CordiAAL: Enhanced Motivation for Cardiological Ergometer Training

through Virtual Groups in Virtual Worlds

Israﬁl Akman

, Rafael Bielen

, Henning Br

ummer

, Egor Kudrjaschow

, Stefan Tschentscher

Oliver Dohndorf

, Heiko Krumm

, Jan-Dirk Hoffmann

, Anke Workowski

and Detlev Willemsen

Lehrstuhl Informatik IV, TU Dortmund University, Martin-Schmeißer-Weg 18, 44221 Dortmund, Germany

Sch

uchtermann-Schiller’sche Kliniken, 49214 Bad Rothenfelde, Germany

Keywords:

Ambient Assisted Living, Bicycle Ergometer, Cardiac Disease, Preventive Programme, Secondary Cardiac

Prevention, Immersive Virtual Reality.

Abstract:

The project CordiAAL purposes the goal of developing a training system for patients with cardiovascular

diseases. The ﬁrst aim is to reduce the anxiety of overexertion by using a continuous monitoring of vital

parameters and the second is to increase the patients’ motivation according to a moderate physical activity by

the virtual environment of the system. The training can be done alone or in common with several patients in a

virtual group, regardless of the patients’ location. When grouped together, the performance of the individual

patient is aligned, so that patients with different ﬁtness levels are able to work together. The evaluations

revealed that the increase in motivation could be achieved by training in the virtual environment in virtual

groups.

1 INTRODUCTION

Cardiovascular diseases are the leading causes of

deaths in industrialized countries. In Germany

there are 40.2% of all deaths based on cardiovascu-

lar diseases (CVD) (Statistisches Bundesamt, 2011).

Mainly elderly people, 65 years and more, die be-

cause of these diseases. In 2011 the mortality rate of

over 65 year old people was 92% (Statistisches Bun-

desamt, 2011). Around 15% of medical expenses,

with 37 billion euros, were spent for these morbidities

for the care, treatment, prevention and rehabilitation

in Germany 2008 (Statistisches Bundesamt, 2008).

These facts increase the meaning of primary and sec-

ondary prevention of people with CVD. Up to 90%

of the cardiovascular morbidities are caused by the

cardiovascular risk factors: dyslipidemia, hyperten-

sion, diabetes mellitus, obesity, distress and physical

inactivity (Yusuf et al., 2004). Large epidemiolog-

ical studies and meta-analyzes of randomized, con-

trolled studies (evidence-based medicine, level IA)

have shown that regular physical endurance activity

is associated with a reduction of cardiovascular mor-

bidity and mortality (Nocon et al., 2008). There-

fore physical inactivity is considered the most impor-

tant modiﬁable risk factor (Bjarnason.Wehrens et al.,

2009). According to results of the National Health

Survey only 41% of German adults (45 up to 65 years

old) are more than 2 hours physically active per week,

about 38% in this age group do not exercise and

around 48% of German people older than 65 years

do not exercise (Gesundheit in Deutschland aktuell

– Telefonischer Gesundheitssurvey (GEDA), 2010).

The ﬁtness level of a person is correlated with the

exercise outcome. The life expectancy of ”Fit Peo-

ple”, people with a high performance, is higher than

for people, who do not perform exercises regularly

(Sattelmair et al., 2011). The risk of a cardiac event

is reduced through regular physical activity by posi-

tive physiological adaptations (Lee et al., 2003) and

the major modiﬁable risk factor is physical inactivity

(Balady et al., 2007).

Because of this background cardiac patients train

their physical activity during cardiac rehabilitation. It

is largely documented that the traditional rehabilita-

tion after discharge from hospital is efﬁcient and ac-

knowledged generally. Several studies conﬁrm that

the cardiac rehabilitation improves the functional ca-

pacity and psychosocial resources. But this success

is not a long term effect (Kotseva et al., 2009), af-

ter one year, the output level is usually reached again.

This lifts the obligation and responsibility of every in-

dividual to make secondary preventive care. In Ger-

many, the secondary prevention of heart patients pro-

Akman I., Bielen R., Brümmer H., Kudrjaschow E., Tschentscher S., Dohndorf O., Krumm H., Hoffmann J., Workowski A. and Willemsen D..

CordiAAL: Enhanced Motivation for Cardiological Ergometer Training through Virtual Groups in Virtual Worlds.

DOI: 10.5220/0004797300780089

In Proceedings of the International Conference on Health Informatics (HEALTHINF-2014), pages 78-89

