[IEEE 2014 IEEE 38th International Computer Software and Applications Conference Workshops...

6
Mobile Pediatric Consultation and Monitoring System through Semantic Web Technology 1 Duygu Çelik, 2 Atilla Elçi, 3 Rıdvan Akçiçek, 4 Bora Gökçe, 5 Pelin Hürcan 1 [email protected], Istanbul Aydin University, Istanbul, Turkey, 2 [email protected], Aksaray University, Aksaray, Turkey, 3 [email protected], Acibadem Health Group Hospitals, Istanbul, Turkey 4 [email protected], Acibadem Health Group Hospitals, Istanbul, Turkey 5 [email protected], Acibadem Health Group Hospitals, Istanbul, Turkey AbstractmHealth is a popular aspect of eHealth applications that is starting to explode exponentially. The use of technology offering solutions with a potential of making patients’ lives easier in the health sector has gradually increased with mHealth. Technology has great bearing in the health sector in order to increase the efficiency to serve patients by helping to diagnose ailments reliably. Mobile systems help observation of child patients in their treatment process and the collection of health data on the course. Therefore, people will be more aware of their children’s health problems and preventive measures in accordance with those health problems. This article is on Mobile Pediatric Consultation and Monitoring System (mPCMS). Details of design and operation are given in this article. mPCMS not only will assist doctors but also help patients in decision making process of the medical treatment due to the information’s ease use. Accordingly, with mPCMS, there will be a contribution to distant consultation by the doctor and the family by collecting the health data before and after the treatment of child patients. KeywordsPediatric Monitoring Ontology, Semantic Web, Semantic Search Systems, E-health systems, Semantic Matching. I. INTRODUCTION Repetitive ear and throat infections have vital symptoms for man recurring health risks that are such as the lower respiratory tract infections, diarrheal diseases, upper respiratory tract infections. The most sensitive system against infection is the respiratory tract in children, consequently respiratory infections are the most common infective illnesses of childhood period. Upper respiratory infections (contagious illnesses), cold, flu, tonsillitis, and otitis are most common. By controlling what and how to diagnose and monitor of a patient children during his/her upper respiratory tract infections we can maximize the patients’ life quality and decrease parental anxiety during this stage. To take under control safe diagnoses and monitoring stages for children patients and their family, we need smart systems with strong knowledge management technology and ease of extension to provide information from additional e-health tools. Semantic Web [1], sometimes called as Web 3.0 is a new Web technology providing to create Ontology [2]. Ontology can be generated containing machine-readable (semantic) annotations of health risk groups for children, their unhealthy symptoms, and possible monitored result types from children body according to patients’ properties for repetitive ear and throat infections. Through the Semantic Web approach, personal health systems are technically based on ontology knowledge management, which is easily extensible to adopt by other additional e-health tools. The ontology that is specified by using Web Ontology Language (OWL) [3] is the shared consensus between personal health services and eHealth or mHealth tools that provide interoperation. This article discusses a Mobile Pediatric Consultation and Monitoring System (mPCMS) through Semantic Web technology that guides parents while gathering information from their child’s body when taking images of throat/tonsil and ear, measuring fever and listening to their lungs during upper respiratory tract infections. This system is planned to be offered as a kit that contains a membership number, a password, a digital thermometer, a pediatric otoscope, and a digital stethoscope compatible with smart phones. This kit would help parents miss less work and potentially decrease on night-time visits to emergency ward. Parts of this type of system have been practiced in many research projects. However, only the mPCMS is gathering four different types of health data together for both consultation and monitoring stages; measuring fever, taking throat/tonsil and ear images, listening to child’s lungs has not been recommended procedures. The CellScope’s 1 (formerly known as Remotoscope) is a smartphone-enabled otoscope project that enables physicians to remotely diagnose ear infections in children. Parents will be able to use the peripheral, which attaches to a smartphone camera lens, to send an image of their child’s inner ear that physicians can use to make a diagnosis. CellScope says ear infections in children make up 30 million doctor visits annually in the US alone. CellScope is a very successful 1 http://mobihealthnews.com/17598/cellscope-smartphone-diagnostic- startup-raises-1m/ 2014 IEEE 38th Annual International Computers, Software and Applications Conference Workshops 978-1-4799-3578-9/14 $31.00 © 2014 IEEE DOI 10.1109/COMPSACW.2014.61 354

Transcript of [IEEE 2014 IEEE 38th International Computer Software and Applications Conference Workshops...

