Journal of Advanced Research in Dynamical and Control Systems, Volume 11, Issue 2, March 2019

ISSN 1943-023X

An Ontology based system for health care people to

prevent cardiovascular diseases
Dr. B R Prakash
Assistant Professor, Department of Master of Computer Applications,
Sri Siddhartha Institute of Technology, Tumkur, INDIA.
[email protected]

Abstract
Nowadays the health care providers plays a very important role in a human health care. Currently the
health issues are addressed in traditional hospital systems by conducting several investigations to
predict the type of diseases like diabetes mellitus, cardiovascular diseases, nephrological diseases, etc.
This investigations doesn’t provide any early significance of human health care system. As being a
human, have a chance of occurring Cardiovascular disease, it is one of the most common disease
worldwide so, it may be considered has a main cause of death. The proposed model will predict the
human health status based on activities performed by him to prevent cardiovascular diseases. So, in
case of health care centre to represent current health care status using Social Networks, having
different conventional methods, in that ontology is one among them. WordSet is the source for
ontology where the information is present. These information’s are presented in the deep semantic
web, these are considered as input to determine the cardiovascular health status based on activity of a
person shared in online Social Networks allowing access between persons and places.
Keywords: Semantic Web, Deep learning, Ontology, Social Networks, Cardiovascular System.

1. Introduction
Deep learning belongs to the class of machine learning methods and concepts. It is a special variation
of representation based learning [1]. In recent years, deep learning sets a new trend in machine
learning, it plays an important role in diagnostic applications and became part of the artificial
intelligence diagnostic process .Deep learning methods are representation-learning algorithms at
multiple levels, with higher level learned features defined in terms of lower level features[2][3].The
objective of deep learning is to discover more abstract features in the higher level of the
representation, which can be easily separated from various explanatory factors extent in the data [4].
Ontology plays a vital role by contributing the common understanding of knowledge between the
people who communicate and the applications that make use of the ontology knowledge [5].The
problem of constructing natural ontologies can be minimized by combining deep web information
with ontology resulting in semantic deep web [6].The development of intelligent healthcare systems
always requires a formulization of medical knowledge. Domain ontologies are suitable for this
purpose but their construction is manually addressed in most of the cases. This results in long and
tedious development processes that halter their real applicability. This is why we need of ontology
learning methods that aid ontology construction process. Ontologies are used to represents explicit
specification of knowledge. Domain knowledge of interest are in the form of concepts, individuals
who are belonging to these concepts and relationships among between the concepts and between
concepts and individuals [7][8].
In recent scenario, heart is one of the most important organ in human body. The term heart disease is
often used interchangeably with the term cardiovascular disease. Cardiovascular disease is one of the
most common diseases that are the leading cause of death in developed countries [9]. Globally, the

77
Journal of Advanced Research in Dynamical and Control Systems, Volume 11, Issue 2, March 2019
ISSN 1943-023X

number of deaths from cardiovascular disease and circulatory disease has increased by one third
between 1990 and 2010 [10].
Heart disease is one of the most common disease globally which is considered as the leading cause of
death [11]. Heart disease describes a wide range of conditions that affect the heart. Heart disease
includes diseases like blood vessel diseases, such as coronary artery disease, heart rhythm problems,
and congenital heart defects, etc. Cardiovascular disease is a class of disease that is caused by
narrowed, blockage of blood vessels that prevent the heart, brain or other parts of the body from
receiving enough blood. Cardiovascular disease symptoms in women may be different from those in
men. Women are more likely to have symptoms such as shortness of breath, nausea and extreme
fatigue whereas men are more likely to have chest pain [6].
Ontology-based decision support system designed to handle and treat the patients who are all affected
by the cardiac disorder. This system analyses the patient’s condition and suggest a treatment to
achieve the fastest possible recovery. The knowledge base in this system consists of an OWL
ontology and a set of SWRL rules and this approach provides supervision and well treatment of
critical patients who are affected by cardiac disorders [12].