ISBN: 978-989-758-010-9

 2014 SCITEPRESS (Science and Technology Publications, Lda.)

vides outpatient heart groups only. However, only 13-

40% of all CVD patients do participate (Bjarnason-

Wehrens et al., 2006). Reasons for this lack are that

these groups are temporally inﬂexible and not nation-

wide available. In addition, there is a high drop-

out rate of new members (Bjarnason-Wehrens et al.,

2006). IT-based systems providing Ambient Assisted

Living functions can be a solution. Therefore the

chair IV of the Technical University of Dortmund

and the Sch

uchtermann-Schiller’sche Kliniken in Bad

Rothenfelde developed a virtual biking application as

a part of a two-semester project, which should moti-

vate patients to do exercises regularly. The CordiAAL

project is an attempt to break the monotony of regu-

lar ergometer training sessions with virtual adventure

components and to sustain the motivation and adher-

ence of CVD patients. Healthcare professionals and

computer science students developed a software for

a bicycle ergometer in combination with a PC and

wireless sensors (ECG, respiration rate, body temper-

ature, oxygen saturation, blood pressure). The system

is integrated into an internet based community, called

RehaWeb, a special website for heart patients, with

a combination of social networking features, edito-

rial contents as well as mobile support and monitor-

ing (Dohndorf et al., 2012). The user can go by bike

through a virtual environment with a movable control

and a virtual reality head-mounted display (HMD),

that the session is as real as possible. The bike con-

sists of a heart-rate controlled load system. The car-

diologist can deﬁne individual corridors for each vi-

tal parameter and can set the optimal load for each

user to prevent overexertion and underexertion. The

training can be fulﬁlled online in a group of up to six

users with different performances or alone in a single

modus. Users can virtually train together without ac-

tually being together. Thereby heart patients will be

motivated through the advanced sensor-based gaming

environments to do exercise. With the deﬁned limits

of the vital data, the users can loose their anxiety of

exercise and get more body awareness. For the do-

cumentation of the training and vital data a report of

each session is generated. The CordiAAL system can

support a sustained prevention and rehabilitation of

people with CVD. The CordiAAL project is based on

the results of ”OSAMI - Open Source Ambient In-

telligence Common”, which was funded by the Fed-

eral Ministry of Education and Research in Germany

(ITEA2) (Busch et al., 2009). This paper will intro-

duce with related works, discuss the requirements and

possible solutions to the system afterwards. The fol-

lowing section will describe the architecture of the

system in detail. Finally, ﬁrst evaluation results of the

system will be presented.

2 RELATED WORK

IT-based systems, that motivate persons to do sports,

can be categorized according to their nature as immer-

sive and non-immersive systems, and regarding to the

kind of environment where they are used. Some sys-

tems can be used at home, but others are made to be

used in physical therapy.

One non-immersive system usable in every envi-

ronment is the Geosocial Network (Boulos and Yang,

2013). This system requires a GPS and an internet-

capable mobile phone. The application deﬁnes a tar-

get in the local area, where the user should travel to.

At an easy level the target is clearly deﬁned by GPS

coordinates, but at a more difﬁcult level the coordi-

nates are hidden by small puzzles. When the user

reaches the target, it is recorded in his proﬁle and

more challenges are unlocked. The proﬁle of each

user is also available in a web portal and is visible

for other players. The training aspect of the Geoso-

cial Network is based on the hiking of the users,

whereas the achievement of objectives and the com-

parison with other players serve an enhanced motiva-

tion. Other examples of such systems are CodeRun-

ner

or Dokobots

The Wiihab system (Anderson et al., 2010) is im-

mersive and made for physical therapy. It is designed

for patients who have restricted moving abilities. This

system requires a Personal Computer, a Wii Balance

Board and a Wii Mote. The Wii Mote is a video

game controller with a built-in motion sensor, the Wii

Balance Board measures the weight and the balance

of the patient. The Wiihab system has ﬁve built-in

games. Every game trains a different aspect, like the

balance of the body or the accuracy of the hand’s

movements. During a game session the training is au-

tomatically adapted to the patient’s health status. Af-

ter every session, the patient gets feedback about his

training success. So, the Wiihab system increases mo-

tivation for rehabilitation training by playing games

and the comparison with other players.

Furthermore, the Kinect system realizes a sensor-

based gaming environment. Similar to the Wii, the

Kinect system motivates the users playfully to per-

form physical activities. This system is immersive

and can be used in every environment. The Kinect

is basically a camera which detects movements of

the body. The user itself is the input device for a

video game. That is unique and increases the im-

mersive factor and hence the motivation. Examples

of immersive and physical activity promoting video

http://www.coderunnergame.com/

http://www.dokobots.com/

CordiAAL:EnhancedMotivationforCardiologicalErgometerTrainingthroughVirtualGroupsinVirtualWorlds

games are Your Shape

or UFC Personal Trainer

These games have certain exercises, training different

parts of the body and recording vital data during the

training.

A very prominent example of an immersive

system, only usable in hospital-like environments, is

the game Re-Mission (Lampert et al., 2009). It is di-

rected to children, who have cancer and aims at in-

ﬂuencing the behavior and attitude of young patients

positively to their disease. The game was created by

Games for Health

. This organization has the objec-

tive to motivate patients’ physical activities and hence

prevent illness and disease with the help of video

games. In this game the patient ﬁghts against the dis-

ease cancer, in particular against tumors and similar

signs of cancer using various weapons.