Page 1: [IEEE 2014 IEEE 38th International Computer Software and Applications Conference Workshops (COMPSACW) - Vasteras, Sweden (2014.7.21-2014.7.25)] 2014 IEEE 38th International Computer

Mobile Pediatric Consultation and Monitoring System through Semantic Web Technology

1Duygu Çelik, 2Atilla Elçi, 3Rıdvan Akçiçek, 4Bora Gökçe, 5Pelin Hürcan1 [email protected], Istanbul Aydin University, Istanbul, Turkey,

[email protected], Aksaray University, Aksaray, Turkey, [email protected], Acibadem Health Group Hospitals, Istanbul, Turkey

[email protected], Acibadem Health Group Hospitals, Istanbul, Turkey [email protected], Acibadem Health Group Hospitals, Istanbul, Turkey

Abstract—mHealth is a popular aspect of eHealth applications that is starting to explode exponentially. The use of technology offering solutions with a potential of making patients’ lives easier in the health sector has gradually increased with mHealth. Technology has great bearing in the health sector in order to increase the efficiency to serve patients by helping to diagnose ailments reliably. Mobile systems help observation of child patients in their treatment process and the collection of health data on the course.Therefore, people will be more aware of their children’s health problems and preventive measures in accordance with those health problems. This article is on Mobile Pediatric Consultation and Monitoring System (mPCMS). Details of design and operation are given in this article. mPCMS not only will assist doctors but also help patients in decision making process of the medical treatment due to the information’s ease use. Accordingly, with mPCMS, there will be a contribution to distant consultation by the doctor and the family by collecting the health data before and after the treatment of child patients.

Keywords— Pediatric Monitoring Ontology, Semantic Web, Semantic Search Systems, E-health systems, Semantic Matching.

I. INTRODUCTION

Repetitive ear and throat infections have vital symptoms for man recurring health risks that are such as the lower respiratory tract infections, diarrheal diseases, upperrespiratory tract infections.

The most sensitive system against infection is the respiratory tract in children, consequently respiratory infections are the most common infective illnesses of childhood period. Upper respiratory infections (contagious illnesses), cold, flu, tonsillitis, and otitis are most common.

By controlling what and how to diagnose and monitor of a patient children during his/her upper respiratory tract infections we can maximize the patients’ life quality and decrease parental anxiety during this stage. To take under control safe diagnoses and monitoring stages for children patients and their family, we need smart systems with strong knowledge management technology and ease of extension to provide information from additional e-health tools.

Semantic Web [1], sometimes called as Web 3.0 is a new Web technology providing to create Ontology [2]. Ontology can be generated containing machine-readable (semantic) annotations of health risk groups for children, their unhealthy symptoms, and possible monitored result types from children body according to patients’ properties for repetitive ear and throat infections. Through the Semantic Web approach,personal health systems are technically based on ontology knowledge management, which is easily extensible to adopt by other additional e-health tools. The ontology that is specified by using Web Ontology Language (OWL) [3] is the shared consensus between personal health services and eHealth or mHealth tools that provide interoperation.

This article discusses a Mobile Pediatric Consultation and Monitoring System (mPCMS) through Semantic Web technology that guides parents while gathering information from their child’s body when taking images of throat/tonsil and ear, measuring fever and listening to their lungs during upper respiratory tract infections. This system is planned to be offered as a kit that contains a membership number, a password, a digital thermometer, a pediatric otoscope, and adigital stethoscope compatible with smart phones. This kit would help parents miss less work and potentially decrease on night-time visits to emergency ward. Parts of this type of system have been practiced in many research projects.However, only the mPCMS is gathering four different types of health data together for both consultation and monitoring stages; measuring fever, taking throat/tonsil and ear images, listening to child’s lungs has not been recommended procedures.

