Doctor of Philosophy - Amazon S3...Medicine submitted by me for thedegree of Doctor of Philosophy is...

ONTOLOGY BASED ONLINE EXPERT SYSTEMFOR EMERGENCY MEDICINE

A Thesis submitted to Gujarat Technological University

for the Award of

Doctor of Philosophyin

Instrumentation & Control Engineering

by

Rutvik Kiritkumar Shukla[129990917006]

under supervision of

Dr. Chetan B. Bhatt

GUJARAT TECHNOLOGICAL UNIVERSITY

AHMEDABAD

Dec – 2019

ONTOLOGY BASED ONLINE EXPERT SYSTEMFOR EMERGENCY MEDICINE

A Thesis submitted to Gujarat Technological University

for the Award of

Doctor of Philosophyin

Instrumentation & Control Engineering

by

Rutvik Kiritkumar Shukla[129990917006]

under supervision of

Dr. Chetan B. Bhatt

GUJARAT TECHNOLOGICAL UNIVERSITY

AHMEDABAD

Dec - 2019

©Rutvik Kiritkumar Shukla

DECLARATION

I declare that the thesis entitled Ontology Based Online Expert System For Emergency

Medicine submitted by me for thedegree of Doctor of Philosophy is the record of research

work carried out by me during the period from September 2012 toDecember 2018 under

the supervision of Dr. Chetan B. Bhatt and this has not formed the basis for the award of

any degree, diploma, associateship, fellowship, titles in this or any other University or

other institution of higher learning.

I further declare that the material obtained from other sources has been duly acknowledged

in the thesis. I shall be solely responsible for any plagiarism or other irregularities, if

noticed in the thesis.

Signature of the Research Scholar : …………………………… Date:….………………

Name of Research Scholar: Rutvik Kiritkumar Shukla

Place : Ahmedabad

CERTIFICATE

I certify that the work incorporated in the thesis Ontology Based Online Expert System

For Emergency Medicine submitted by Shri. Rutvik Kiritkumar Shukla was carried

out by the candidate under mysupervision/guidance. To the best of my knowledge: (i) the

candidate has not submitted the same research work to any other institution for any

degree/diploma, Associateship, Fellowship or other similar titles (ii) the thesis submitted is

a record of original research work done by the Research Scholar during the period of study

under my supervision, and (iii) the thesis represents independent research work on the part

of the Research Scholar.

Signature of Supervisor: ……………………………… Date: ………………

Name of Supervisor: Dr. Chetan B. Bhatt

Place: Ahmedabad

Course-work Completion Certificate

This is to certify that Mr. Rutvik Kiritkumar Shuklaenrolment no.129990917006 is a

PhD scholar enrolled for PhD program in the branch Instrumentation & Control

Engineering of Gujarat Technological University, Ahmedabad.

(Please tick the relevant option(s))

He/She has been exempted from the course-work (successfully completed during

M.Phil Course)

He/She has been exempted from Research Methodology Course only (successfully

completed during M.Phil Course)

He/She has successfully completed the PhD course work for the partial requirement

for the award of PhD Degree. His/ Her performance in the course work is asfollows

Grade Obtained in Research Methodology(PH001)

Grade Obtained in Self Study Course(Core Subject)

(PH002)BB AB

Supervisor’s Sign:__________________

Name of supervisor: Dr. Chetan B. Bhatt

Originality Report Certificate

It is certified that PhD Thesis titled Ontology Based Online Expert System For

Emergency Medicine by Shri. Rutvik Kiritkumar Shukla has been examined by us. We

undertake the following:

a. Thesis has significant new work / knowledge as compared already published or are

under consideration to be published elsewhere. No sentence, equation, diagram,

table, paragraph or section has been copied verbatim from previous work unless it

is placed under quotation marks and duly referenced.

b. The work presented is original and own work of the author (i.e. there is no

plagiarism). No ideas, processes, results or words of others have been presented as

Author own work.

c. There is no fabrication of data or results which have been compiled / analysed.

d. There is no falsification by manipulating research materials, equipment or

processes, or changing or omitting data or results such that the research is not

accurately represented in the research record.

e. The thesis has been checked using <Turnitin Software> (copy of originality report

attached) and found within limits as per GTU Plagiarism Policy and instructions

issued from time to time (i.e. permitted similarity index <=25%).

Signature of the Research Scholar : …………………………… Date: ….………


Place :Ahmedabad

Signature of Supervisor: ……………………………… Date: ………………


Place: Ahmedabad

PhD THESIS Non-Exclusive License toGUJARAT TECHNOLOGICAL UNIVERSITY

In consideration of being a PhD Research Scholar at GTU and in the interests of

thefacilitation of research at GTU and elsewhere, I, Rutvik Kiritkumar Shukla having

Enrollment No.129990917006 hereby grant a non-exclusive, royalty free andperpetual

license to GTU on the following terms:

a) GTU is permitted to archive, reproduce and distribute my thesis, in whole or in part,

and/or my abstract, in whole or in part ( referred to collectively as the “Work”) anywhere

in the world, for non-commercial purposes, in all forms of media;

b) GTU is permitted to authorize, sub-lease, sub-contract or procure any of the acts

mentioned in paragraph (a);

c) GTU is authorized to submit the Work at any National / International Library, under the

authority of their “Thesis Non-Exclusive License”;

d) The Universal Copyright Notice (©) shall appear on all copies made under the authority

of this license;

e) I undertake to submit my thesis, through my University, to any Library and Archives.

Any abstract submitted with the thesis will be considered to form part of the thesis.

f) I represent that my thesis is my original work, does not infringe any rights of others,

including privacy rights, and that I have the right to make the grant conferred by this non-

exclusive license.

g) If third party copyrighted material was included in my thesis for which, under the terms

of the Copyright Act, written permission from the copyright owners is required, I

haveobtained such permission from the copyright owners to do the acts mentioned in

paragraph (a) above for the full term of copyright protection.

h) I retain copyright ownership and moral rights in my thesis, and may deal with the

copyright in my thesis, in any way consistent with rights granted by me to my University

in this non-exclusive license.

i) I further promise to inform any person to whom I may hereafter assign or license my

copyright in my thesis of the rights granted by me to my University in this non-exclusive

license.

j) I am aware of and agree to accept the conditions and regulations of PhD including all

policy matters related to authorship and plagiarism.

Signature of the Research Scholar:


Date: Place:Ahmedabad

Signature of Supervisor:


Date: Place:Ahmedabad

Seal:

Thesis Approval Form

The viva-voce of the PhD Thesis submitted by Shri. Rutvik Kiritkumar Shukla

(Enrollment No. 129990917006) entitled Ontology Based Online Expert System For

Emergency Medicine was conducted on …………….………… (day and date) at Gujarat

Technological University.

(Please tick any one of the following option)

The performance of the candidate was satisfactory. We recommend that he/she be

awarded the PhD degree.

Any further modifications in research work recommended by the panel after 3

months from the date of first viva-voce upon request of the Supervisor or request of

Independent Research Scholar after which viva-voce can be re-conducted by the

same panel again.

The performance of the candidate was unsatisfactory. We recommend that

he/sheshould not be awarded the PhD degree.