3 REQUIREMENTS AND

SOLUTION APPROACHES

In order to serve its purpose as an aid in the prophy-

lactic and rehabilitation phase the envisaged project

CordiAAL has to meet following social, technical and

medical requirements:

Social Requirements

Easy usage of the Application: Because the pri-

mary target user group consists of elderly patients

which usually have little experience in dealing

with modern media, the user-friendliness is of

great signiﬁcance. We take up this challenge

by sticking to the ﬁve rules as described in

(Caprani et al., 2012, p. 101). In respect of these

rules, we ﬁrstly ensure a fast running system

and secondly realize a very intuitive operation of

the application, where, for example choices are

limited in order not to confuse the user. Thirdly,

the guidance through the application is very clear

so that no ambiguities are given. Not least of all,

the usage is evaluated intensively. An easy usage

of the modiﬁed RehaWeb website is guaranteed,

since it was evaluated for the original version

(Dohndorf et al., 2012), and the extensions are

maintaining the original RehaWeb principles and

styles.

Building Trustfulness: In the sensitive area of me-

dical applications, the mediation of trust is a nec-

essary prerequisite with respect to a high user ac-

http://www.ubi.com/DE/Games/Info.aspx?pId=10769

http://www.golem.de/1104/82633.html

http://gamesforhealth.org/

ceptance. For that purpose, the users are informed

about the relevant issues of data security, safety

and data privacy. In the medical ﬁeld the analy-

sis of training sessions by medical professionals

increases trust.

Increasing Motivation to Secure Long-term Use:

Patients use to neglect their training in case of

a motivation lack (Shahsavari et al., 2012). We

focus on that with the main feature of CordiAAL,

the group training mode, which keeps the mo-

tivation on a high level. Boredom is prevented

through a common training with like-minded

patients. Not least of all, the HMD and the voice

chat are conducive to this. Moreover monotony

does not arise due to the existence of several

maps.

Reduction of Fears: As it is generally known, pa-

tients with a cardiac disease commonly have to

cope with fears (Eifert et al., 2000). Thus, the

software CordiAAL pursues the goal of reducing

these fears in several ways. Technically, thresh-

olds for the heart rate, deﬁned by medical profes-

sionals with respect to the health level of the pa-

tients, are used in each training session. Further-

more the training load during a training session

is adapted instantly according to the current heart

rate so that the risk of a further cardiac event dur-

ing the training is minimized.

Socially, the user gets in contact with like-minded

patients that suffer from a similar health issue,

which helps to distract the user’s fears. Further-

more, the training is not felt to be a training, in-

stead it is felt to be a game.

Technical Requirements

Achievement of Data Security and Privacy: We

extend the RehaWeb project which already en-

sures these requirements. Our extension contains

the web interface for the cardiologist where he

can deﬁne and adjust vital thresholds for a patient

and a detailed web statistic for each training. The

statistics can only be accessed by authorized per-

sons.

Recording of Vital and Load Parameters: The vi-

tal data have to be delivered from the body sensors

to the client software. These data are the basis

for the dynamic ergometer load control system.

The vital data used in CordiAAL comprise the

ECG, the respiration rate, the body temperature,

the SpO2 level and the blood pressure.

Controlling Functions and Dynamic Group

Control: The load and the RPM of the ergometer

HEALTHINF2014-InternationalConferenceonHealthInformatics

deﬁne the speed of the virtual bike. The load is in

turn inﬂuenced by the virtual world environment,

e.g. the gradient or the ground character. During a

group training, patients must have the possibility

to continue the training within their group even

if they have a lower ﬁtness level compared to

other group members. Therefore we developed

a corresponding load controlling algorithm: the

dynamic group control. During a group training

this provides a regulation of the ergometer load

of a patient in a way that the patient is capable

to work with a higher RPM, since the load is

decreased when his health status turns critical.

Thus, the patient is able to stay virtually near to

the group cycling with a lower load but a higher

RPM, and therefore a higher velocity, whereas

other patients’ load and RPM may vary.

Real-time: In order to create a sufﬁciently realis-

tic virtual environment during the cycling within

CordiAAL we have to meet real-time constraints.

The rendering during a training shall not introduce

signiﬁcant delays, even during a group training,

where the other bicycles of the participating pa-

tients have to be rendered also. So, the connection

must not slow down the virtual training. These

constraints also have to enable the mentioned con-

trolling functions. Therefore we use Unity3D for

rendering which works well with modern comput-

ers. In addition the position data of group mem-

bers is communicated periodically using small

UDP-messages.

Medical Requirements

Prophylactic and Rehabilitation Measure: A suf-

ﬁcient training effect must be guaranteed. We

achieve this goal by providing several maps for

virtual training environments, where each of them

is categorized by a difﬁculty level. This leads to

an optimal training for the patients with respect

to their ﬁtness level. Moreover the load control

supports efﬁcient and risk-reduced training. This

results in improving the patients’ sporting perfor-

mance as well as their state of health.

Recording of Vital Data: For the acquisition of

physical constitution in the course of cardiac re-

habilitation and prophylaxis a broad spectrum of

medical sensors is supported. They monitor the

most important medical data, the heart rate, the

respiration rate, the body temperature, the SpO2

level and the blood pressure.

Analyzing and Visualizing Vital Data: The record-

ed vital data have to be analyzed by the

application in order to control the ergometer load.

The data have to be visualized on the screen.

In addition, the vital data are also visualized in

RehaWeb in a much more detailed way. This en-

ables the cardiologist to examine these data during

the training phase of each patient and therefore

plan the next training sessions.