The CellScope’s1 (formerly known as Remotoscope) is asmartphone-enabled otoscope project that enables physicians to remotely diagnose ear infections in children. Parents will be able to use the peripheral, which attaches to a smartphone camera lens, to send an image of their child’s inner ear that physicians can use to make a diagnosis. CellScope says ear infections in children make up 30 million doctor visits annually in the US alone. CellScope is a very successful

1 http://mobihealthnews.com/17598/cellscope-smartphone-diagnostic-startup-raises-1m/

2014 IEEE 38th Annual International Computers, Software and Applications Conference Workshops

978-1-4799-3578-9/14 $31.00 © 2014 IEEE

DOI 10.1109/COMPSACW.2014.61

354

Page 2: [IEEE 2014 IEEE 38th International Computer Software and Applications Conference Workshops (COMPSACW) - Vasteras, Sweden (2014.7.21-2014.7.25)] 2014 IEEE 38th International Computer

approach for mhealth project which has a simple clip-on system converting a standard smartphone into a digital otoscope with data transmission to a remote physician. It provides high-quality imaging and the ubiquity of wireless communication, a smartphone otoscope presents a cost effective and medically sound solution.

Another study is the StethoMic2 that is a low-cost digital stethoscope prototype for off-the-shelf components of smart phones. StethoCloud currently runs on Windows Phone, Android and legacy J2ME compatible phones and an iOS version of the app is currently in development. In addition, StethoCloud is unique from other digital stethoscopes because they harness the power of cloud computing to create the artificial intelligence that makes their stethoscopes "smart". They used Microsoft's Windows Azure platform to build the world's largest database on breath and heart sounds. Through this database, they planned to train their machine learning algorithms to extract features and perform knowledge discovery of breath sound patterns and their related pathologies. This training will enable the machine to accurately diagnose pneumonia, asthma and other respiratory diseases with at least the same level of sensitivity and specificity as human diagnosticians.

mPCMS project differs from CellScope and StethoMic.mPCMS does not consider only gathering ear image or listening lungs for controlling repetitive ear infections or lung diseases in children; it also takes images of throat/tonsil and measures fever of the child and guides parents on how to gather such ehealth data from their child’s body during upper respiratory tract infection stages.

Depending on such kind of practices, it is observed that mobile phones have an important contribution to patients’ easy access to information and services of health services.Besides, the most functional side of mobile applications is their help to a better communication between doctors and patients. Getting appointments, access to health information, documentation of medical records and issues such as decision making get easier.

In future, such mHealth systems may ease appropriate health tracking for individuals by accessing to annotations of their personal health data through ontology knowledge bases.

The proposed system’s mobile application will enable consumers to use an interface as a web service to perform Mobile Pediatric Consultation and Monitoring System (mPCMS) transactions online. The system uses its own children disease ontology knowledge base that is called Children Disease Ontology (CDO). Following sections cover the system mechanism, the ontology knowledge base, implementation of the system, a case study and finally conclusions. In the next section, the detailed description of system’s ontology knowledge base and the working mechanism of the system are discussed.

2 http://www.stethocloud.com/howitworks.html

II. PEDIATRIC CONSULTATION AND MONITORING SYSTEM (MPCMS)

A. System Mechanism The mPCMS has four main procedures that are used by

parent, namely, taking instant images of throat/tonsil and ear inflammation, measuring instant fever and listening to child’s instant voice of lungs during inflectional term. The system is a kit that contains some peripheral parts to gather these instant health data such as a digital thermometer, a pediatric otoscope, and a digital stethoscope compatible with smart phones. The system analyses the gathered health data from child in a smart way and then suggests useful next medical steps for parents to follow securing doctor's permission. These may be how to apply cold therapy duringhigh fever, finding closest pharmacy or hospital, alarm stage-contacting doctor and rearranging medicine usage, guiding to understand the urgency of the situation and improvement in progression of the child health in monitoring stage.

The mPCMS is complete with required peripherals and an ontology-based software application that is designed based on semantic search, matching and inference techniques. User interfaces as well being driven for ease of use, catering to user needs and for use anywhere / anytime; it enables consumers to use an interface as a web service to perform its transactions online. mPCMS application provides authentication to a member consumer who can subsequently upload gathered health data to system database; then the system analyses the current and previous data in semantic ways suggesting the family what to do. The possible courses of advise may be: take an appointment from your doctor, contact the doctor immediately providing the symptom data, contact the doctor for suggestion remotely through the system etc… The consumers get a report from these gathered health data that also indicates the urgency of the situationand improvement progression in monitoring stage.