2. Literature Survey
In recent scenario, health care becomes one of the most important concerns. This paper presents an
ontology called Cardiovascular Disease Ontology (CVDO), which is structured on the principle of
OBO Foundry and based on Basic Formal Ontology (BFO) and Foundational Model of Anatomy
(FMA). The CVDO aims at identifying and completing DOID Cardiovascular Disease classes on the
base of general model of disease called Ontology for General Medical Science (OGMS) and OGMS
needs should be supplemented by certain methodological rules in order to discover the end point of a
disease course and the material basis of the disease should be selected in the casual chain disorders
[13].
Ontologies are used to represent the explicit specification of knowledge in specific domain knowledge
of interest in the form of concepts and relations among them. This paper proposes a Ontology called
Medical Ontology, which describes the domain of heart failure (HF). The aim of constructing
Ontology for a domain like heart failure is identified as an important step in systematization of
existing medical knowledge. This paper proposes that the current progress of heart failure Ontology is
one of the main research results of the EU Heartfaid project. The development of heart failure
Ontology is done mainly by technical people by referring medical literature, guidelines of heart failure
published by European Society of Cardiology. The Ontology presents a detailed study of the Heart
failure domain and also a useful framework that support for building knowledge based system in the
heart failure domain [14].
Heart disease is one of the most common disease globally which is considered as the leading cause of
death. The cardiologists in Palestine says that at a rate of 27.5% of all deaths, the heart disease was
the main cause of death among the Palestinians population. The patient’s symptoms are not enough to
give an accurate diagnosis since several types of heart diseases have the similar symptoms. Hence the
results of several clinical tests are required to recognise the differential diagnosis between different
types of heart disease. Currently the domain of heart disease in Palestine does not contain any specific
system and also semantic approaches are not employed in the available medical system they are just
dependent on only database-oriented methodologies which are not flexible, lack intelligence, etc.
Hence this work aim is to improve the diagnosis of heart disease through Semantic Web technologies.
When a proposed system was tested using a sample set of patients(30 patients) with heart disease, the
result generated from the system have shown that the system have correctly diagnosed 27 patients out
of 30 patients by reaching the accuracy of 90%. The main contribution of this research is done

77
Journal of Advanced Research in Dynamical and Control Systems, Volume 11, Issue 2, March 2019
ISSN 1943-023X

through using Ontology and SWRL rules to diagnose all types of diseases that are related to hear. An
Ontology called domain Ontology (HeartOnt) has been built which covers the domain knowledge of
heart diseases [6].
Heart disease describes a wide range of condition that affects the heart. This paper proposes a system
for assisting remote ECG diagnosis and this system consists of performing painless non-invasive
medical exams using a portable ECG device. ECG makes use of signal processing tools and also
describes various methods for removing noise on ECG signals. ECG are performed when a patient
complains of light-headedness, palpitations, or Syncope since abnormal heart ability to pump blood
and provide the body with oxygen. The classification between normal and abnormal heart beats are
performed by a Multilayer Perception Neural Network [15].
Health Monitoring System plays an important role with potential to change the way that the healthcare
is currently delivered. This paper proposes a development and design of GSM (Global System for
Mobile) based smart wearable system with 3-axis accelerometer, 3-lead ECG recording system and
real-time NIBP analysis system. The system is capable of detecting sudden fall signals, hypotension
and cardiac abnormalities and thereby which is suitable for real-time monitoring and all types of
diagnosis purposes. The system is even capable of sending the notifications to the healthcare experts
soon after detecting any of the health abnormalities in order to take necessary steps to overcome the
abnormalities detected and this is the main objective to design and implement a low cost wearable
health monitoring system. A software platform: NI LabVIEW is used to perform the complete
analysis and evaluation of the signals acquired from the patient since the wearable technologies plays
an important role in recognizing transient activity-related features in order to predict risk of
cardiovascular events and also to help those who residing in remote locations and incapable of
affording expensive healthcare facilities in urban hospitals [16].
Recent evolution which have been taken place in the area of deep learning have been proved that it is
achieving tremendous progress in the field of computer vision and extremely favourable for several
natural language processing tasks like sentiment analysis, question answering, and machine
translation and other fields of machine learning [17].
In recent scenario, the importance of deep learning has spread beyond both academia and industry
with several inspiring real-world applications since it is a new, timely and promising area of machine
learning. Also the significance of deep learning are being enlarged into the other fields like Social
media [P18][P19], Social network analysis [20], bioinformatics [21], medicine and healthcare [22].
Deep learning is a subset of machine learning and deep learning is a representation of multiple levels.
The concept in deep learning algorithm is automatic extraction of complex data representation at
higher levels of abstraction. And this how deep learning is more significant for addressing some
important issue in Big Data Analytics, including extracting complex patterns from massive volumes
of data, fast information extraction, simplification of discriminative tasks, etc. [23].
The Semantic Web is an extension of the World Wide Web which is used to make search
automatically and also that enables machines to analyse and respond to human requests on the basis of
their meaning since Semantic Web is concerned with the meaning. The relevant information is
retrieved from the web by making the system to understand without any human involvement [24].
The Semantic Web aim is to automate several tasks performed by the humans with World-Wide Web
today [5]. The Semantic Web main aim is to add logic to the current web [6]. Several existing
applications and research projects can be seen in the medical and health sciences based on Semantic
Web technologies. Semantic Web could plays a very important role in information retrieval of
biomedical vocabularies, terminologies, and taxonomies [25].