-------------------------------------------------- ---------------------------------------------------

Name and Signature of Supervisor with Seal 1) (External Examiner 1) Name and Signature

------------------------------------------------------- -------------------------------------------------------

2) (External Examiner 2) Name and Signature 3) (External Examiner 3) Name and Signature

i

ABSTRACT

In order to provide immediate and urgent care to the critically ill patient, there is an

overwhelming need for efficient and well-prepared guidelines for the emergency service

provider. Specifically, developing countries like India are facing a lack of well trained and

experienced medical professional in the rural area. In addition to this doctor to patient ratio

in these countries are also poor and declining day by day with a rise in population. With

novel and well-defined policies of government has tried to address this issue toa certain

extent. Even though, because of socio-economical condition, geographical situation,

unavailability of well-trained EM staff and lack of medical experts in the rural area, the

people living in this part of the countries are still not getting enough medical facilities.

Computer-basedexpert systems are very well known since a long time for assisting the

individuals in the absence of experts by utilizing the knowledge which they possess. Expert

systems have found application in various fields. The medical field is also an important

sector where these knowledge-based computer systems can prove to be a boon for saving

the lives of people. But these systems depend on the individual competency and

effectiveness of designing the system based on individual perspective. This puts a

limitation of making the whole system rigid and unscalable. This often results in an

inefficient and nonreliable system in terms of decision-making skill. The semantic web is

the new concept introduced by the scientists to make the machine interpretable web which

helps the machine to infer new information based on available information over the web.

Ontologies are the prime component of the semantic web, which is used to model the

available information in semantic form. Ontologies are based on a shared and consensual

domain knowledge agreed by a community. Ontology is expressed by languages known as

OWL (Ontology Web Language). One of the popularways of describing ontology is

expressing it in terms of RDF (Resource Description Format). Ontology provides a way of

expressing data in a generic, extended and integrated form and makes the overall system

flexible and scalable. The system proposed here utilizes the java based client-server MVC

architecture. JENA API is used here for the purpose of integration between knowledge-

base stored in OWL format with JAVA servlet. The additional database related to patient,

disease, and treatment is stored in a MySQL database.This system takes the most basic

vital parameter needed for primary assessment of patient’s condition from the EM

paramedic from client side and passes this information to server side where the primary

ii

risk level stratification will be calculated by the ontology and score is available to the

paramedic. The secondary assessment asks additional parameters to the paramedic and

generates a list of probable disease from the ontology. This system also displays the

interactive steps of carrying out the treatment.

iii

Acknowledgment

Completing a PhD is a tough task that requires hard work and a lot of efforts. This is often

an overwhelming but also great experience which I would not have been able to complete

without the assistance and support of so many people. Thus, it is my great pleasure to

thank all those people.

First of all, I would like to thank almighty for giving me the strength to carry out this

humongous task. I would like to deeply thank Dr. C. B. Bhatt, my supervisor, for his

guidance, encouragement, and support over these years. This research work would not

have been possible without his constructive pieces of advice, his systematic guidance and

patience support thought out this duration of my research work.

I would like to express my sincere gratitude to Dr. Saurin Shah and Dr. J B Patel, my

doctoral progress committee members. Their rigorous style of reviewing and constructive

feedback with valuable suggestion helped me a lot to decide the possible course of action.

I take this opportunity to thank Dr. UdgeethThaker, Bankers multi-specialty hospital,

Baroda. His audacity of utilizing existing technologies for patient care has worked as a

push to my research work. I would also like to thank Dr. Vishal Sadatiya and

management of Shree Giriraj Multi Speciality Hospital, Rajkot for allowing me to access

the patient database. Especially I thank Dr. Vishal Sadatiya, who spent his valuable time

whenever required for discussing the medical aspects of this work and provided relevant

material as well.I would also thank Mr. Nitin Joshi, Executive of telemedicine project at

Apollo hospital, Ahmedabad. In the earlier phase of my research work he helped me to

understand the working methodology of the telemedicine project.

I would like to thank Mr. Parth Modi for helping me to understand and develop the

system. His ability to work continuously and his passion inspired me a lot while we work

together. I would also like to thank Dr. M. K. Shah who motivated me to carry out my

research work and always there to help. I am also thankful to Prof. M. D. Khediya, Ms.

M. V. Patel and other colleagues of the IC department, VGEC, Chandkheda for their

cooperation in every possible means. I would thank specially to my research colleague Mr.

Mihir Dhudhrejiya for his precious support. I would also like to express my sincere thank

to Prof. M. J. Modi, Prof. M. P. Jani, Dr. D. H. Makwana and colleagues of IC

department, GEC, Rajkot for their support and encouragement.

iv

I also like to express my sincere gratitude to my parents Shri. Kiritkumar M. Shukla and

Smt. Jayshreeben K. Shukla without their support this work wouldn’t have been

possible. Also, I want to express my appreciation to my son Hard and wife Mrs. Setu for

their concern and cooperation.

Lastly, I would thank all the people who directly or indirectly helped me during this very

important phase of my life.