Firstly, a parent need subscribe to the mPCMS services which involves getting the kit and registering his/her child to the system. Then, the parent should enter login information and patient card of his/her child such as; name, age, weight, etc… The parent chooses a registered doctor and starts to send gathered ehealth data from his/her child at any time. While gathering data from his/her child, the parent will see a menu interface of four main procedures of the system: taking images of throat/tonsil and ear, measuring fever and listening to their lungs during diagnosing or monitoring stages especially in the case of upper respiratory tract infections. In following these procedures, parents will be able to use the peripherals, in taking the images of inner ear and throat/tonsil that doctors can use to make a diagnosis and monitoring. Also, parents will be able to use a digital thermometer attached to their smartphone through a short cable or Wi-Fi to measure instant fever of the child. Lastly, parents will be also able to use a digital stethoscope attached to their smartphone through a short cable to get instant record of the child’s lungs. The mPCMS kit includes a user manual to guide parents while gathering data during stages of upper respiratory tract infections.

355

Page 3: [IEEE 2014 IEEE 38th International Computer Software and Applications Conference Workshops (COMPSACW) - Vasteras, Sweden (2014.7.21-2014.7.25)] 2014 IEEE 38th International Computer

The mPCMS checks instant fever data with previous fever data to understand current situation during monitoring stage. To do this, system compares instant fever data with previous fever data and runs its own semantic rules on ontology according the type of disease. In order to perform this comparison, the system needs to perform reasoning with the concepts and relationships defined in the Children Disease Ontology that is discussed in next section. The comparison process of retrieved concepts from the ontology is performed by Inference Engine of the system (Fig. 1).

Children Health DatabaseChildren InformationHealth InformationFever ResultsRequirements........

2

1

USER

CHILDRENHEALTH

DATABASE

CHILDRENDISEASE

ONTOLOGYKNOWLEDGEBASE

1

Children Disease Ontology KnowledgebaseChildren Identification InformationChildren Diseases and their symptomsSymptoms of fever in children diseases characterized........

2

INFERENCE ENGINE THROUGH SWRL RULES

Concepts of Disease Groups +Children Patient Gathered Fever Data= Suggestions

3

RELAXGREEN

URGENTRED

RISKYYELLOW

TEMPRATURE

OC

RESULTS

4

Fig. 1. System working mechanism through the Inference Engine

B. Children Disease Ontology (CDO) The ontology involves symptoms of the lower and upper

respiratory tract infections caused by viruses or bacteria. Some of those symptoms are created as object type properties in ontology such as: nasal discharge, ear discharge, fever, sneezing besides coughing, headache, nasal discharge, sore throat-ache, watery eyes and ocular discharge, muscle pain, weakness, and anorexia etc. The aim of Children Disease Ontology (CDO) is to represent an abstract model of the different types of lower and upper respiratory tract infections diseases which contains symptoms information including the types and ranges offever and the recommended steps according to fever ranges.In other words, the ontology enables to represent children

characteristics, lower and upper respiratory tract infections diseases and symptoms, and fever ranges in a semantic way.

The ontology is used by the proposed system in inferencing:

- the list of instant and previous temperature health data of the child;

- the list of fever values and symptoms of different types of lower and upper respiratory tract infections diseases and the fever range information according to the disease (from ontology).

Fig. 2. A part of Children Disease Ontology (CDO) of the mPCMS on the Protégé editor.

Protégé editor provides to generate automatically the CDO in the format of OWL 2.0. Ontological structure of the CDO and its semantic rules are created by using Protégé editor3 [4] in the form of SWRL4 (Semantic Web Rule Language). The semantic children disease contexts are described in the CDO using OWL semantic tags such as <owl:class>, <rdfs:subClassOf>,<owl:DatatypeProperty>, and <owl:ObjectProperty>. Table 2 and Fig.2 depicts a portion of the Children Disease Ontology (CDO) in the OWL 2.0 form that keeps many Classes and ObjectProperty, AnnotationProperty and DatatypeProperty properties about the children disease domain such as "Respiratory_System_Disorder", "Finding", and “Patient", "ObservationalResults" etc. The

3 http://protege.stanford.edu/download/registered.html#p4.14 http://www.w3.org/Submission/SWRL/

356

Page 4: [IEEE 2014 IEEE 38th International Computer Software and Applications Conference Workshops (COMPSACW) - Vasteras, Sweden (2014.7.21-2014.7.25)] 2014 IEEE 38th International Computer

"Respiratory_System_Disorder" contains also various subclasses "Pheumania", "Chronic_Lung_Disease", "Emphysema_of_Lung" etc. The ontology also contains many properties that are declared as "hasBacteriaName", "hasSymptom", "hasFeverRange", "hasSynonym""hasComplication" and so on. properties of the domain.