77
Journal of Advanced Research in Dynamical and Control Systems, Volume 11, Issue 2, March 2019
ISSN 1943-023X

In the field of computer science, Ontologies are emerged in different concepts. Object and relation
between the objects are represented in specific domain in Ontology [26]. Ontology is one of the
tremendous technologies derived from WordSet where information can be retrieved from Semantic
Web for the cross reference of the given input regarding health social activity of a person based on the
posts updated in online social networks allowing access between person and places [27].

3. Model Description
3.1 System Architecture
The architecture of the proposed system consists of following components as depicted in figure 1.
This figure provides the interaction between all these componetns. This is one of the expert system
like others, knowledge base is the core of this system which includes fact base and rule base. The fact
base is extracted using user interface from the deep semantic web based on the physical activities they
perform and that information is posted in any social media.
The rule base consists of SWRL rules and the ontology classes. The ontology classes give the
structure of classes along with the relationships between the classes, where the ontology classes are
build using Protѐgѐ ontology editor.
The inference engine is the core of the expert system which depends on the facts and the rules to
reason the required decision. In this paper Pallet [29], is used as OWL-DL reasoned, where it consists
of several features like data-type reasoning and debugging, integration of rules, reasoning conjunctive
queries. More decision rules could be inferred and added to the list of available rule base in this
section. The final decision results will be introduced to the user through the user interface alongside
with the explanation about this decision inferred from the explanation module.

Figure 1: The System Architecture

3.2 System Methodology

METHONTOLOGY[30], is one of the methodology which is considered as most comprehensive
ontology engineering methodology among several methodologies presented to building and
developing an ontology-based expert systems.

77
Journal of Advanced Research in Dynamical and Control Systems, Volume 11, Issue 2, March 2019
ISSN 1943-023X

Figure 2: Life-cycle of Ontology development

As depicted in the Figure 2, the ontology development process involves phases as illustrated. In all
these phases some of the phases are independent, and most of them are interfere with each other and
can be operate simultaneously. The proposed domain ontology depends on these phases and the
development process as follows:
The first phase is to determine the purpose of building the ontology, its formation and its scope. The
purpose of this domainOWL is to be a knowledge representation to prevent the cardiovascular
diseases by doing the physical activites. Since it is expressed in a formally defined language(OWL),
hence its degree of formality is “semi-formal”. This paper proposes a system can be used by the end
user to prevent the cardiovascular diseases based on the reasoning process embedded in the ontology.
The knowledge capturing phase is the beginning of the design by capturing the knowledge required to
build a domainOWL. Here structured and non-structured requirements, concepts, object properties,
data properties and the relations between those are determined. The main ontology structure and its
concepts are gather by analysing formal and informal text from the deep semantic web.
The knowledge conceptualization phase includes the concepts and properties used in designing the
ontology. These are useful to construct the class hierarchy and to determine the properties and the
relations between these classes.
Integration stage specifies that if the developed ontology can have benefit from other existing
ontologies. As shown in table 1 it has some classes and some of its instances.