Thanking You

Rutvik Kirikumar Shukla

v

Table of Content

Abstract i

Acknowledgment iii

List of Abbreviation viii

List of Figures x

List of Tables xii

1 Introduction 1

2 Review of Emergency Medicine and Expert System 4

2.1 Emergency Medicine 4

2.2 Emergency Medicine in India 5

2.3 Role of ICT in Health care 6

2.3.1 Telemedicine 7

2.3.2 Role of ICT in Emergency Health Care System 8

2.3.3 Key Elements for accessing the quality of ICT in emergency

healthcare

11

2.3.4 Some of the ICT application in health care in India 15

2.4 Expert system In Medicine 20

2.4.1 Expert System – A brief review 20

2.4.2. Classification of ES 20

2.4.3 Expert system in Medicine 22

2.4.4. Few examples of ES used in India 26

2.5 Summary 27

2.6 Definition of the Problem 27

3 Patient Assessment Tools in Emergency Medicine 29

3.1 Introduction 29

3.2 Early warning scoring system 30

3.2.1 APACHE II 30

3.2.2 APACHE III 31

3.2.3 MEWS 31

3.2.4 NEWS 32

3.2.5 PHEWS 35

3.2.6 Discussion 37

vi

3.3 Primary Assessment tool in Emergency health care system 38

3.3.1 Perfusion Status Assessment 38

3.3.2 Respiratory Status Assessment 39

3.3.3 Conscious State Assessment (Glassgow Coma Scale) 40

3.4 Contribution to this research work: 40

3.5 Summary: 41

4 Tools of Knowledge-Based System 42

4.1 CommonKADS: A Modeling Approach for Knowledge engineering 42

4.1.1 Introduction 42

4.1.2 CommonKADS Modelling Framework 43

4.1.3 Discussion 49

4.2 Introduction to Ontology 50

4.2.1 Ontology 51

4.2.2. Advantages of using ontology 52

4.2.3 Principles for the Design of Ontologies 52

4.2.4 Types of Ontology 54

4.2.5 Ontology Languages 55

4.2.6 Reasoning 58

4.2.7 SPARQL 58

4.2.8 Ontology Editors 59

4.2.9 Steps for creating ontology in Protégé 61

4.2.10 Discussion 62

4.3 Summary 62

5 Architecture and Development of Expert System - “Meditrace 64

5.1 Overall System Architecture 64

5.2 Ontology Development Phase 68

5.2.1 Define class and class hierarchy 68

5.2.2 Define Object and Datatype property 70

5.2.3 Defining facet, range, and domain of the property 72

5.2.4 Creating Individuals 74

5.3 Ontology Model Creation 75

5.3.1 Loading Ontology file 75

5.3.2 SPARQL query to retrieve the information from the ontology 76

vii

5.3.3 MySQL Database server 77

5.3.4 Use case for Developed Expert system 78

5.3.5 Description of Activity diagram 80

5.4 Implementation of Ontology-based Expert system for Emergency Medicine

– Meditrace

81

5.4.1 Emergency Staff registration page: 81

5.4.2 Login page 82

5.4.3 Emergency Risk Level Assessment 82

5.4.4 Display page of Risk level score 83

5.4.5 Differential Diagnosis 83

5.4.6 Primary Assessment Result screen 85

5.4.7 Treatment 85

5.4.8 Emergency Assessment Patient Report 87

5.5 Emergencies included in developed expert system 88

5.5.1 Cardiac Emergency 88

5.5.2 Respiratory Emergency 89

5.5.3 Treatment Guidelines 90

5.6 Summary 91

6 Results & Validation 92

6.1 Testing Dataset 92

6.2 Validation Results 95

7 Conclusion & Future Scope 98

7.1 Conclusion 98

7.2 Future Scope 99

List of References 101

List of Publications 112

viii

List of Abbreviation

APACHE Acute Physiology and Chronic Health Evaluation

API Application programming interface

AVPU Alert/Verbal/Pain/Unresponsive

CDSS Clinical Decision Support System

COPD Chronic Obstructive Pulmonary Disease

DIT Department of Information Technology

ED Emergency Department

EHR Electronic Health Record

EM-DSS Emergency Medicine Decision Support System

EMS Emergency Medicine Service

ES Expert System

EWS Early warning scoring system.

GCS Glasgow Coma Scale

HIS Hospital Information System

HR Heart Rate

ICT Information and communication technology

ICU Intensive Care Unit

JSP Java Server Pages

KBS Knowledge-Based System

KE Knowledge Engineering

MDDS Metadata & Data Standards

MEWS Modified Early warning scoring System

MoHFW Ministry of Health and Family Welfare

MVC Model View Controller

NEWS National Early Warning Scoring System

NTN National Telemedicine Network (NTN):

OWL Ontology Web Language

PH Pre-Hospitalization

PHEWS Pre-Hospital Early Warning Score

RDF Resource Description Format

RDFS RDF Schema

ix

RR Respiration Rate

SBP Systolic Blood Pressure

SPARQL SPARQL Protocol And RDF Query Language

W3C World Wide Web Consortium

WHO World Health Organization

XML eXtensibleMarkup Language

x

List of Figures

FIGURE 4.1 Organizational Model for Emergency Medicine Expert

System

44

FIGURE 4.2 Agent, Task and Communication Model for Emergency

Medicine Expert System

46

FIGURE 4.3 Percentage of ontology languages currently used by a user 56

FIGURE 4.4 Percentage usage of ontology editors by respondents 60

FIGURE 5.1 Architectural framework of the developed system 64

FIGURE 5.2 Simplified architecture of JSP and Java servlet technology 65

FIGURE 5.3 An overview of the classes of the ontology 68

FIGURE 5.4 Part of some concepts in the developed ontology 69

FIGURE 5.5 Object and Datatype property 70

FIGURE 5.6 Facet, domain, and range of properties 71

FIGURE 5.7 Example of instances in the ontology 74

FIGURE 5.8 JAVA code for loading and saving an ontology model 75

FIGURE 5.9 Example of SPARQL query to retrieve risk level of patient 75

FIGURE 5.10 Use case diagram of the developed system 77

FIGURE 5.11 Activity Diagram 78

FIGURE 5.12 Emergency Staff registration page 80

FIGURE 5.13 Login screen 81

FIGURE 5.14 Emergency assessment screen 81

FIGURE 5.15 Risk assessment score page 82

FIGURE 5.16 Differential diagnosis assessment screen 83

FIGURE 5.17 Assessment result screen 84

FIGURE 5.18 Treatment screen step-1 84



FIGURE 5.21 Treatment screen last step 85

FIGURE 5.22 Admin screen for patient assessment report 86

FIGURE 5.23 Screen showing assessment of one patient 87

FIGURE 5.24 Treatment steps for COPD 89

FIGURE 6.1 Patient database with Gender variation 92

xi

FIGURE 6.2 Patient database with Age variation 92

FIGURE 6.3 Patient database with variation in Past history 93

FIGURE 6.4 Patients NEWS score and its range variation 93

FIGURE 6.5 Patient database with cardiac emergency variation 94

FIGURE 6.6 Patient database with respiratory emergency 95

FIGURE 6.7 Disease prediction probability 96

FIGURE 6.8 The success rate for different diseases 96

xii

List of Tables

TABLE 3.1 MEWS scoring system 32

TABLE 3.2 NEWS scoring system 34

TABLE 3.3 NEWS threshold trigger level 35

TABLE 3.4 Parametric Comparison of various EWS 36

TABLE 3.5 Perfusion status assessment 38

TABLE 3.6 Respiratory Status Assessment 39

TABLE 3.7 GCS score 40

TABLE 4.1 NEWS table for score calculation and clinical risk

determination

47

TABLE 4.2 Knowledge elicitation form for diagnosis of EM disease 48

TABLE 4.3 Knowledge elicitation form for possible treatment of EM

disease

48

TABLE 5.1 List of object properties with their domain and range 71

TABLE 5.2 List of Data type properties with their domain and range 72

TABLE 5.3 Data dictionary in database 76

Introduction

1

CHAPTER 1

Introduction

An expert system is a computerized system or software program developed from the

knowledge of experts and information from the application domain. It can also be

considered as an assistive system which helps to solve the problem effectively in the

absence of the experts. An expert system has found its application in different domains

with the growth in computer technologies. The health care sector is one of the most

significant areas, which needs considerable focus specifically in developing countries. An

expert system can play a very important role in the medical sector in several ways such as

patient management, disease diagnosis, laboratory analysis, treatment planning, and

medical education. Most of the developed expert system depends on individual

competency and effectiveness of designing the system. This makes the system inefficient,

rigid, unreliable and unscalable. The semantic web is the novel concept introduced by the

researcher to make the machine interpretable web. Ontologies are the prime component of

the semantic web, which is used to model the availableinformation in semantic form.

Ontologies are based on a shared and consensual domainknowledge agreed by a

community. Ontology is expressed by languages known as OWL(Ontology web

language).Ontology provides a way of expressing data ina generic, extended and integrated

form and makes the overall system flexible and scalable.Emergency health care sector

requires immediate intervention by the medical professional for saving the patient’s life.

Lack of trained and experienced emergency staff is a serious hurdle in making the

emergency health care delivery effective. In addition to this unavailability of expert and

specialist medical staff in a rural region is another obstruction in the pre-hospital health

care system in several regions. The requirement of assistance in the process of disease

diagnosis, risk level detection, and therapeutic guideline is one of the promising areas for

an expert system. The research work presented here provides an ontology-based online

expert system for emergency medicine.