The CDO structure is inspired by the universal research project Bioportal5. Bioportal’s PEDTERM6 ontology contains the terms associated with pediatrics, representing information related to child health and development from pre-birth through 21 years of age; contributed by the National Institute of Child Health and Human Development. Table I displays particulars of PEDTERM ontology.

TABLE I. BIOPORTAL’S PEDTERM ONTOLOGY

Number of classes: 1771 Number of individuals: 0 Number of properties: 7 Maximum depth: 9 Maximum number of children: 50 Average number of children: 4 Classes with a single child: 125 Classes with more than 25 children: 7

The PEDTERM ontology is considered as an upper or base ontology of the client ontologies such as our CDO. However, the CDO is combined the PEDTERM ontology and expanded with more new added properties according to the mPCMS procedures such as "hasBacteriaName", "hasSymptom", "hasFeverRange", "hasSynonym""hasComplication" and so on. properties of the domain. There a lot of ObjectProperty, AnnotationProperty and DatatypeProperty properties according to the functional structure of the proposed system.

TABLE II. ONTOLOGY OWL SYNTAX IN CHILDREN DISEASE ONTOLOGY (CDO) FILE

<?xml version="1.0"?> <Ontology xmlns="http://www.w3.org/2002/07/owl#" xml:base="http://www.semanticweb.org/ontologies/CDO" <Declaration> <Class IRI="#Respiratory_System_Disorder"/> <Class IRI="#Finding"/> <Class IRI="#Patient"/> <Class IRI="#ObservationalResults"/> <Class IRI="#Chronic_Lung_Disease"/> <Class IRI="#Pheumania"/> <ObjectProperty IRI="#hasComplication"/> <ObjectProperty IRI="#hasFeverRange"/> <ObjectProperty IRI="#hasSymptom"/> </Declaration> <SubClassOf> <Class IRI="#Pheumania"/> <Class IRI="#Respiratory_System_Disorder"/> </SubClassOf> <SubClassOf> <Class IRI="#Chronic_Lung_Disease"/>

5 https://bioportal.bioontology.org/6 https://bioportal.bioontology.org/ontologies/PEDTERM

<Class IRI="#Respiratory_System_Disorder"/> </SubClassOf> <SubClassOf> <Class IRI="#Emphysema_of_Lung"/> <Class IRI="#Respiratory_System_Disorder"/> </SubClassOf> <SubClassOf> </Ontology>

C. Inference Engine (IE) The mPCMS checks instant fever data with previous

fever data gathered from same child to understand his/her current situation during monitoring stage. To do this, system compares last fever data with previous fever data and runs its own semantic rules on ontology according the type of disease. In order to perform this comparison, the system needs to perform reasoning through the concepts and relationships defined in the Children Disease Ontology. The ontology is used in inferencing stage to extract:

- the list of instant and previous fever data in Celsius of the child from database.

- the list of fever values and symptoms of different types of lower and upper respiratory tract infectious diseases and the fever range information according to the disease (from ontology).

In order to perform this inference task, the system needs to operate some inferencing rules according to defined the concepts and relationships defined in the CDO (Fig. 1).

Via Ontology-data integrity-data Identification

.

.

…CBABacterialPhn

.

.

…CBAViralPhn

OntologyQuerying

PhenomeniaA B C

.

.

.

.

A B

A

InfectionIs_a

Cough, Fever, Rigors etc.

hasSymptomsIs_a

Is_a

hasFeverRanges

39.0-42

Fig. 3. Semantic Matching and Inferencing

For the concept of Phenomena health problem is related to Fever concept through the hasFeverRanges property (Fig. 3). The value of hasFeverRanges property is 39.0-42.0 Celsius that appears in the CDO. Then, the proposed system starts to match similar findings in the historical record of the child patient’s health card. If doctor diagnosed Phenomena and the hasFeverRanges property is outside the ranges 39.0-42.0 Celsius, then next medical step or suggestions should be considered. Consequently, these semantic descriptions can be retrieved from domain ontologies then send them to the system to guide understand the urgency of the current situation and improvement progression of the child health in monitoring stage.