4 ARCHITECTURE

The project CordiAAL is based on a typical

client/server system (see ﬁgure 1). On the one hand

a web server with database access runs on the server-

side. This server is responsible for the requests of the

RehaWeb community while the data access is realized

via a RESTful-API (Pautasso et al., 2008). On the

other hand an application server acts as a communica-

tion interface between the different ergometer training

applications. On this application server the vir-

tual group information is processed via the position-

tracking application as well as the real-time commu-

nication via the voice chat application.

Figure 1: System architecture of the project CordiAAL.

The client-side consists of three different groups

of actors that interact with the system. Firstly the

administrator realizes both the administration of each

patient and the maintenance of the RehaWeb website.

The cardiologists are able to deﬁne or to adjust vital

parameters of their patients. Moreover, in RehaWeb

they can supervise the recorded statistics. The pa-

tients immerse in the virtual world by using the

ergometer and the HMD and are able to perform

a virtual group training. Furthermore they can use

RehaWeb’s social network function in order to ﬁnd

out virtual routes or arrange a virtual group training.

The integrated devices comprise the virtual reality

headset for displaying the immersive 3D world as

well as a bicycle ergometer for controlling the load

and a harness for recording the vital data (see ﬁg-

ure 2). All of these components deﬁne a cyber-

physical system (CPS) (Lee, 2008). In such systems

the strict separation between the real and virtual world

becomes neutralized. The physical environment gets

CordiAAL:EnhancedMotivationforCardiologicalErgometerTrainingthroughVirtualGroupsinVirtualWorlds

recognized by the introduced sensors and is mixed

with parameters of the virtual reality. The obtained

information is processed and thus the physical envi-

ronment is inﬂuenced correspondingly. For instance

the revolutions of the pedals and the movement of the

steering wheel control the velocity and the direction

of the player respectively. On the other hand a moun-

tain or bad ground covering in this world causes a

higher adjustment of the ergometer’s load, but always

in consideration of the deﬁned thresholds. The vital

parameters ensure medically appropriate adjustments.

Thus, this CPS combines the patient, the ergometer

control and the Unity3D game engine to a virtual en-

vironment, that facilitates a group training and adjusts

the individual performance levels.

Figure 2: Sensors environment of the system CordiAAL.

4.1 Social Network

The RehaWeb system pursues the goal of motivat-

ing cardiac patients for sustainable training. Virtual

group trainings on selected virtual 3D worlds can be

planned in the RehaWeb community together with

friends. The collected data are transmitted to the

RehaWeb server and are reviewable there in the form

of a training diary for both the patient and the medi-

cal staff. The patient is able to talk with other patients

about the development of his own performance and

other prevention-relevant topics in the forum.

The Architecture of the social network RehaWeb

consists of two layers. The ﬁrst one comprises the

server-side. The server component has been designed

with a service-oriented and layered architecture. The

Tomcat application server contains the community

software as a Web2.0 user interface which is con-

nected to a MySQL database. The server provides

central storage and processing of all data. The access

is provided via a RESTful-API. The second layer con-

sists of the user’s web browser. Java-ServerFaces 2.0

is used in order to generate dynamic HTML code.

The Structure of the web portal RehaWeb is con-

ceived hierarchically as a tree structure. The top level

of the web portal is called the root. It is followed by

the level of the main navigation and its subnavigation.

In RehaWeb the level of the main navigation is repre-

sented as a horizontal bar above the green strip. The

further sublevels are displayed as a subnavigation in

the left column (see ﬁgure 3).

Figure 3: Portal structure of the website RehaWeb.

4.2 Communication

The CPS CordiAAL shows three types of communi-

cation. These are described in the following.

The ﬁrst one is the web communication. The com-

munication between the web browser and the server is

based on HTTP/HTML.

The second type is the access to the services of the

server. These services are composed of the access,

processing and storage of all necessary data before

and after the training. The services are designed in

respect of REST principles (Fielding, 2000). They are

stateless and encapsulate every communication with

the database. Each service class is responsible for one

type of resource and provides methods according to

the CRUD principle (Novick, 2008). JSON messages

as the transport format are transmitted over HTTP.

The third type regards the time critical commu-

nication between the ergometer training applications.

For the realization of a virtual group training, the

current data concerning each patient’s current state

within the virtual world has to be exchanged via the

position-tracking application. For instance the pa-

tient’s ID and his position vector are transmitted. An-

other aspect concerns the audio communication be-

tween the patients of a group, realized by the voice

chat application. On the application-side the open

source .NET Audio and Musical Instrument Digital

HEALTHINF2014-InternationalConferenceonHealthInformatics

Interface (MIDI) library NAudio

has been used to fa-

cilitate a data stream. The audio samples are transmit-

ted to the voice chat application over UDP messages

and forwarded to the members of the group (see ﬁg-

ure 4).

Figure 4: Voicemail distribution in different groups.

4.3 Training

The training is executed either as a single training or

as a group training. The challenge of this work is to

perform a training in virtual groups and the medical

load control of different patients. These aspects are

explained in the following.