Introduction

2

The thesis includes a total seven chapters. The first chapter is of introduction. Immediately

after this introduction, the next chapter consists of basic understanding and requirement of

emergency medicine concentrating specifically on pre-hospital emergency health care

delivery system. The chapter includes a comprehensive review of the health care system in

India and the application of ICT in the health care sector. It also covers the various efforts

made by the different researcher for improving the quality of emergency health care

delivery system. In order to make this system applicable in the Indian context, it is

necessary to explore various challenges faced by the health care sector in Indian scenario.

Various sociopolitical, organizational, financial, regulatory and technological challenges

are discussed in that chapter in detail with respect to the Indian situation. The basic

information about the expert system and its classification is also discussed in chapter 2.

Expert systems are used in the medical field fora long time. Some of the most popular

medical expert systems are discussed in detail in the later section of that chapter. The

chapter ended by formulating the research problem.

Chapter 3 includes various primary assessment tools used in the emergency department for

assessing the patient condition. The primary risk level stratification is performed by early

warning scoring system. The first section covers basic information about the requirement

of early warning scoring (EWS) system. It includes different EWS used in various

department of in the hospital and out of hospital situation. The section also includes

introduction about different scoring system with a detail discussion of NEWS scoring

system which is used in this research work for risk level stratification purpose. Parametric

comparison of different scoring system is an essential criterion of selection of EWS. The

second section of the chapter includes various primary assessment tools available for

perfusion status, respiratory status, and conscious status assessment.

The management and development of knowledge-based system require various tools. The

fourth chapter includes a well-known modeling approach adapted for knowledge

engineering, called commonKADS. It includes the construction of six main models used to

construct a knowledge-based system. It includes organization, agent, task, communication,

knowledge, and design model. These models are used to fragment the whole complex

problem into a smaller and modular structure which simplifies the process of system

development. The second section includes the overview of the ontology and semantic web.

Introduction

3

Ontology is a hierarchical structure of the most significant concepts related to aparticular

domain, an association between classes and their properties. The knowledge encoding of

ontology requires certain formal languages known as ontology languages. The section also

includes an explanation about a few popular ontology languages available and their

selection process. OWL (ontology web language) is one of the most used standard

ontology language written in XML format. The chapter ends with different types of

ontology editors and their features. It also lists out the steps for creating the ontology in

one of the most popular open source ontology editor called protégé.

The fifth chapter includes the overall architecture and the development of an expert system

for emergency medicine called Meditrace. It is deployed at www.meditrace.in. This system

provides risk level stratification of a patient, differential diagnosis of disease and treatment

guideline to paramedic staff in an emergency situation. The process of ontology

development and ontology model creation is discussed in detail in that chapter. This

system is designed with an aim to suggest the risk level of patient based on the available

physiological parameters. It also seeks additional information from the paramedic for

further assessment of a patient’s health and to facilitate the process of diagnosis. The later

section includes user information and their access to the system, different screens added in

the system and their significance. The chapter ends with the selection lists of emergencies

which are incorporated in the developed system.

Chapter 6 consist of the process of patient data collection and the validation of the system.

It includes the variation of databases selected for the validation purpose. The success rate

of the disease prediction probability is also shown for the whole patient group. The thesis

ends with a summary of the research carried and brief ideas of the possible future scope of

this research work in chapter 7.

Review of Emergency Medicine and Expert System

4

CHAPTER 2


This chapter includes an overview of emergency medicine and an expert system. The first

section includes an introduction to the emergency health care delivery system and its

essential components. After that, the existing situation of emergency care in India will be

discussed. It also includes in detail discussion about key components of accessing the

emergency health care system in India. The next section includes an introduction to the

expert system and its general classification. This section includes different types of expert

systems available listing some of their applications in various domains. Then, expert

systems in the medicinal field called Clinical Decision Support System (CDSS) and few

earlier CDSS systems are discussed in the chronological event of their development. After

that, it includes a few examples of expert systems used in India in various application

domains. The chapter ends with problem definition and the contribution of the thesis.

2.1 Emergency Medicine

A medical emergency is an unexpected wound or medical complaint (physiological or

psychological) requiring immediate medical care. The person might be in danger of any

health impairment or loss of life or maybe incapacitate or vulnerable as a result of a

physical ormental condition. Emergency medical care focuses on giving urgent and timely

medical interventions to stabilize such patients and prevent any possible disability and

death.

Time is one of the most critical factors while the patient is attended in an emergency. If the

patient is suffering from an acute heart attack then immediately the patient stops breathing

and the heart stops pumping then within four to six minutes, irreversible brain damage may

occur. So providing help within this time frame is the most critical process. The study also

Emergency Medicine in India

5

shows, for out-of-hospital cardiac arrests, it was found that the chance of survival is less if

victims were not resuscitated before they reach the hospital. About 50 percent of road

traffic accident death happens in an initial 15-20 minutes of the accident due to severe

injury to the main organs of the body including brain, heart and major blood vessels. In

addition to this around 35%, people have died in the next 1-2 hour of chest and head

injuries. The time between injury and initial stabilization is the key factor for saving the

patient’s life.

The main goal of emergency medical care will be- First, to guarantee timely detection of a

medical emergency, urgent provision of First Aid and efficient resuscitation. Second is to

ensurethe prompt &safe transportation of the patient to the most suitable emergency

medical department of a hospital.And third, the subsequent provision of more definitive

treatment.

Considering this, pre-hospital care is the most important area every emergency medical

personnel have to look for. Pre-hospital care should be simple, sustainable and efficient.

Paramedical personnel plays a very important role in this using their dedicated and

equipped vehicle. The paramedical team should comprise of: EMT basic, EMT Paramedic

and EMT advance, with each one of them having dedicated roles and

responsibilities(Sharma & Brandler, 2014).

2.2 Emergency medicine in India

India is right now amidst a monetary and demographic transition. Since the country is

facing a serious epidemiological transition due to urbanization with changing lifestyle, it

results ina rapid expansion of cardiovascular and cerebrovascular illness, diabetic

problems, Chronic Obstructive Pulmonary Disease (COPD) and so on. Moreover,

communicable diseases (acute respiratory infections, acute diarrhoeal diseases,

tuberculosis, malaria, etc.)keep on increasingconsiderable amount ofburden of disease in

the country. Other than these, some of the unintentional injuries (road traffic accidents,

fires, falls, etc.) and intentional injuries (self-inflicted injuries and those due to violence)

also represent a significant burden of disease in the nation. Many of these conditions

require emergency care in their acute stages or are acute in nature (Myocardial Infarction

(MI), acute hemorrhages)(Joshipura, Hyder, & Rehmani, 2004).


6

The concept of “quality of health care”emerges from our understanding of the goals of

healthcare. The main objectives of any medicinal service includes health improvement of

thepopulation, sensitization towards people’s need and financial protection against ill-

health expenses(WHO World Health Organization (WHO), 2000). India as a developing

nation with large rural areas and populations has some of the crucial issues to deal with,

such as prevent chronic infectious disease, the lack of adequately trained health care

personnel and health care facilities, and a limited number of health care programs.