357

Page 5: [IEEE 2014 IEEE 38th International Computer Software and Applications Conference Workshops (COMPSACW) - Vasteras, Sweden (2014.7.21-2014.7.25)] 2014 IEEE 38th International Computer

D. Case Study The following scenario discusses: a parent login to the

mPCMS application to upload instant health data to the system database through WiFi connection for checking these data by his/her doctor distantly. After login to the mPCMS app, parent will see his/her child health records and agathering data menu. Before gathering data, the parent needs to connect required peripheral parts of the mPCMS app to take instant samples. Then, the app menu is asking to the parent below needed procedures/steps to do by family.

1. Lung listening - Digital stethoscopes (smart phone compatible) - Audio recording and upload audio from the gathering data menu (depicted in Fig. 4).

2. Tonsil image - In front of the camera lens adapter or via otoscope compatible with smart phones and upload image from the gathering data menu (depicted in Fig. 4).

3. Ear image - pediatric otoscope compatible with shelf of the smart phones and upload image from the gathering data menu (depicted in Fig. 4).

4. Digital thermometer - pediatric digital thermometer compatible and connect to the smart phones and upload the temperature data from the gathering data menu (depicted in Fig. 4).

STETHOSCOPEConnected to smart phones

via cable.

THERMOMETER Connected to smart phones

via cable.

OTOSCOPE+ CAMERAoff-the-shelf component for smart

phones.

OTOSCOPE+ CAMERAoff-the-shelf component for smart

phones.

Fig. 4. The Mobile Pediatric Consultation and Monitoring System (mPCMS)

Currently, mPCMS app applies reasoning only on fever data during inferencing phase. The gathered previous and last measured fever data collection are applied on the some inference rules in ontology by considering the patient child characteristic data (age, height, weights, previous diseases etc.) and current existing disease. Here, these kind of data are converted to semantic declarations to run inferring mechanism of the system through its SWRL rules. The aim of this is that the existing fever data knowledge will provide another meaningful ehealth data such as expected symptoms or fever ranges, and suggestions as highlighted earlier. The historical fever data is kept as graphical format that is a

report from entire gathered health data that gives the information of urgency of the situation and improvement progression in monitoring stage. The Pellet [5] supports reasoning with SWRL rules. Pellet interprets SWRL using the DL-Safe Rules notion which means rules will be applied to only the named individuals in the ontology. The SWRL rules are applied to inferring the colored results to understand current fever state of the child. However, the issue will be detailed in our next article.

E. Tools Employed

SYSTEM WEB SERVICES

FOR PARENTS AND DOCTORS

mPCMS SYSTEM PANEL

US

ER

IN

TE

RFA

CE

CH

ILD

RE

N D

IS

EA

SE

O

NT

OLO

GY

ON

PR

OT

EG

E

OW

L A

PI,P

ELLE

T A

PI

US

AG

E F

OR

JA

VA

WE

B

SE

RV

IC

ES

CO

DIN

G

AS

E

E

Fig. 5. Semantic Web used tools and programming stages of the mPCMS.

For the ontology knowledge bases of Mobile Pediatric Consultation and Monitoring System (mPCMS) are semantic-based with OWL used in creating domain ontologies of the field, it eases semantic search and inference. In developing ontology, the Protégé 4.1 [4] with OWL 2.0 support is the preferred tool. As Java7 programing language is used in the functional architecture of the system,inference over the ontology knowledge bases is through OWL API ver3.4.10 (2014-01-18) (Java based Ontology Parser) [6]. Important algorithms of the methodology base of the system are as follows:

- Jaro-Winkler distance [7]

- ‘Semantic Matchmaking’ algorithm [8-10]

- Natural Language Processing (NLP) approaches8,

and

- Regular-expressions9.

These algorithms are used in the Step 2 programming stage in the Fig. 5. Based on these methods, the concepts of children instant symptoms and instant fever information are matched with existing symptoms and fever ranges according

7 http://java.sun.com/products/archive/j2se/6u7/index.html 8 http://en.wikipedia.org/wiki/Natural_language_processing 9 http://en.wikipedia.org/wiki/Regular_expression

358

Page 6: [IEEE 2014 IEEE 38th International Computer Software and Applications Conference Workshops (COMPSACW) - Vasteras, Sweden (2014.7.21-2014.7.25)] 2014 IEEE 38th International Computer

to disease’s type through children disease ontology (semantic matching) and then the returned result guides to understand the urgency of the situation and improvement progression of the child health in monitoring stage (three states: Red (urgent), Yellow (risky-continue controlling), Green (relax)-depicted Fig. 1). It identifies the current situation based on the previous fever values and symptoms to parent and doctoras a report before taking the next medical step decision by his/her doctor. The semantic matching approach is similar to that of some recent studies of semantic service search in accordance with the users’ needs [8-10].