Medical Monitoring

The following data are collected for the medical su-

pervision:

• ECG

• Respiration rate

• Body temperature

• SpO2

• Blood pressure

For the analysis of the SpO2 and blood pressure

data we make use of the Corscience NiBP2010 sen-

sor. The sensor communicates via an USB interface

with our system. Moreover the Zephyr BioHarness3

sensor, which enables a precise recording of the vital

parameters is used. The BioHarness3 communicates

via a Bluetooth interface.

Load Control

Enabling the individual control of each patient’s load

within a group training needs to deal with several sen-

sors and a load control algorithm.

http://naudio.codeplex.com

Sensors & Actors: The ergometer is the central en-

tity for controlling the intensity during the training.

Within our system we have used the ergometer

Ergo bike2000 of the company Daum Electronics.

Via a RS232 interface data can be exchanged between

the ergometer and the PC. These data are the cur-

rent speed, the RPM and the load data to control the

ergometer. The handle bars are modiﬁed in a way that

they are rotatable and a realistic steering is achieved.

Controlling One Person: The load control for one

person is based on a typical control loop (Schulz,

2007) (see ﬁgure 5).

Figure 5: Control circuit of the load control.

The regulation monitors the current heart rate

value (y(t)) and compares it with the desired value

(w(t)). Thus, disturbances (d(t)), for example slopes

and overloads of the patient, are recognized and com-

pensated (u(t)). The compensation is achieved by re-

ducing or increasing the load.

Controlling a Virtual Group: In order to support

virtual groups of patients with different ﬁtness levels,

the group control reduces the load of low ﬁtness level

patients. In detail, this means that a group member

is not outpaced by the others due to his own lower

ﬁtness level. He has the possibility to cycle virtually

near to the others. The load (d(t)) of a patient whose

heart rate gets into the threshold area is reduced so

that this patient is able to train with a higher RPM

under less effort to retain the group velocity within

the virtual world.

Therefore patients who have undergone a load re-

duction are able to cycle with the same RPM as be-

fore. Thus, this mechanism enables the weaker pa-

tients and the ones with a lower ﬁtness level to stay

virtually near the other group members.

Immersive Game Environment

For the purpose of realizing an immersive world we

make use of a virtual reality head-mounted display as

well as of a game engine for displaying the 3D world.

Designing the application as a distributed system en-

ables multiplayer games over the internet.

CordiAAL:EnhancedMotivationforCardiologicalErgometerTrainingthroughVirtualGroupsinVirtualWorlds

Virtual Reality Head-mounted Display: To serve

its purpose as a motivational tool and to improve the

display quality of the 3D world a virtual reality head-

mounted display is integrated into the system, in de-

tail the Oculus Rift (developer version) is used. The

head-mounted display has got a large ﬁeld of vision

and fast head-tracking sensors. The Oculus Rift pro-

vides a DVI and a HDMI input interface as well as

an USB interface for transferring tracking data to the

ergometer training application.

Unity3D: Unity3D is a complete development en-

vironment, designed to develop video games and

rendering. The target platforms are Windows,

Mac OS X, video game consoles and mobile phones.

The development environment includes a visual edi-

tor, the C# editor Mono and different modules.

The main task of the visual editor is to create and

manipulate 3D objects. The 3D objects can be created

in programs such as Blender or Maya 3D. For details,

logic and behavior components can be attached to 3D

objects. Components deﬁne objects’ properties like

sound, look, material, texture, behavior and physical

properties. Own components can be written in C#,

JavaScript and Boo.

For the display of 3D objects Unity3D makes use

of its own 3D engine. The 3D engine also handles

the optimization of the scene so that the drawing is

efﬁcient.

For music and sounds Unity3D uses the FMOD li-

brary. Input devices such as mouse, keyboard and

joystick can be integrated directly into Unity3D.

Unity3D is free, but there is also a pro version. The

pro version allows to write components in C++ and

has some additional improvements in the 3D engine,

the pro version supports dynamic light, shadow ef-

fects and more efﬁcient rendering.

With Unity3D version 3.0 comes an Asset Store

in which resources uploaded by other users, for in-

stance 3D models, textures and sounds, can be bought

or downloaded for free.

Virtual Group Dynamics: In the single training

mode we have local access to the static 3D world.

This mode requires no further information about other

players. There is only communication to the server

before and after the training. Before the training the

information for the lobby, i.e. the currently training

virtual groups or the planned training groups, is re-

trieved. After a training session the statistic data of

the current session is sent to the server and stored in

the database. In the group training mode each pa-

tient’s local computer computes an own 3D-scenario,

namely the view of the respective patient. How-

ever, the different scenarios are coordinated with each

other. They use the same map and the positions of the

players are communicated continuously. The corre-

sponding distributed system is presented in ﬁgure 6.

Figure 6: Behavior of the group dynamics on a distributed

system.

The 3D world is maintained on each computer, the

other players are displayed dynamically according to

the information about each group member. In order

to obtain this information, the application exchanges

messages in the following form:

• Patient ID

• Patient name

• Group ID

• Position vector (x, y, z)

• Direction vector (x, y, z)

• Velocity

• Position of the steering wheel

• Training time

The distribution of the messages to the corresponding

group members is based on the group ID within the

position-tracking application. The further data con-

tained in the message are needed for displaying 3D

objects locally on the ergometer training application.

