is 103 Kg. and, despite of her cerebral palsiedcondition, she can maintain her standing position for a long time. Furthermore, she cansit independently.

The second patient who participated in this study was Mohamed. He is a 12 yearsold boy and his weight is 65.2 Kg. He can also maintain a standing position, but showsmore difficulty than Fatima.

Finally, the third child is Alba. She has got 10 years old and her weight is 25 Kg.She suffers spastic tetraparesis4. She has exaggerated stretch reflexes that causesher to respond to rapid passive stretching with vigorous muscle contractions. She cannot maintain a standing position more than few seconds and requires assistance to sitcorrectly.

3.2.2 Tests

Fatima performed the test two times. The first one in sitting position and the second onein standing position. Both tests were carried out without any support tool.

To perform the test in sitting position, she sat down on a stool and leant their feet onthe floor. Only a 75 percent of her weight was distributed to the Wii Balance Body. Theremainder 25 percent was distributed directly to the floor through their feet. During thetrial, the therapist observed too much control of her movement because their feet weresupported.

In equilibration phase, the graphs of variance5 (Figure 3.6) show how she has nodifficulties in sitting on the Wii Balance Board and maintain the equilibrium. There is notpractically movement.

4Spastic tetraparesis. A type of cerebral palsy that affects muscle tone of the four limbs.5The variance is calculated in real time using a moving average that is updated in each moment of the

phase.

57


Figure 3.6: Variance of motion in equilibration phase of Fatima’s sitting test.

After 12 seconds starts the measurement phase (Figure 3.7). In this step, the patientwas asked to move her body on the Wii Balance Body. During this movement, Fatimadisplaced her body forwards and right but she could not move her body backwards andleft more than the equilibration position. That’s the reason why the averages in antero-posterior and lateral balance are displaced, 5% forward and 3% at right, in spite of hercapability of maintain her body equilibrate in the center of the 2D graphic.

Figure 3.7: Balance distribution in measurement phase of Fatima’s sitting test.

In consequence of this information, we can come to the conclusion that this patienthas trouble to move her body backwards and left when she is sitting. With a visuallyexploring of Fatima’s behavior, it is possible that we not become aware of this defect and,considering these conclusions, the physiotherapists can deduct the type of therapy to putinto practice.

Three minutes after of this first trial, a standing test was performed. In this case, shewore slippers.

Equilibration phase (Figure 3.8) lasted 8 seconds. During this time, she tried to adopta stable position. Lateral variance shows how she could maintain equilibrium, but antero-posterior graph indicates constant movement in this direction.

58


Figure 3.8: Variance of motion in equilibration phase of Fatima’s standing test.

Measurement phase lasted 1 minute. In this step, lateral balance was enough stableregardless of a hop in the graph. The physiotherapist annotated it was produced becauseFatima moved their feet (Figure 3.9).

In antero-posterior graph, appears something like a sine wave. This means thatFatima moved her body forward and backward to not looses her equilibrium.

Figure 3.9: Balance distribution in measurement phase of Fatima’s standing test.

Her antero-posterior balance average was 3 percent backward, while her lateralbalance average was completely centered.

To conclude, Fatima is capable to maintain an optimal center of gravity (centeredin the 2D graph) when she is standing, but she has trouble to maintain her equilibriummotionless in antero-posterior axis and this increases the risk of falls.

In the case of Mohamed, he performed two standing tests. The first one wearingslippers and the second one barefooted. Both tests were carried out without any supporttool.

The first test lasted a total of 16 seconds. Eight seconds in equilibration phase andalso eight seconds in measurement phase.

59


According to physiotherapist’s observations, Mohamed could maintains on the WiiBalance Board but he needed to move their feet so as not to leaves the platform.

In equilibration phase, the variance graph (Figure 3.10) shows how the amount oflateral movement is practically zero. In antero-posterior axis, there is a few movement butit is not excessive.

Figure 3.10: Variance of motion in equilibration phase of first Mohamed’s standing test.

Examining the measurement phase, it can be seen how the user moves a bit forwardbut rectifies his position. Moreover, in lateral direction he maintains the stability until theend of the test when he looses the equilibrium and move their feet.

His antero-posterior and lateral balance averages were practically centered, 1 percentbackward and 1 percent at left.