III. CONCLUSION

This paper describes a Mobile Pediatric Consultation and Monitoring System (mPCMS) through Semantic Web technology. Only, the mPCMS is gathering four different types of health data for both consultation and monitoring stages; measuring fever, taking throat/tonsil and ear images, listening to child lungs has not been recommended procedures. This type of pediatric consultation and monitoring in a mobile environment should be useable wherever children may be, such as dormitory, nursery, and hospital or school environment. In this project, images of the ear drum and tonsil via a otoscope, last measured fever degree in Celsius via a digital thermometer, and the audio of listening lungs via a digital stethoscope compatible with smart phones are gathered from the patient child body that are sent in digital to the doctor who can assess the grade of infection and decide next medical steps or if the child needs medical attention or not.

The system contains its own ontology knowledge base, that is, the Children Disease Ontology (CDO). The knowledge base involves symptoms of the lower and upperrespiratory tract infections caused by viruses or bacteria.

Currently, the system’s inferencing mechanism considers only the historical and last fever data of patient and the patient child’s characteristic data (age, height, weights, previous diseases etc.) and current existing disease. In future, we planned to develop the system, especially the system’s inferencing mechanism that will consider entire gathered four different health data (throat, tonsil and ear images and audio file-voice of the child’s lungs) through image and voice analysis in semantic way to suggest next medical steps to both doctor and patent’s family.

ACKNOWLEDGMENT

The project presented in this paper is submitted to TUBITAK10: The Scientific and Technological Research Council of Turkey – 1501 for funding program, ‘Mobile Pediatric Consultation and Monitoring System’) cooperated with Semantica Internet and Software Services Trd. Ltd. Co.11 and The Acıbadem Hospitals Group.12

10 http://www.tubitak.gov.tr/en11 http://www.semantica.com.tr/en12 http://www.acibademinternational.com/

REFERENCES

[1] Berners-Lee, T., Hendler, J., and Lassila.,O. (2001).The Semantic Web, Scientific American, 284(5) 34-43.

[2] Gruber, T. (2007). What is an ontology? Last accessed on January 30, 2014 from, http://www-ksl.stanford.edu/kst/what-is-an-ontology.html.

[3] OWL 2.0, Web Ontology Language Overview, W3C Recommendation, Online: http://www.w3.org/TR/owl2-overview/,last visited: March 2014.

[4] Protégé OWL Ontology Editor, Protégé 4.1 tool website, Stanford University. http://protege.stanford.edu/, last visited: March 2014.

[5] Sirin E, Parsia B. PELLET. An OWL DL Reasoner. In: International Workshop on Description Logics (DL2004), Whistler, 2004.

[6] OWL API (for OWL 2.0), http://owlapi.sourceforge.net/, last visited: March 2014.

[7] Winkler, W. E. (1999). The State of Record Linkage and Current Research Problems. Statistics of Income Division, Internal Revenue Service Publication R99/04. Available from, http://www.census.gov/srd/www/byname.html.

[8] Paolucci, M. (2002). Semantic Matching of Web Service Capabilities, Springer Verlag, LNCS, International Semantic Web Conference.

[9] Çelik, D., and Elçi, A. (2011). Ontology-based Matchmaking and Composition of Business Processes, Book Chapter on Semantic Agent Systems-Foundations and Applications (SASFA), Atilla Elçi, Mamadou T. Kone, and Mehmet A.

Orgun (Editors:), V. 344 in Studies in Computational Intelligence by Springer-Verlag, pp: 133-157.

[10] Çelik, D., and Elçi, A. (2013). A Broker-Based Semantic Agent for Discovering Semantic Web Services through Process Similarity Matching. Science China Information Sciences Volume 56 Issue 1 pp:012102:1–012102:24. http://link.springer.com/article/10.1007/s11432-012-4697-1

359