5 APPLICATION

CordiAAL consists of two main applications:

the RehaWeb website and the ergometer training

application. The ergometer training application is re-

sponsible for the whole training sessions of the pa-

tients and the RehaWeb website for user administra-

tion and community functions.

5.1 RehaWeb

During the development of the CordiAAL project

the existing RehaWeb application was extended

and adapted, so that it is now possible to combine

both applications. The RehaWeb website provides

HEALTHINF2014-InternationalConferenceonHealthInformatics

Figure 7: RehaWeb Home Surface.

functionality for three types of user groups: patient,

cardiologist and administrator. Every group has got

different permissions within RehaWeb.

Patient: Patients can log in after registering on

RehaWeb. After log in they can look at their last

workout statistics.

They can arrange to meet with other patients for

a virtual group workout within the CordiAAL

ergometer training application. Furthermore it is

possible to communicate in a forum to exchange

information and experiences with other patients.

Moreover they can exchange private messages

with friends inside the RehaWeb community.

Figure 8: Vital Parameters Surface.

Cardiologist: Cardiologists can use RehaWeb to

manage the vital parameters of their patients

(see ﬁg. 8). This is a very important function,

so that an individual and best possible workout

is guaranteed for each patient. It helps to avoid

overloads of the patients. The cardiologists are

able to set the values for the heart rate, respiration

rate, body temperature and SpO2. Initial values

have to be set for each patient after examination

by the clinic.

Additionally they have permissions to view

Figure 9: Detailed Statistic Surface.

and supervise detailed training statistics of all

their patients. Figure 9 shows an excerpt from a

sample statistic which has been recorded during a

training session.

Administrator: The administrator has got the per-

missions to manage all the users of RehaWeb.

Only he is able to add a new virtual route which

the patients can choose for their virtual training

within the ergometer training application.

5.2 Ergometer Training Application

The ergometer training application is the main

application of the CordiAAL system and responsible

for the training sessions. It controls the workout,

manages the sensors and takes care of the health

CordiAAL:EnhancedMotivationforCardiologicalErgometerTrainingthroughVirtualGroupsinVirtualWorlds

status of the training members. A training session can

be divided into three main sections: Training Prepa-

ration, Training Execution and Training Conclusion.

Training Preparation: First of all after starting the

application, the patients log in to the ergometer

training application with their user data from

RehaWeb. After a successful login the patients are

able to view their last workout statistics or they

can choose between a single and a group training

session. By selecting a group training it is pos-

sible to start a workout together with up to six

other patients at the same time using the same vir-

tual route. In this training mode all patients can

see each other like in a multiplayer game. The

other participants are covered in different colors

and they are able to communicate via voice chat.

After pushing the training start button, the patients

are asked to connect all sensors and devices. The

system waits until the patients have connected all

sensors. In an overview all connectable sensors

and devices are displayed with their connection

status.

Before the workout starts, the patients have to

Figure 10: Training Settings.

conﬁgure the settings of the training. They can

select a virtual route which they want to drive

in, set the duration of the workout and in group

training mode they can set the maximum number

of other patients driving with them (see ﬁg. 10).

It is also possible to join an existing training in-

stead of creating a new one. By clicking on a dis-

played training from the list, a short status infor-

mation is shown. It gives information about the

remaining time, which virtual route is used and

how many other patients participate in this work-

out. The patients get only training sessions dis-

played they are able to join. Either the training is

a public one or the patient has previously arranged

it on RehaWeb. If they want to join one of the cur-

rent training sessions they simply have to select

one and click on the start button. The workout

starts immediately. As well there is a menu item

reachable from the main menu, where the patients

get an overview list of training sessions which will

start in the future.

Figure 11: Workout Screen.

Training Execution: From beginning to end of a

training session the application records all sen-

sor data like heart rate, SpO2, temperature, speed,

distance, time and rpm in an interval of one sec-

ond. Some of this information is displayed on the

screen during the training. So the patients can

see directly what the current load in watt is, the

speed in km/h, meters of the altitude proﬁle and

the driven distance. On the left side of the dis-

play the names of the other group members and

their colors are shown. But the most important

information displayed is the heart rate status bar.

It is a big status bar in top of the screen, which

is corresponding to the current heart rate of the

patient (see ﬁg. 12). Thus the patients can eas-

ily see if they perform their training in the right

way. The status bar is colored from green to red.

These scale correlates to the settings the cardiolo-

gist managed on the RehaWeb website. Training

inside the green range is optimal. The yellow and

red range indicate to drive faster or slower to get

back into the green area. Additionally a message

occurs on the screen with instructions. If the phys-

ical load is still too high, the system will regulate

the load of the ergometer down, until the heart rate

of the patient will be out of the red area. After the

predeﬁned time span the workout will ﬁnish au-

tomatically, but the patient is capable to abort the

training session by clicking an exit button before

the training time is elapsed.

Figure 12: Heart Rate Status Bar.