Figure 3.11: Balance distribution in measurement phase of first Mohamed’s standing test.

As per physiotherapist assessment, Mohamed seems to have more trouble tomaintain an equilibrated posture than Fatima. Nevertheless, comparing both statisticsseems that both patients have a similar behavior and approximately the same difficulties.

In the second standing test, the equilibration phase only lasted three seconds whilethe measurement phase lasted for 1 minute.

Contrary to the previous trial, the variance in antero-posterior axis in equilibration

60


phase was minimum. However, the lateral variance started with large amount of motionin this axis, but soon after was decreased and he got in a stable posture. He just need alittle more than a second to recover the stabilization (Figure 3.12).

Figure 3.12: Variance of motion in equilibration phase of second Mohamed’s standingtest.

Finally, the graphs corresponding to the measurement phase (Figure 3.13) shows howthe lateral motion is small, but he can not stay stabilized in antero-posterior axis.

Figure 3.13: Balance distribution in measurement phase of second Mohamed’s standingtest.

In the second test, it seems that Mohamed oscillates his body more than the first one.It can be due to different causes such as the test duration, the footwear or his own fatigue.

Nevertheless, comparing again with the statistics of Fatima, it seems that bothchildren keep on a similar pattern to maintain equilibrium on the Wii Balance Board.

In the case of Alba, she performed only a test. It was carried out in standing positionand she wore normal shoes.

Because her condition of spastic tetraparesis, she can not stand firm for a long time.That’s the reason why the physiotherapist decided focuses in the measurement phase,and equilibration phase almost lasted a second (Figure 3.14).

61


Figure 3.14: Variance of motion in equilibration phase of Alba’s standing test.

The measurement phase lasted for nine seconds (Figure 3.15). Alba started thisphase tilted to the back, but during the first four seconds, she struggled to achievean equilibrated center of gravity moving her body forward. As a consequence of thisdisplacement, she also produced a lateral oscillation. The last five seconds, she couldmaintain perfectly the center of gravity with correctly antero-posterior and lateral balancedistribution.

The antero-posterior balance average was 3 percent forward and lateral balanceaverage was totally centered.

Figure 3.15: Balance distribution in measurement phase of Alba’s standing test.

In this trial, the physiotherapist has had to go fast between each phase due the limitedtime of the test. In this example, the first four seconds of the measurement phase reallybelong to the equilibration phase because Alba is still trying to balance her body.

As the physiotherapist could not predict the child’s reaction, he changed to the mea-surement phase when he supposed that Alba was in her maximum stabilized position.Therefore, the same test with different physiotherapists or different physiotherapist’s re-action can produce different statistics. That’s the reason why it would be a good ideaconsider that the system was who decides when proceed to the next phase. With thisimprovement, the statistics would be less subjective.

62

Chapter 4

Analysis of the biometric detectionsystem

The aim of the biometric detection system in this project is provide more comfort to thephysiotherapist when he or she is working with his or her patients. The basic idea is thatwhen the physiotherapist selects the autodetection mode1, he or she does not have to beaware to select the user who is going to carry out the test. Moreover, this method makesthe process more agile and simplifies the physiotherapist interaction with the application.

The experimental tests with the children of Associacio Provincial de Paralisi Cerebralde Tarragona APPC [8] (Fatima, Mohamed and Alba) were the first trials carried out withthis software. It is why we could not analyze the biometric detection system with thembecause there were not any information in the database to compare. However, the studyof the data extracted in these trials, allowed us to develop useful statistics that aided usin the design of the biometric detection system.

The challenge of designing a biometric detection system in this area is not an easytask because has several issues to be considered. We are treating with cerebral palsychildren. This involves that different tests with the same child can produce differentstatistics. It is due several circumstances such as the patient disposition in the momentof the test, the duration of the test or the therapist instructions.

An important factor to consider is the patients’ evolution. Aside from these trials, all ofthem are receiving their usual therapy. This can produce an enhancing of their posturalcontrol that difficults the detection of a user in long term. Furthermore, children’s growthbring an increment of their weight that can cause balance modifications. Moreover,specific features such as typology, support tools or footwear have influence over the testresults.