Concept Name Instances Value

Laying Yes/No
Sitting Yes/No
Regular Movements Standing Yes/No
Ascending stairs Yes/No
Descending stairs Yes/No
Normal walk Yes/No
Rare Movements
Nordic walking Yes/No

77
Journal of Advanced Research in Dynamical and Control Systems, Volume 11, Issue 2, March 2019
ISSN 1943-023X

Running Yes/No
House cleaning Yes/No
Vacuuming Yes/No
Housekeeping work
Ironing Yes/No
Folding laundry Yes/No
Cycling Yes/No
Driving Bike riding Yes/No
Car driving Yes/No
Carom Yes/No
Mental ability games Chess Yes/No
Board games Yes/No
Badminton Yes/No
Tennis Yes/No
Physical games Cricket Yes/No
Volleyball Yes/No
Football Yes/No
Table 1: DomainOWL conceptualization table
Implementation phase is responsible to implementing the ontology in one of the available ontology
editors. This domainOWL is creating using Protѐgѐ 4.3 as editor which has the ability to check the
lexical and syntax errors. It is integrated with several types of reasoners which guarantee the
completeness, consistency, and not redundancy in the defined ontology. Figure 3 shows the class
hierarchies in an OWL ontology to be viewed and incrementally navigated, allowing comparison of
the asserted class hierarchy and the inferred class hierarchy.
Evaluation phase is one it enhances the developed ontology over time and requirement of the end
user. Documentation phase is going in parallel with all other phases where all the activities happened
to develop the ontology and all its aspects are maintained.

Figure 3. Asserted class hierarchy and the inferred class hierarchy produced by Protѐgѐ

77
Journal of Advanced Research in Dynamical and Control Systems, Volume 11, Issue 2, March 2019
ISSN 1943-023X

In figure 3, where it depicts the asserted class hierarchies in that the thing class is referred as main
parent class of that hierarchy in Protѐgѐ. It refer person in out model, a person may be male or female.
The subclass regular movement, rare movements, physical games, mental ability games,
housekeeping works, driving are represents the various activities carried out by a person. The siblings
of these classes are instances of the activities, where all of these are associates with stipulated fixed
time interval. Some of the calories are burn off when these instances of activities are performed, from
this the cumulative result is determined. Then that cumulative result is divided by the total calories
form all those activities an inference is drawn. Figure 4 show the tree structure build by OWLViz
plug-in with in Protѐgѐ.

Figure 4. DomainOWL tree structure produced by Protѐgѐ

4. Discussion
In daily routine people are participated in several activities in those activities some of the activities are
considered in table 1. The prediction is when a person performed an activity for a stipulated time
period a specific amount of calorie is required; it may vary from person to person. The activities are
carried out for some days or months those are predicted as follows.
In this discussion D represents days and A represents activity, there are discrete in nature and they are
known as discrete random variables.
Here, D and A are two discrete random variables defined in joint distribution probability, its function
is given by
𝑃(𝐷 = 𝑑, 𝐴 = 𝑎) = 𝑓(𝑑, 𝑎) 𝑒𝑞𝑛(1)
Where it satisfies these conditions
1. 𝑓(𝑑, 𝑎) ≥ 0

2. ∑ ∑ 𝑓(𝑑, 𝑎) = 1
𝑑 𝑎

Suppose D can assume any one of m values like d1,d2,d3.......dm and A can assume any one of n values
like a1,a2,a3..........an. Then the probability of these events are given as

77
Journal of Advanced Research in Dynamical and Control Systems, Volume 11, Issue 2, March 2019
ISSN 1943-023X

𝑃(𝐷 = 𝑑𝑖 , 𝐴 = 𝑎𝑗 ) 𝑒𝑞𝑛(2)

The probability that D=di is obtained by adding all entries corresponding to di is given as
𝑛

𝑃(𝐷 = 𝑑𝑖 ) = 𝑓1 (𝑑𝑖 ) = ∑ 𝑓(𝑑𝑖 , 𝑎𝑗 ) 𝑓𝑜𝑟 𝑗 = 1,2,3 … . 𝑛 𝑒𝑞𝑛(3)

𝑗=1

Similarly the probability that A=aj is obtained by all the entries in corresponding to aj is given as
𝑚

𝑃(𝐴 = 𝑎𝑗 ) = 𝑓1 (𝑎𝑗 ) = ∑ 𝑓(𝑑𝑖 , 𝑎𝑗 ) 𝑓𝑜𝑟 𝑖 = 1,2,3 … . 𝑚 𝑒𝑞𝑛(4)

𝑖=1

The eqn(3) and eqn(4) can be written as

𝑚 𝑛

∑ ∑ 𝑓(𝑑𝑖 , 𝑎𝑗 ) = 1 𝑒𝑞𝑛(5)
𝑖=1 𝑗=1

The eqn(5) is simply the statement that the total probabilities of all entries is 1.
The obtained values are lies in joint distribution, where joint distribution that occurs naturally in many
discrete situation.