Training Conclusion: After the workout, the pa-

tients are asked to specify a Borg value (Borg,

HEALTHINF2014-InternationalConferenceonHealthInformatics

1970) for the training, which is an important clue

for the cardiologist about the subjective inten-

sity of the training (see ﬁg. 13). Thereafter all

Figure 13: Surface to enter the Borg Value.

recorded data from the sensors and the application

are transmitted and stored in a database hosted

by the CordiAAL server. If it is impossible to

transmit the data to the server, e.g., because of

a broken internet connection, the recorded data

will be saved on the local hard disk. Next time

the ergometer training application is started the

application tries to submit the data again. Finally

a short summary of the current training session re-

sults together with average values like heart rate,

distance and time are displayed (see ﬁg. 14).

Figure 14: Workout Summary.

6 EVALUATION

The evaluation aims at two different aspects: Analyz-

ing how far the CordiAAL system assists the users in

the execution of a training session and increasing the

motivation for a sustainable training. Secondly, veri-

fying the application in terms of design, usability and

ergonomics of the user interface.

In the evaluation process, the subjects performed

an afore designed scenario, which relates to a normal

use of the software. The scenario comprises a training

session, inspecting the statistics and a review of the

completed session. During the training session, it will

be analyzed how the displayed information affects the

subjects, how they experience the virtual environment

through the virtual reality head-mounted display, and

how far the virtual training affects the increase of mo-

tivation. Here, an important aspect is whether a vir-

tual group training inﬂuences the motivation and how

it stimulates the subject’s ambition, so that training

sessions are no longer felt as a liability but more as a

game.

To evaluate the aforementioned aspects, a ques-

tionnaire was presented to the subjects after the

training. The questions were divided into three cat-

egories and rated on a scale from one to ﬁve (1 = ”the

statement is not true”, 5 = ”the statement is true”).

First, the subjects were asked to answer questions for

self-assessment, these questions helped to determine

how motivated the subjects are in doing sports. In the

second category of the questionnaire, subjects were

asked to rate the usability of the system. At the end of

the test subjects were asked to rate the virtual training

and specify whether this type of training could in-

crease their motivation.

Overall 10 participants, including 2 women and 8

men were tested in the age range of 25 years up to

36 years. The sequence of execution took about 30

minutes to complete, with two training sessions by 10

minutes.

In order to present the usability evaluation results,

ﬁve categories were deﬁned (see ﬁgure 15). As you

can see, the software is considered as user-friendly,

an average value of 3,7 out of 5 was achieved. One of

the most important aspects of this case was to make

the software as intuitive as possible, so long learning

phases are avoided and this barrier pulls no reduction

in motivation.

Figure 15: Evaluation of the usability questions by cate-

gories ”Understand Solutions, Learnability, Operability, At-

tractiveness, Usability Compliance”.

To evaluate the training, the subjects self-assessed

their sportiness and their general sport-related moti-

CordiAAL:EnhancedMotivationforCardiologicalErgometerTrainingthroughVirtualGroupsinVirtualWorlds

vation to conclude how far the training in the virtual

environment could increase their motivation. To il-

lustrate the inﬂuence of the training with CordiAAL

we have taken two aspects into account. The moti-

vation for doing sports with and without any general

technical aids, and the same with CordiAAL as a spe-

ciﬁc technical aid. Figure 16 clearly illustrates that

the subjects have difﬁculty doing sustainable sporting

activities and also the willingness of using aids is low.

In turn, the increasing motivation through the virtual

training is clearly evident. An increased value for the

readiness to use CordiAAL as a technical aid for reg-

ular sport activities was also indicated. The most im-

portant motivation factor was thereby the training in a

virtual group. It was stated that the training and com-

munication with other group members minimize the

effort of training and increase the ambition.

Figure 16: Evaluation of the questions of motivation and

willingness to use the aids, divided according to data ob-

tained from the self-assessment and evaluation after training

with CordiAAL.

7 CONCLUSIONS

The user study clearly demonstrates that the main ob-

jectives of the CordiAAL system were achieved, real-

izing a playfully ergometer training which motivates

users to a regular physical activity, bringing users

with different ﬁtness levels together and to give them

safety at the same time.

This is done by a permanent monitoring of vital

data and an adjustment of the ergometer load based

on these data. Since the study was not conducted

with subjects with heart disease, for ethical reasons,

the question of the decrease of anxiety remains open.

The next stage in the evaluation would be a long-

term study to check if the motivation stays high-

leveled in a long-term use.

REFERENCES

Anderson, F., Annett, M., and Bischof, W. F. (2010). An-

nual Review of Cybertherapy and Telemedicine 2010.

Studies in Health Technology and Informatics.

Balady, G., Williams, M., Ades, P., Bittner, V., Comoss, P.,

Foody, J., Franklin, B., Sanderson, B., and Southard,

D. (2007). Core components of cardiac rehabili-

tation/secondary prevention programs. Circulation,

115(20):2675–82.

Bjarnason-Wehrens, B., Held, K., and Karoff, M. (2006).

Herzgruppen in Deutschland: Status quo und Perspek-

tiven. Herz, 31:559–565.

Bjarnason.Wehrens, B., Drsch, M., Gielen, S., Halle, M.,

Hambrecht, R., and Kindermann, W. (2009). Leitlinie

krperliche aktivitt zur sekundrprvention und therapie

kardiovaskulrer erkrankungen. Clinical Research in

Cardiology Supplement (CRICS), pages 1–44.