In order to solve these problems, we have developed a keep up to day biometricdetection system. It means that when a user practices the test, the system updatesautomatically the user attributes with the newest statistics.

1In the initial window of the program.

63


To decrease balance differences between typologies, we have took into account thisattribute. That’s to say, when a user carries out a standing test, the other typologies testsin the database are ignored by the biometric detection system.

To do an initial selection of the tests which are candidates, first of all we filtrate thesein function of the users’ weight and their antero-posterior and lateral balance averages.Different thresholds for weight and balance have been adjusted to discard the tests whichget away from the user features.

For the candidates which accomplish these features, the system creates a matrix thatrepresents the surface of the Wii Balance Board. Each position of the matrix involves anarea of the Wii Balance Board2 and is set with the percentile of times the user has beenin its area of the Balance Board.

This matrix can be represented graphically as a scattergram (Figure 4.1). To comparetwo scattergrams, their correspondent matrix are subtracted. As minor result obtained,the pair of scattergrams are more similar.

Figure 4.1: Example of an scattergram. The total area represents the total surface of theWii Balance Board. In this case, the user is inclined backward and right.

To test the reliability of this biometric detection system, a group of 6 healthy userswere selected. Their healthy conditions difficults the autodetection task because theirbalance distribution were shapely and similar between them. Furthermore, two of themhave similar weight than Fatima, another two have similar weight than Mohamed and theother two users have similar weight between both them. Table 4.1 lists the users and theirweights.

User WeightCarles 90.3 Kg.Josep 89.6 Kg.Joan 69.3 Kg.Sara 62.2 Kg.Ana 53.3 Kg.

Maria 53.2 Kg.

Table 4.1: Users selected for autodetection test.

Two initial tests were performed to introduce the users in the system. The first one instanding position and the second one in sitting position. Next day, they repeat the sametests with the autodetection mode enabled.

2The size of this area depends on a threshold.

64


Each test consisted of a 30 seconds measurement phase. The equilibration phasewas ignored because it does not affect in the autodetection system.

On the one hand, the autodetection system obtained an effectiveness of 67% instanding tests. Only two of six tests failed detecting the user (Figure 4.2). In Josep’stest, the system was mistook with Carles, and in Ana’s test, the system was mistook withMaria.

1.

2.

3.

4.

5.

6.

Figure 4.2: Standing tests’ scattergrams. Images at left are initial test, images at rightare second test in autodetection mode. 1. Carles, 2. Josep, 3. Joan, 4. Sara, 5. Ana, 6.Maria.

These confusions ocurred for several reasons. First of all, the users have practicallythe same weights. Furthermore, body balance is similar in both anteroposterior andlateral axis. Finally, the scattergrams of the second test have more similitude with theusers that the system detects.

On the other hand, the same six users performed the initial test in the sitting position,but only five carried out the second one. Figure 4.3 shows the scattergrams of the fiveusers’ sitting tests.

65


1.

2.

3.

4.

5.

Figure 4.3: Sitting tests’ scattergrams. Images at left are initial test, images at right aresecond test in autodetection mode. 1. Josep, 2. Joan, 3. Sara, 4. Ana, 5. Maria.

This time, the autodetection system obtained an effectiveness of 80%. Maria’s testwas just wrong. It can be seen how her tests have not any similitude between them.

We became aware that all of the users leant between 2 and 17 percent of their weightson the floor through their feet except Maria, who leant the 82% of her weight on the floor.This means that the test was evaluated only with the 18% of her total weight.

The fact that users recline their feet on the floor may have an important influenceon the statistics. The reason why she reclined their feet on the floor considerably is theheight she was seated. With a low seat, the pressure of the feet on the floor is bigger.

That is why users should sit in a high chair to recline their body on the Wii BalanceBoard and not distort the results due the pressure of their feet on the floor.

The way to sit also has importance in the results. It is very difficult determine if theuser is sitting totally centered on the Wii Balance Board. The minimum displacement onthe platform may produce different statistics. Moreover, it is complicated that a user sitsin the same position all the tests. That is why the tests in the sitting position are morecomplicated to compare than the tests in the standing position.

In the kneeling position, the user reclines the toes on the floor. As in the sittingposition, this may affect the results of the tests in function of the pressure of the toes

66


on the floor. To reduce this distortion, the user should lean forward to transmit the mostweight possible to the knees.