2.3 Role of ICT in Health care

Information and communication technology (ICT) plays a critical role in unifying

communications, making people to access, process, store and transmit data through fully

integrated audiovisual, data communications, and electronics systems(Henriquez-

Camacho, Losa, Miranda, & Cheyne, 2014).Since 1999, ICT has a crucial role to play

almost in every sector of the society, including the health care sector eHealth. eHealth

technologies offer a reduction in cost and advancement in health information exchange and

improve health care access ultimately enhances the effectiveness of the health care delivery

system. In 2014, more than 90 percent of people in developing countries are an active

subscriber of a cell phone. The widespread use of mobiles and their ease of use give rise to

mobile health mHealth.

The use of ICT in the health care sector has focused mainly on health care following these

three principles ways: Improving the functioning of health care systems, improving the

delivery of health care and improving communication about health.

Role of ICT in improving the delivery of health care utilizes ICT for better diagnosis,

better training and sharing of knowledge amongst workers in primary and rural health care

which includes: biomedical literature search and retrieval, continuing professional,

development of health workers, telemedicine and remote diagnostic support, diagnostic

imaging, critical decision support systems, quality assurance systems and disease

surveillance and epidemiology(Ariani, Koesoema, & Soegijoko, 2017).

Role of ICT in Health Care

7

2.3.1 Telemedicine

Telemedicine, a term that came in existence in the 1970s, means "healing at a distance",

signifies the usage of ICT to improve patient outcomes by increasing access to care and

medical information.

As per World Health Organization (WHO), 1998 "The delivery of health care services,

where distance is a critical factor, by all health care professionals using information and

communication technologies for the exchange of valid information for diagnosis,

treatment, and prevention of disease and injuries, research and evaluation, and for the

continuingeducation of health care providers, all in the interests of advancing the health of

individuals and their communities”(WHO World Health Organization (WHO), 1998).

As per the international telecommunication report (ITU) 1998, Telemedicine is potentially

an efficient means of providing specialized medical services to a remote location. It has

also stated that telemedicine promises to improve the quality of medical care and decreases

cost, particularly in under-served urban and rural areas(Wright, 1998).

As per the WHO report 2010,access, equity, quality, and cost-effectiveness are key issues

facing health care in both developed and less economically developed countries(WHO

World Health Organization (WHO), 2010). Modern information and communication

technologies (ICTs) (including computers, internet, and cell phones) are revolutionizing

how individuals communicate with each other, seek and exchange information, and

enriching their lives. The survey undertaken in 2009 has highlighted the role of ICT in

health care- Telemedicine, especially in developing countries. It was highlighted that

telemedicine is the biggest opportunity for increasing access to health care. It was also

reported that telemedicine can successfully improve the quality and accessibility of

medical care by allowing evaluation, diagnosis, and treatment from a distant service

provider. The secondary benefit of telemedicine was also reported in that survey asserts

telecommunication channels used in telemedicine can be effective tools for connecting

remote sites. This allows communication between health care professionals located at rural

and remote sites acrossthe globe, overcoming geographical barriers. Telemedicine also

provides opportunities for learning and professional development by enabling the provision


8

and dissemination of general information and the remote training of health-care

professionals.

As per report by the Innovation Working Group (IWG) ASIA task force on telemedicine,

The deployment of information and communications technology for improving the reach

and penetration of healthcare services, in the form of telemedicine and mobile health (m-

health) services is a potential solution to mitigate strains faced by healthcare systems

across the world. They have highlighted the challenges faced by various nations in health

care delivery in a remote and rural area. The report has highlighted the health care issues

faced by developing countries with an increasing rate of population. The potential benefits

of telemedicine include: Effective management of chronic disease, care of physically and

mentally challenged patients, patient empowerment, for assistive primary care providers,

health financing, community & population health improvement, addresses a shortage of

healthcare workforce. It was also highlighted in the report that telemedicine can be useful a

lot in emergency health services(The Innovation Working Group (IWG) ASIA Task Force

on Telemedicine, 2014).

2.3.2 Role of ICT in Emergency Health Care System

In 1998, a team ofresearchers from Biomedical Engineering Laboratory, Athens, Greece,

has also explored the role of telemedicine in pre-hospital patient care and management

using wireless technology in Ambulance (Pavlopoulos, Kyriacou, Berler, Dembeyiotis, &

Koutsouris, 1998). They have also highlighted the lack of skills & training in ambulance

personnel (EM - paramedic, EM - Technician, EM - Basic), who manages the emergency

first. They proposed the first ambulance architecture where the mobile unit is located in an

ambulance while the consultation unit is located in a hospital and both of them are

connected wirelessly by the GSM link. The mobile unit consists of a bio-signal monitor

and a portable PC. The system gets data from the bio-signal monitor, stores in the local

hard drive and then transfers this data to the hospital through GSM modem. The system

has shown stability and robustness in real-life emergency conditions. But that system

utilized the technologies available at that time and lacks portability and lack of GIS/GPS.

This system is the stepping stone towards the discovery of ICT application in the

emergency health care system.


9

Karlsten and Sjoqvist(Karlsten&Sjoqvist, 2000)have suggested the usage of telemedicine

and decision support in the emergency ambulance (pre-hospital). Early diagnosis in the

ambulance itself can improve the handling of the patient at the hospital and can save the

patient by initiating proper and timely treatment. In that case, the EM staff can transfer

some of the basic information regarding the patient to the central facility including the

location of the patient, name of the patient, main symptom, priority based on risk level. It

was the very first usage of telemedicine in ambulance suggested by the researcher and they

have developed the system for implementing this concept. This system has facilitated the

process of quality control and follow-up. This is also the first system to incorporate the use

of a Decision Support System (DSS) in pre-hospital emergency care.

The research proposed by Pavlopoulos and team in the year 1998, taken further by again

the team of researcher from biomedical engineering laboratory in the year 2003 under the

supervision of Kyriacou(Kyriacou et al., 2003). They have developed multi-purpose

healthcare telemedicine systems by establishing a communication link from a mobile

network. The system was focused on developing combined real-time and store and forward

facilities using the base unit and telemedicine unit. This integration is very much useful in

handling an emergency in ambulances or at rural health centers. This system allows the

transmission of vital bio-signals (ECG, SpO2, HR, NIBP, Temp, Respiration rate) and still

images of the patient. The transmission of data is possible through GSM mobile

telecommunication network or a satellite link. The consultation site is equipped with

multimedia to view the patient and database to store and manage the collected data. The

system was tested in real time emergency health care situations. The system has shown

improvement in the percentage of incidents in the emergency case and employed in some

of the ambulances in Greece and Sweden. This system includes the scope of transmitting

live data as well, but still, the system faces technological constraints in terms of feasibility

as it transfers waveform and images.

Health information and quality authority report(2014), has also listed some of the prime

usages of ICT in the national ambulance service. The report includes computer-aided

dispatch – incident tracking system, emergency response resource location, incident

address verification, satellite navigation systems used by emergency response personnel,

communication between control centers and emergency response staff, mobile data

terminals and patient care report(Health Information and Quality Authority, 2014).