Conclusion
In the field of computer science ontologies affects the development of various types of systems like
expert systems. Here activity information of person is one of the data that to be analysed to determine
the hidden information and extract the knowledge to improve the cardiovascular health of a person.
By applying data mining algorithms ontology based health care system to prevent cardiovascular
diseases to be implemented. It can be serve as a assistant tool for general health care practitioners,
fitness trainers to make better clinical decisions and fitness activities. To share the system knowledge,
domainOWL was build using Protѐgѐ tool.
References :
[1] Oliver Faust, Yuki Hagiwara, Tan Jen Hong, Oh ShuLih, U Rajendra Acharya, “Deep learning for
healthcare applications based on physiological signals: A review”, Computer Methods and Programs in
Biomedicine, April 2018, DOI: 10.1016/j.cmpb.2018.04.005.
[2] Riccardo Miotto, Fei Wang, Shuang Wang, Xiaoqian Jiang and Joel T. Dudley, "Deep learning for
healthcare: review, opportunities and challenges”, Briefings in Bioinformatics, 2017, 1-11, DOI:
10.1093/bib/bbx044.
[3] LeCun Y, Bengio Y, Hinton G. Deep learning. Nature 2015;521:436–44.

[4] YoshuaBengio, Workshop on Unsupervised and Transfer Learning, “Deep Learning of Representations for
Unsupervised and Transfer Learning”, JMLR: Workshop and Conference Proceedings 27:17–37, 2012.
[5] Kushala. V., Dr.Supriya. M, “Deep Learning Ontology: Dimensions in the Field of Agriculture, A Survey”,
International Journal of Latest Technology in Engineering, Management & Applied Science (IJLTEMAS)
Volume VII, Issue I, January 2018 | ISSN 2278-2540.

[6] Yoo Jung An, “Ontology learning for the semantic deep web”, January 2008.

[7] Hosam Mohammed Alagha, “Diagnosing Heart Diseases Using Ontology and SWRL Rules”, March/2017.

77
Journal of Advanced Research in Dynamical and Control Systems, Volume 11, Issue 2, March 2019
ISSN 1943-023X

[8] Sherimon, P., Vinu, P., Krishnan, R., &Takroni, Y., “Ontology based system architecture to predict the risk
of hypertension in related diseases”, International Journal of Information Processing and Management, 2013,
4(4), 44.

[9] Go, A. S., Mozaffarian, D., Roger, V. L., Benjamin, E. J., Berry, J. D., Blaha, M. J., Franco, S. (2014).
Executive summary: heart disease and stroke statistics--2014 update: a report from the American Heart
Association. Circulation, 129(3), 399.

[10] Lozano, R., Naghavi, M., Foreman, K., Lim, S., Shibuya, K., Aboyans, V., Ahn, S. Y.,(2013). Global and
regional mortality from 235 causes of death for 20 age groups in 1990 and 2010: a systematic analysis for the
Global Burden of Disease Study 2010. The Lancet, 380(9859), 2095-2128.

[11] Lloyd-Jones et al., “Diagnosing Heart Diseases Using Ontology and SWRL Rules”, 2010.

[12] Martínez-Romero, M., Vázquez-Naya, J. M., Pereira, J., Pereira, M., Pazos, A., &Baños, G. (2013). The
iOSC3 system: using ontologies and SWRL rules for intelligent supervision and care of patients with acute
cardiac disorders. Computational and mathematical methods in medicine, 2013.

[13] Adrien BARTON, Arnaud ROSIER, Anita BURGUN and Jean-François ETHIER, “The Cardiovascular
Disease Ontology”, Frontiers in Artificial Intelligence and Applications, September 2014, DOI: 10.3233/978-1-
61499-438-1-409.

[14] AlanJovic, DraganGamberger, GoranKrstacic,“Heart failure ontology”.