Unfortunately, we could not check the biometric detection system in the kneelingposition. However, the results that we would obtain would be very similar to the results inthe sitting position due it also has a support out of the Wii Balance Board.

Taking into account the similitudes between the users selected in their weights,balances, and the scattergrams, we can consider satisfactory the results of the biometricdetection system. There is no autodetection test which was mistook with a cerebralpalsied user because their scattergrams tend to be more dispersed. We believe thatthe biometric detection system applied in children with cerebral palsy may increase itseffectiveness due to this greater scattering.

67

Chapter 5

Conclusions

There are no many research in the use of the Wii Balance Board in physiotherapy ofchildren with cerebral palsy and most of the proposals which use this balance board haveemployed the Nintendo commercial games to carry out the therapy. The short time thatthe Nintendo Wii is on sale and the complexity of the technology required to combine thishardware with customized software are the reasons for this absence of research in thisfield.

Until now, the physiotherapists could only diagnose the postural balance through theirpersonal point of view. This method is not reliable because the diagnose is based onthe therapist’s personal opinion. In other words, the same session, seen by differenttherapists may have different results. That is why, it is not an objective diagnose.

We have developed an application which allows therapists to diagnose the balancebody conditions of a patient based on the statistics of a test. This is a powerful tool fortherapists because of they can describe the postural balance of a child with a methodthat is not a simple personal opinion but a graphic form that converts the diagnosesobjectiveness.

This method was tested by physiotherapists of the Associacio Provincial de ParalisiCerebral de Tarragona (APPC) [8] and they are still using this software in the APPCcenter and in their own private offices.

The therapists are motivated with this project because they perceive an improvementin the way of working. They assure tha the cerebral palsied children are more involvedwith this task. Playing with the 2D graph, they can improve their levels of responding andstimulation control, and therefore their quality of life.

Though the application works properly, there are some aspects that can be improved.When the therapist is performing a test, he has to be aware of the patient’s reactionand with the interaction of the software. When he detects that the user is stabilized, thetherapist should press the button to change to the next step. In a way, the results dependon the therapist’s reaction.

68


The automatic phase step could improves the system, with the obtaining of resultseven more objective. Moreover, it facilitates the therapist task because of he only needsto start the process, then the system automatically changes to the next steps and finalizesthe test in the right time.

This enhancement is not trivial to develop because it requires an algorithm capable todetect if the user is going to fall or just has simply moved the feet to determine when thesystem has to change to next phase and to decide when the test must be finalized.

After study the results obtained in the children with cerebral palsy, we deduce thatanother important factor is if the user is watching the screen. When the user watches thereal-time graph, he tends to correct their body balance to situate the square in the middleof the graph.

In the case of use the application as a measurement tool, if the users correct theirbody balance, the statistics do not help to illustrate the users’ balance troubles. That iswhy, when the physiotherapist utilizes the application to measure the body balance of thepatient, he should not see the real-time graph.

Several studies have been proposed to control body balance in children withdisabilities. However, they are usually very costly, difficult to obtain or they need adedicated room.

This proposal only needs the Wii Balance Board and a common PC or laptop withBluetooth connection. This system offers many advantages such as low-cost, easy toobtain, and it can be used at any place.

The biometric detection system is very useful for the therapists because facilitates theinteraction with the application. The results obtained in this study are very satisfactory,however, its effectiveness can be improved considerably.

There is the need that the users be placed on the Wii Balance Board in the sameway all the times to not distort values. Also it is important to control the pressure that theuser applies out of the platform when he or she is sitting or kneeling. Moreover, usersvary postural skills over time, that makes more difficult the task of the biometric detectionsystem.

More research in this area is needed to improve the effectiveness of this system.Perhaps, with different thresholds according to the typology of the test and implementingan improved matching algorithm.

This study is only focused in children with cerebral palsy. Other people with differentmultiple disabilities are not addressed. Further research is necessary to extend thisproject in other collectives of people with disabilities.

69


Future work

Our proposal can adopt two different uses. The first one is as a measurement tool. Itallows physiotherapists have an statistical control of the distribution of the body balancein children with cerebral palsy.