10

Ahmed, Ishaque, and Nawaz (2014) have investigated the impact of ICT on increasing

efficiency in emergency medical services. ICT is becoming an essential part of health care

life-critical systems and has substantially reduced morbidity and mortality rates. He also

mentions that it is possible to transmit the patient vital sign values over mobile networks as

a part of pre-hospital care. He suggested a framework for emergency medical services in

Pakistan. The suggested framework includes a central dispatch center, ambulatory vehicle,

and care facility. The data communication takes place over cellular networks already

established in the country. Still, the country is not utilizing the full potential of ICT in

emergency care(Ahmed, Ishaque, & Nawaz, 2014).

In 2014, the team of researchers under the guidance of Zahhad has implemented a wireless

emergency telemedicine system for patient monitoring and diagnosis(Abo-Zahhad,

Ahmed, & Elnahas, 2014). The system consists of a mobile care unit connected to the

patient body, data communication networks preferably GSM/GPRS (for real-time) or

Internet (store and forward), Remote server with central database and local monitoring

facility and management/monitoring units consisting emergency service, medical

personnel or physician. The mobile care unit acquires the ECG, SpO2, Temperature, and

BP sensor to get the bio-signals from the patient's body. This system has presented the

user-friendly web-based interface for medical staff to observe current vital signs for remote

treatment. This system primarily designed for monitoring purpose which lacks intelligence

and automation in terms of diagnosis.

Two of the senior researchers associated with planning and development of healthcare

solutions in Fujitsu Kyusu System have developed the project on information support

solution in Emergency Medical Service proposing close collaboration between paramedics

and medical institution(Sonoda & Ishibaei, 2015). Sonada and Ishibai (2015) have

considered hospital reference information to be shared among paramedics through the fire

fighting command system, the use of a tablet to share information between paramedics and

medical institutions, Connecting Health Information Exchange (HIE) network, and

emergency transport support system. This system was able to support paramedics to obtain

various types of information from the established network. This makes the transportation

times shortest and allows utilizing information about the patient including medical history,


11

allergy, and other data. This Fujitsu funded project has opened another dimension to assist

paramedics in an emergency condition.

Badr(Badr, 2016)has examined the role of ICT in pre-hospital emergency medical services.

Badr has also stressed on requirement of a specialized and unique type of intelligent

transport system in emergency medical services. It was also suggested that ICT usages as

telemedicine had a positive impact on emergency pre-hospital medical care.

The smart ambulance system is another attempt to identify the role of ICT in pre-hospital

emergency care(Gupta, Pol, Rahatekar, & Patil, 2016). The system was implemented into

client-server architecture to make it a small size application and keep the data available at a

central location. The system was the attempt to utilize Internet - of - things (IoT) in the

emergency health care system. The system helps to identify appropriate hospital and used

to transmit continuous real-time data of patient's health to hospital personnel. This reduces

the time complexity and helps to provide faster health care service.

Koceska and the team from Macedonia in the year 2019, proposed the system of a mobile

wireless monitoring system for pre-hospital emergency care(Koceska et al., 2019). This

system has utilized wireless bio-sensors for monitoring the vital parameters of a patient

and this data will be transferred and monitored by the paramedic available in ambulance.

With the available internet connection, this data can also be transferred to a central location

for further investigation or necessary guidance. This system displays real-time vital data

measurement, historical trends of these parameters, Glasgow coma scale, and place of

injury and also incorporates triage procedure. Primarily, this system can be used as a

complementary system in EMS, allowing continuous real-time monitoring of patients vital

sign wirelessly and on-scene triage. This system does the designated task efficiently but it

lacks intelligence and assistance for taking a proper decision.

2.3.3 Key Elements for accessing the quality of ICT in emergency healthcare

For successful development and applications of ICT in healthcare, especially in developing

countries, there are few existing and potential challenges. The key elements for accessing


12

the quality of ICT in emergency health care in India are listed below and each of them is

discussed in detail:

Access to Health care

Financing of Emergency care

Policy & Standards

Human resource for emergency care

Research & Evaluation

Coordination & Collaboration

2.3.3.1 Access to Health care

The quality of emergency care indicates the overall performance of the health sector.The

health care system is specifically suffering from inequality in terms of location and

socioeconomic status. There is a vast gap in the numbers of hospitals, dispensaries, PHCs

available in a rural area and those in an urban area considering the average population. As

per the Government of India (GoI) health policy (Ministry of Health and Family Welfare

Government of India, 2017), the Government tried to attract and retain doctors in rural

areas by giving them various financial and nonfinancial incentives.These steps and other

policies and strategies proposed ultimately used to improve the quality of health care. The

government has framed some strategies for providing effective and emergency medical

care to the accident-prone zone. In addition to this, some major players in private health

care sectors are also taking an initiative to improve pre-hospital emergency care by

collaborating with either government or any public sector player. This includes Mobile

Telemedicine Bus for rural health care delivery by AIIMS cochin and PGIMER

Chandigarh, Sky health centers by world health partners & Melinda gates foundation,

Lifeline express train by impact India foundation and Indian railway, ATM-based kiosk for

rural healthcare by Yolo health ATM and KIOSK, Tele eye care service in CHC in PPP

with Apollo hospital services(S K Mishra, 2018). Air ambulance care is also introduced in

India but in a limited region but this service is offered only by the private sector. Some of

them are Panchmukhi Air Ambulance, VMEDO ambulance, Skylift aviation, etc. These

initiatives are primarily covering the population in an urban area, while the rural

population of this country is still lacking the primary treatment as a part of emergency

health care.


13

2.3.3.2 Financing of Emergency Care

As per GOI policy (2017), the main aim is to ensure adequate investment by increasing

healthcare expenditure by the government as a percentage of GDP form the existing 1.15%

to 2.5% by 2025. Emergency care in India is still not receiving much priority while

allocating the public fund. In addition to this small percentage of money is allocated out of

the total health care budget. Looking at the current situation in developing countries, one

would say that the major sharing of the financial burden because of poor health is due to

diseases and other conditions related to emergency care. Therefore, families are forced to

choose between indigence and financial obligation of medical expenses or death risk or

impairment due to the unavailability of emergency care. Extensive and Dedicated funds are

required from either the government or from the private sector for the overall improvement

of quality of emergency health care in India. In GoI policy (2017), the Government has

started to provide free emergency care services in all public hospitals(Razzak &

Kellermann, 2002).

2.3.3.3 Policy and Standards

At the policy level, the major challenge is to make telemedicine an integral part of the

healthcare delivery system in India. There are few initiatives taken by the Department of

Information Technology (DIT) and the Ministry of Communication and IT (MCIT) for the

preparation of guidelines and standardization of telemedicine infrastructure in

India(Mishra, Singh, & Chand, 2012).As per GOI-Policy (2017), the policy has proposed

collaboration with the private sector for comprehensive primary health care by focusing on

one of the telemedicine services. Still, India needs to frame proper guidelines and

standards of many medical review criteria for many types of health care. Standards and

regulatory frameworks are hardly available for health care quality assurance(Bhat, 1996).

The recent efforts of the Ministry of Health and Family Welfare (MoHFW) have taken an

effort to develop central drugs standards control organization (SUGAM), Electronic Health

Record (EHR) standards, and Metadata & Data Standards (MDDS)(“e-Health &

Telemedicine | Ministry of Health and Family Welfare | GOI,” n.d.).