Borg, G. (1970). Perceived exertion as an indicator of so-

matic stress. Scandinavian journal of Rehabilitation

Medicine, 2:92–98.

Boulos, M. N. K. and Yang, S. P. (2013). Exergames for

health and ﬁtness: the roles of GPS and geosocial

apps. Technical report, International Journal of Health

Geographics.

Busch, C., Litvina, A., and Willemsen, D. (2009). Kar-

diale Telerehabilitation auf Basis einer ﬂexiblen Platt-

form f

ur verteilte Systeme - das OSAmI-D Projekt.

e-Health, pages 218–221.

Caprani, N., O’Connor, N. E., and Gurrin, C. (2012). Touch

Screens for the Older User. In Cheein, D. F. A., editor,

Assistive Technologies, chapter 5. InTech.

Dohndorf, O., G

oring, A., Krumm, H., Schneider, A., Som-

mer, A., Sladek, S., Busch, C., Hoffmann, J.-D., and

Willemsen, D. (2012). RehaWeb – An information

system for cardiologic rehabilitation assistance in the

third phase. Technical report, TU Dortmund Univer-

sity.

Eifert, G. H., Thompson, R. N., Zvolensky, M. J., Edwards,

K., Frazer, N. L., Haddad, J. W., and Davig, J. (2000).

The cardiac anxiety questionnaire: development and

preliminary validity. In Behaviour Research and Ther-

apy, volume 38, pages 1039–1053. Elsevier Ltd.

Fielding, R. T. (2000). Architectural styles and the design

of network-based software architectures. PhD thesis,

University of California. AAI9980887.

Gesundheit in Deutschland aktuell – Telefonischer Gesund-

heitssurvey (GEDA) (2009/2010). Gesundheit in

Deutschland aktuell – Telefonischer Gesundheitssur-

vey (GEDA): Durchschnittliche sportliche Aktivit

pro Woche (Anteil der Befragten in Prozent).

Kotseva, K., Wood, D., Backer, G. D., Bacquer, D. D., Py-

oeraelae, K., and Keil, U. (2009). Cardiovascular pre-

vention guidelines in daily practice: a comparison of

euroaspire i, ii, and iii surveys in eight european coun-

tries. The Lancet, 373(9667):929 – 940.

Lampert, C., Schwinge, C., and Tolks, D. (2009). Der

gespielte Ernst des Lebens: Bestandsaufnahme und

Potenziale von Serious Games (for Health). Zeitschrift

ur Theorie und Praxis der Medienbildung, 15/16:1–

16.

Lee, C., Folsom, A., and Blair, S. (2003). Physical activity

and stroke risk: a meta-analysis. Stroke, 34(10):2475–

81.

HEALTHINF2014-InternationalConferenceonHealthInformatics

Lee, E. A. (5-7 May 2008). Cyber Physical Systems:

Design Challenges. Object Oriented Real-Time Dis-

tributed Computing (ISORC), 2008 11th IEEE Inter-

national Symposium on, pages pp.363–369.

Nocon, M., Hiemann, T., M

uller-Riemenschneider, F., Tha-

lau, F., Roll, S., and Willich, S. N. (2008). Association

of physical activity with all-cause and cardiovascular

mortality: a systematic review and meta-analysis. Eur

J Cardiovasc Prev Rehabil, 15:2239–246.

Novick, A. (2008). Implementing crud operations using

stored procedures: Part 1.

Pautasso, C., Zimmermann, O., and Leymann, F. (2008).

Restful web services vs. ”big”’ web services: mak-

ing the right architectural decision. In Proceeding of

the 17th international conference on World Wide Web,

WWW ’08, pages 805–814, New York, NY, USA.

ACM.

Sattelmair, J., Pertman, J., Ding, E., Kohl 3rd, H., Haskell,

W., and Lee, I. (2011). Dose response between physi-

cal activity and risk of coronary heart disease: a meta-

analysis. Circulation, 124(7):789–95.

Schulz, G. (2007). Regelungstechnik 1 Lineare und

Nichtlineare Regelung, Rechnergesttzter Regleren-

twurf. Oldenbourg Wissenschaftsverlag GmbH.

Shahsavari, H., Shahriari, M., and Alimohammadi, N.

(2012). Motivational factors of adherence to cardiac

rehabilitation. Iranian Journal of Nursing and Mid-

wifery Research, 17:7.

Statistisches Bundesamt (2008). Herz-

/Kreislauferkrankungen verursachen die h

ochsten

Krankheitskosten. Wiesbaden.

Statistisches Bundesamt (2011). 2011: Todesf

alle 2011

leicht r

uckl

auﬁg – h

auﬁgste Todesursache Herz-

/Kreislauferkrankungen. Wiesbaden.

Yusuf, S., Hawken, S., Ounpuu, S., Dans, T., Avezum, A.,

Lanas, F., McQueen, M., Budaj, A., Pais, P., Varigos,

J., and Lisheng, L. (2004). Effect of potentially mod-

iﬁable risk factors associated with myocardial infarc-

tion in 52 countries (the interheart study): case-control

study. Lancet, 364(9438):937–52.

CordiAAL:EnhancedMotivationforCardiologicalErgometerTrainingthroughVirtualGroupsinVirtualWorlds