The other functionality of this system is as a therapy tool. Using the real-time graph ofthe test window, the therapist can incentivize the patient to move the square into thedesired direction. With this, the patient practices a therapy focused in improves herpostural trouble.

To develop this second functionality, the idea is to convert the real-time graph in avideogame for patients. An image can appears into the graph in the desired position bythe therapist. The user should ’touch’ this image with the square leaning her body to thisdirection. To increase the patient’s attention, the therapist may select an stimulant imagefor the patient.

Until now, we have collected statistics of patients to check their postural limitationsusing the software as a measurement tool. Hereafter, we will employ this software as atherapy tool. As we know the defect of each patient, we can focus in correct this posturaltrouble.

After this therapy, another measurement test will be carried out to verify if the therapyhas had an improvement of patients’ postural balance.

70

Bibliography

[1] Michal Kuczynski, Karina Slonka. Influence of artificial saddle riding on posturalstability in children with cerebral palsy. Gait and Posture, volume 10, pages 154-160,October 1999.

[2] Fernandes Lorrain C., Chitra Jeba, Metgud Deepa, Khatri S.M. Effectiveness ofartificial horse riding on postural control in spastic diplegics. Indian Journal ofPhysiotherapy and Occupational Therapy. ISSN 0973-5674.

[3] Ching-Hsiang Shih, Ching-Tien Shih. Assisting people with multiple disabilities to usecomputers with multiple mice. Research in Developmental Disabilities, volume 30,pages 746-754. 2009.

[4] Ching-Hsiang Shih, Ching-Tien Shih. Development of a computer input system forpeople with disabilities using a commercial mouse and switches. Disability andRehabilitation: Assistive Technology, November 2009; 4(6): 414-421.

[5] Ching-Hsiang Shih, Ching-Tien Shih, Kun-Tsan Lin, Ming-Shan Chiang. Assistingpeople with multiple disabilities and minimal motor behavior to control environmentalstimulation through a mouse wheel. Research in Developmental Disabilities, volume30, pages 1413-1419. 2009.

[6] C. Mauri, A. Solanas, T. Granollers, J. Bages and M. Garcıa. Interactive TherapeuticMulti-sensory Environment for Cerebral Palsy People. Proceedings of the 12th IFIPTC 13 International Conference on Human-Computer Interaction: Part II. Uppsala,Sweden. Pages 696-699. 2009.

[7] APPC 2008 ICT Research and Application.z http://www.youtube.com/watch?v=98 chQxx4w

[8] Associacio Provincial de Paralisi Cerebral de Tarragona (APPC).http://www.appctarragona.org/

[9] CREA Sistemes Informatics. http://www.crea-si.com/

[10] Centre de recherche Informatique et Creation Musicale (CICM).http://cicm.mshparisnord.org/

[11] Denise T. Reid. Benefits of a virtual play rehabilitation environment for children withcerebral palsy on perceptions of self-efficacy: a pilot study. Pediatric Rehabilitation,2002, Vol. 5, No. 3, pages 141-148.

71


[12] William Li, Tom Chau, Sophie Lam-Damji, Darcy Fehlings. Usability of a virtual realitytherapy system for children with hemiplegic cerebral palsy.

[13] C. DeMatteo, M. Law, D. Russell, N. Pollock, P. Rosenbaum, S. Walter. The reliabilityand validity of Quality of Upper Extremity Skills Test. Phys Occup Therp Pedi., vol. 13,pages 1-18, 1993.

[14] Sc@ut. Sistema de Comunicacion Aumentativa y Adaptativa. URLhttp://scaut.ugr.es/

[15] Jose Luis Gonzalez, Marcelino J. Cabrera, and Francisco L. Gutierrez. Usingvideogames in special education. Proceedings of the 11th international conferenceon Computer aided systems theory, pages 360-367. Las Palmas de Gran Canaria,2007. ISSN 0302-9743.

[16] Sc@ut, comunicador aumentativo para Nintendo DS. URLhttp://www.youtube.com/watch?v=QO5FQqCr4Zc

[17] Meredith R. Golomb, Monica Barkat-Masih, Brian Rabin, Moustafa Abdelbaky,Meghan Huber, Grigore Burdea. Eleven months of home virtual reality telerehabili-tation - Lessons learned. URL http://www.ti.rutgers.edu/

[18] N. Taylor, P. L. Sand, R. H. Jebsen. Evaluation of hand function in children. Archivesof Physical Medicine and Rehabilitation, vol. 54, pages 129-35, March 1973.