14

2.3.3.4 Human resources for Emergency care

The unavailability of trained staff at all tiers is the major limitation of the Indian health

care delivery system. As a result of asymmetricaldevelopment in the health care system,

the trained professionals are pondered in the urban region while the rural sector faces a

tremendous shortage. Hence, to involve them in healthcare delivery through telemedicine

requires specific approaches to make the teleconsultation effective as well as efficient.

Lack of paramedical and medical staff at all levels is a serious constraint in the provision

of proper and efficient emergency care, specifically in a rural region. Because of this GoI

policy (2017), it recommends a scheme to develop human resource and specialist skills.

The policy has also targeted the human resource/skill gap. The policy states "workforce

performance of the system would be best when we have the most appropriate person, in

terms of both skills and motivation, for the right job in the right place, working within the

right professional and incentive environment”. The policy has also recommended ensuring

the availability of paramedics as per standard guidelines. The availability of suitable health

human resource at all tiers is the key to the development of the emergency care system in

the nation(Sharma & Brandler, 2014).

2.3.3.5 Research and Evaluation

Research is the key component indicates the progressive development in the emergency

health care delivery system. As compared to developed countries, India is not paying much

attention to this factor. GoI policy (2017) has recognizedthe need for building research and

public health skills for preventive and primitive care. The policy has also highlighted the

need for research in developing a new vaccine, e-health, telemedicine, medical

education,drug discovery, diagnosis of disease and treatment sector. The policy has

focused on “Health research plays a significant role in the development of a nation’s

health.”Research should focus on utilizing available technology for defining proper

diagnosis and proper treatment with a limited source of information and

manpower(Hofman, Primack, Keusch, & Hrynkow, 2005).


15

2.3.3.6 Coordination and Collaboration

To deliver urgent and effective emergency care service it requires proper coordination at

different tiers of health care providers, multiple government and non-government agencies.

It was suggested by the experts in the ICT in the emergency medicine field to make a

central coordinating agency for monitoring, controlling and facilitating Emergency

Medicine Services (EMS) under the ministry of health and family welfare. The integration

of emergency care with other health care system improves the overall health care scenario

at country level.As per GoI policy (2017), the policy recommended inter-sectoral

coordination at the national and sub-national levels to optimize health outcomes. The

policy also suggested exploring the collaboration of primary care services with a "not-for-

profit" organization having a track record of public services. The policy supports

collaboration with private sectors in capacity building, skill development program, Disaster

management, Enhancing accessibility in private care, disease surveillance, Make in India

and Health Information System (HIS)(Joshipura et al., 2004).

2.3.3.7 Technological Resource

The technological resource is the most important resource required for effective

implementation of telemedicine.With respect to technological acceptance, India is far

behind its Asia Pacific counterparts such as Australia, Japan, South Korea, Singapore, and

Malaysia. In 1998, An autonomous government organization, Center for Development of

Advanced Computing (C-DAC), developed and deployed the very first Hospital

Information System (HIS) software by collaborating with Sanjay Gandhi Post Graduate

Institute of Medical Sciences (SGPGIMS), Lucknow. With the development of

communication and active collaboration with the private sector, some of the basic

technological constraints can be overcome to establish a cost-effective telemedicine

system(Chandwani & Dwivedi, 2015).

2.3.4 Some of the ICT application in health care in India:

The integration of ICT into the existing health system has dramatically improved health

care delivery in different ways(Mishra, Kapoor, & Singh, 2009). Few of the attempts made


16

by different sectors in the field of ICT based health care (e-Health) in India, are discussed

here(“e-Health & Telemedicine | Ministry of Health and Family Welfare | GOI,” n.d.):

Ministry of Health and Family Welfare (MoHFW), Government of India:

GoIhas taken various initiatives by utilizing ICT for improving efficiency and reach of the

health care delivery system. Some of the important initiatives are:

Central Drugs Standard Control Organization “SUGAM”:A “single window”

system developed to serve multiple stakeholders (Pharma Industry, Regulators, and

Citizens) involved in the process of the Central Drugs Standards Control

Organization. "SUGAM" allows online application submission, tracking,

processing and grant of approvalsonline mainly for drugs, clinical trials, ethics

committee, medical devices, vaccines, and cosmetics.

Vaccine Tracker (mobile application) (Indradhanush Immunization):The

application designed specifically to present in detail information about various

types of vaccines available in India and their schedule. The “Indradhanush”

application is developed to increase the parent’s awareness of their children’s

vaccination.

NHP (National Health Portal)Swasth Bharat (mobile application):This

application was developed with an aim of providing in detail information with

respect to a healthy lifestyle, disease conditions (A-Z), symptoms, first aids,

treatment options,and public health alerts.

Pradhan MantriSurakshitMatritvaAbhiyan (PMSMA) (mobile application):It

offers a common platform to the retired obstetricians, radiologists and physicians to

engage themselves voluntarily in free antenatal services to the pregnant woman at

government health facilities.

mDiabetes Program: This program contributes to increasing awareness amongst

people about diabetes and various suggestions to prevent diabetes by living a

healthy and active lifestyle. mDiabetes will also contribute to the process of early

diagnosis, give better adherence to drug or dietary control, self-care, as well as

helps to prevent complications among patients with diabetes. mDiabetes is based on

proven algorithms for diabetes prevention and care and builds on previous


17

international experiences in using mobile technologies to deliver these

interventions.

Hospital Information System (HIS): HIS is one of the most comprehensive

software technologies desperately needed in today’s era. HIS is to be implemented

in hospitals to automate various hospital processes to attain better efficiency and

service delivery in Public Health facilities even up to the Community Health Center

(CHC) level. This system is designed with the aim of managing efficient hospital

workflow, ultimately to improve health care delivery services to the patient and to

enhance the efficiency of different hospital processes.

(Electronic Health Record) EHR Standards:A uniform system for maintaining

the patient record in electronic form is the key consideration in this particular

scheme. The standards are needed to maintain Electronic Medical Records /

Electronic Health Records (EMR / EHR) by the hospitals and healthcare providers

in the nation. A team of experts was set up by the government to propose and

develop EMR / EHR Standards. After due consideration of the recommendation of

the committee and the comments, the “Electronic Health Record Standards for

India” have been finalized and approvedby the Ministry of Health and Family

Welfare, Government of India.

Metadata & Data Standards (MDDS): The MDDS is developed with an intention

to bring semantic interoperability among all health IT systems. This is a

prerequisite for establishing interoperability among disparate health information

systems. Health Domain MDDS has created more than 1000 data elements and 142

code directories. Most of these standards are drawn from global standards however

these are developed keeping in view local health information systems requirements.

National Telemedicine Network (NTN): It is also termed as “National

Telemedicine Portal”. Ministry has employed a project on e-health including

telemedicine on national medical college network (NMCN) for connecting medical

colleges across the nation with an aim of promoting e-learning and National Rural

Telemedicine Network. For this purpose, they have identified the National

Resource Center (NRC) and Rural Resource Center( RRC)

Establishment of Satellite Communication (SATCOM) based Telemedicine

Nodes :In 2001, Indian Space Research Organization(ISRO) through Department

of Space (DoS), had initiated a nationwide Telemedicine (TM) program and


18

deployed TM systems hardware, software, communication equipment as well as

satellite bandwidth for 384 Hospitals with 60 specialty hospitals connected to 306

remote/rural/district/medical college hospitals. Eighteen (18) Mobile Telemedicine

units were also enabled for Satellite connectivity.