[15] Luis A. SoutoMaiorNeto, Robson Pequeno, Carlos Almeida, Katia Galdino, Fabricia Martins, Antonio V.
de Moura, “AmethodforIntelligentSupporttoMedicalDiagnosisinEmergencyCardiacCare”,978-1-5090-6182-
2/17/$31.00 ©2017 IEEE.

[16] SouvikTewary,ShreyosiChakraborty,JoshitaMajumdar,Rahul Majumder,DebasishKundu, Sudipta Ghosh,

Sauvik Das Gupta, “A Novel Approach towards designing a Wearable Smart Health Monitoring System
measuring the Vital Parameters and Emergency situations in Real-Time and providing the necessary Medical
Care through Telemedicine”, 978-1-4673-7918-2/16/$31.00 ©2016 IEEE Students' Conference on Electrical,
Electronics and Computer Science.

[17] GiulioPetrucci, Chiara Ghidini, Marco Rospocher, FBK-irst, “Ontology Learning in the Deep”.

[18]NhatHaiPhan, YueWang, XintaoWu,DejingDou, “DifferentialPrivacyPreservationforDeepAuto-Encoders:

AnApplicationofHumanBehaviorPrediction”, Proceedings of the Thirtieth AAAI Conference on Artificial
Intelligence (AAAI-16).

[19] Yuan, Z.; Sang, J.; Liu, Y.; and Xu, C. 2013. Latent feature learninginsocialmedianetwork.
InACMMM’13,253–262.

[20] Perozzi, B.; Al-Rfou, R.; and Skiena, S. 2014. Deepwalk: Online learning of social representations. In
KDD’14, 701– 710.

[21] Chicco, D.; Sadowski, P.; and Baldi, P. 2014. Deep autoencoder neural networks for gene ontology
annotation predictions. In ACM BCB’14, 533–540.

[22] Song, Y.; Ni, D.; Zeng, Z.; He, L.; et al. 2014. Automatic
vaginalbacteriasegmentationandclassificationbasedonsuperpixelanddeeplearning. JournalofMedicalImagingand
Health Informatics 4(5):781–786.

[23] MaryamMNajafabadi,FlavioVillanustre,Taghi M Khoshgoftaar,NaeemSeliya, RandallWald1and

EdinMuharemagic, “Deeplearningapplicationsandchallengesin bigdataanalytics”, JournalofBigData (2015)
2:1,DOI 10.1186/s40537-014-0007-7.

[24] B sangeetha, R.Vidyapriya, Monikasree R, “Semantic Ontology Mapping using Support Vector Machine”,
International Journal of Computer Applications (0975 – 8887) International Conference on Innovations in
Computing Techniques (ICICT 2015).

77
Journal of Advanced Research in Dynamical and Control Systems, Volume 11, Issue 2, March 2019
ISSN 1943-023X

[25] Lorence, D. P., &Spink, A. (2004). Semantics and the medical web: a review of barriers and breakthroughs
in effective healthcare query. Health Information & Libraries Journal, 21(2), 109-116. Lozano.

[26] BaydaaTaha Al-Hamadani, RaadFadhilAlwan, “An Ontology-Based Expert System for General
Practitioners to Diagnose Cardiovascular Diseases”,Advances in Computational Sciences and Technology ISSN
0973-6107 Volume 8, Number 1 (2015) pp. 53-65 © Research India Publications.

[27] H. R. Divakar, Dr.B.R.Prakash, “Classics of Deep Learning Approach for Human Behaviour Ontology: A
Survey”, International Journal of Computer Trends and Technology (IJCTT) – Volume 51 Number 1 September
2017.

[28] N.Phan et al., 2016. Ontology-based deep learning for human behaviour prediction with explanations in
health social networks, Information Sciences.

[29] Sirin, E., B. Parsia, B.C. Grau, A. Kalyanpur, and Y. Katz, “Pallet: A Practical OWL-DL reasoned”.
Web Semantics: Science, Services and Agents on the World Wide Web, Vol. 5 no 2:pp. 51-53. 2007

[30] Ferndndez, M., A. Gomez-Perez, and N. Juristo, METHONTOLOGY: From Ontological Art Towards
Ontological Engineering, A.T.R SS-97-06, Editor. 1997

77