[19] R. H Bruininks. Bruininks-Oseretsky Test of Motor Proficiency - Owner’s Manual.Circle Pines, MN: American Guidance Service, 1978.

[20] Reid DT. (2002) The use of virtual reality to improve upperextremity efficiency skillsin children with cerebral palsy: a pilot study. Techn Disabil 14: 53-61.

[21] Holden MK, Dyar T. (2002) Virtual environment training: a new tool for neurorehabil-itation. Neurol Rep 26: 62-74.

[22] You SH, Jang SH, Kim YH, Kwon YH, Barrow I, Hallett M. Cortical reorganizationinduced by virtual reality therapy in a child with hemiparetic cerebral palsy. Dev MedChild Neurol. 2005 Sep; 47(9): 628-35.

[23] Bruininks R. Bruininks-Oseretsky Test of Motor Proficiency. Minnesota: AmericanGuidance Service, 1978.

[24] Taub E, Wolf S. Constraint-induced movement technique to facilitate upper extremityuse in stroke patients. Topics Stroke Rehab 3: 38-61, 1997.

[25] Van Der Lee JH, Beckerman H, Knol DL, de Vet HC, Bouter LM. Clinimetricproperties of the motor activity log for the assessment of arm use in hemipareticpatients. Stroke 35: 1410-1414, 2004.

[26] Fugl-Meyer AR, Jaasko I, Leyman I, Olsson S, Steglind S. The post-strokehemiplegic patient. A method for evaluation of physical performance. Scand J RehabilMed 7: 104-122, 1975.

72


[27] Christine Brumback, Liubo Borrisov, Jeffrey Galusha, y Annette DiIorio. Intellivision.The Sixth International Conference Ubiquitous Computing, Nottingham, UK, 2004.University of Nottingham.

[28] Deutsch JE, Borbely M, Filler J, et al. Use of a low-cost, commercially availablegaming console (Wii) for rehabilitation of an adolescent with cerebral palsy. Phys Ther.2008.

[29] Janet Eyre. Nintendo Wii may help cerebral palsy children. Jul 2008.http://www.telegraph.co.uk/news/uknews/2276616/Nintendo-Wii-may-help-cerebral-palsy-children.html

[30] Pearson, E. and Bailey, C. (2007). Evaluating the potential of the Nin-tendo Wii to support disabled students in education. In ICT: Providingchoices for learners and learning. Proceedings ascilite Singapore 2007.http://www.ascilite.org.au/conferences/singapore07/procs/pearson-poster.pdf

[31] Palisano RJ, Rosenbaum PL, Walter SD, et al. Development and reliability of asystem to classify gross motor function in children with cerebral palsy. Phys Ther.1997; 39:214-223.

[32] Wechsler Intelligence Scale for Children (WISC). URLhttp://en.wikipedia.org/wiki/Wechsler Intelligence Scale for Children

[33] Wii Remote controller. http://en.wikipedia.org/wiki/Wii Remote

[34] Berg Balance Scale. URL http://www.aahf.info/pdf/Berg Balance Scale.pdf

[35] Wii Balance Board. URL http://en.wikipedia.org/wiki/Wii Balance Board

[36] Clark RA, et al. Validity and reliability of the Nintendo Wii Balance Board for assess-ment of standing balance. Gait Posture (2009), doi:10.1016/j.gaitpost.2009.11.012.

[37] Ching-Hsiang Shih, Ching-Tien Shih and Ming-Shan Chiang. A new standingposture detector to enable people with multiple disabilities to control environmentalstimulation by changing their standing posture through a commercial Wii BalanceBoard. Research in Developmental Disabilities, Volume 31, Issue 1, January-February 2010, Pages 281-286.

73

Motor Rehabilitation for People with Cognitive Impairments ...deim.urv.cat/~francesc.serratosa/2010_09_07_Joardi_Alsina_MEIS_Mas… · Motor Rehabilitation for People with Cognitive

Documents