Ministry of Electronics & Information Technology: MoE&IThas launched several

health services to assist the health care system:

Telehealth Consultation:Tele-medicine Remote Diagnostic Kit: It is an integrated

wireless healthcare monitoring medical device that helps in monitoring Blood

Pressure, Heart Rate, Blood Oxygen, body temperature, Total Cholesterol,

Haemoglobin, and Blood Glucose. To provide grass root level access points for

health consultation among communities through the digital medium. In 2016, the

Common Service Center (CSC)Special Purpose Vehicle (SPV) launched has its'

own Telehealth consultation services throughout India through Allopathic,

Homeopathic and Ayurvedic doctors across the country.

Initiatives are taken by the large hospital:

Sanjay Gandhi Postgraduate Institute of Medical Sciences (SGPGIMS), Lucknow

has established a School of Telemedicine & Biomedical Informaticsfrom the

financial grant received by the Government of Uttar Pradesh and the Department of

Information Technology, Ministry of Communication & IT, Government of India

with an intention to create various levels of human resources in Healthcare

Information Technology (Telemedicine, Hospital Information Management

System, Nursing Informatics, Digital Medical Library, Medical Multimedia &

Animation, Bioinformatics, Medical Imaging Informatics, Cancer Informatics,

Artificial Intelligence & Clinical Decision support system in Medicine, e-learning

in Medicine, Surgical Informatics, Virtual Reality & Medical Simulation etc.). This

is the firsteducational institution of its kind devotedentirely to the promisingfield of

health care informatics in a public funded academic health institutional

setup(“nmcn - National Telemedicine Portal,” n.d.).

The state of Uttar Pradesh with the support from the Banaras Hindu University

Varanasi has taken an active e-learning initiative by setting up telemedicinelinkages

with some of the selected district hospitals through telemedicine network in order

to get the access to the super-specialty tertiary health care facility.


19

The Telemedicine Centre at Postgraduate Institute of Medical Education &

Research, Chandigarh has initiated basic telemedicine facilities and to deliver

highly specialized quality service to the people of this area covering majority parts

of this region i.e. Chandigarh, Haryana, Himachal Pradesh, Jammu & Kashmir, and

parts of Uttar Pradesh, Uttaranchal, and Rajasthan. This Telemedicine Centre has

an access to 24 district hospitals and 3 medical colleges of Punjab for Tele

consultations and to the Post Graduate Institute of Rohtak, SGPGI Lucknow,

AIIMS Delhi, IGMC Shimla, RPG Tanda, Medical College Jammu and many

others for interactive sessions through Video Conferencing(“nmcn - National

Telemedicine Portal,” n.d.).

The AIIMS, New Delhi, the Department of Telemedicine Facility is dedicated to

providing all aspects of Telemedicine services to the physicians and other health

care delivery staff. AIIMS has an in-house Telemedicine link with (National Drug

Dependence Treatment Centre (NDDTC) - Ghaziabad, Uttar Pradesh &

Comprehensive Rural Health Services Project at Ballabhgarh in Haryana. AIIMS

is one of the premier institutes of the country offers additional telemedicine

services like Tele-CMEs, Tele-Consultation (Online & Offline), Telephonic

Consultation, Tele-Conferences, Tele-Live surgery, Tele-Evidence, etc. to various

medical colleges and hospitals all over India and these services are also extended to

other countries as well(Pinki_CF, n.d.).

One of the very successful frameworks in the private telemedicine sector is that of

the Apollo Telemedicine Networking Foundation. In 1999, the Apollo group has

established a non-profit organization known as Apollo Telemedicine Networking

Foundation (ATNF). The primary duty of this foundation is to offer remote

consultation to patients, for whom due to distance and spiraling costs, access to

quality health care is difficult.The Telemedicine Specialty Centers of Apollo

hospital at various places across the country includes Chennai, Hyderabad, Delhi,

Ahmedabad, Kolkatta, Bangalore, and Madurai acts as telemedicine specialty

(referral) centers(Krishnan, Aditi, & others, 2009).


20

2.4 Expert system In Medicine

2.4.1 Expert System – A brief review

An expert system is the computerized system or software program captures the knowledge

of human experts and uses this knowledge to solve the problem as the expert does. An

expert system can be considered as an assistive system that helps to solve the problem in

the absence of the experts. These computer-based systems found their existence in various

fields depending on the nature and extent of the problem. An expert system is the rule-

based Artificial Intelligence application designed to solve the problem in the intended

domain. ES provides a powerful and flexible means for obtaining solutions to various

problems that can't be solved by other conventional methods.

2.4.2 Classification of ES

Rule-based ES contains the information gathered from a human expert and represents the

information in the form of rules such as IF-THEN. These rules are later applied to the data

to infer the proper conclusion. These inferences are in the form of a computer program

which helps in the reasoning process of information in the rule base and knowledge base,

ultimately forming the conclusion. Rule-based expert systems are used in production

planning, Agriculture planning, tutoring system, Communication system fault diagnosis,

load scheduling, DNA histogram interpretation, etc.(Marchevsky, Truong, & Tolmachoff,

1997).

Knowledge-based ES is a human-centered. It derived from the AI and it attempts to

understand and initiates human knowledge in computer systems. The four principal

components of KBS includesa knowledge base, an inference engine, a knowledge

engineering tool, and a dedicated user interface. A knowledge base contains the knowledge

necessary for understanding and formulating the problem and ultimately to solve that. The

inference engine is the brain of the entire computer system. Knowledge engineering

includes the knowledge acquisition phase which accumulates and transfers the problem-

solving expertise into a computer program for building the knowledge base. The user

interface is a means of communication with the user. KBS also finds varieties of

application: Medical treatment, Personal finance planning, climate forecasting, crop

production planning, power electronics design, chemistry modeling, etc. (Tripathi, 2011).

Expert System in Meidicine

21

Neural network based ES includes Artificial Neural Network (ANN) is a model

whichimitates a biological neural network. This model is used to realize software

simulations for the enormously parallel procedures that include handling components

interconnected in system design. As in biological system neuron receives the inputs from

the other neuron's dendrites in an electrochemical way, an artificial neuron will also

receive the input. The output of artificial neurons resembles the signals transferred from

the axons of a biological neuron. These counterfeit signals can be changed additionally to

the physical changes happening at neural synapses. Applications that are implemented

using the artificial neural network are: Fault diagnosing, decision making, machine

learning, waste treatment, biomedical application, etc. (Hudli, Palakal, & Zoran, 1991).

Fuzzy based ES is developed based on fuzzy logic which deals with uncertainty. This

procedure, which utilizes the scientific hypothesis of fuzzy sets, reproduces the procedure

of ordinary human thinking by permitting the computer to act less precisely and

legitimately than conventional computers. This approach focuses on the result considering

it’s not always true or false. The result may be somewhere in between these two

extremities. Some applications based on fuzzy based ES are power load forecasting, online

scheduling, radiography classification, performance indexing, medical diagnosis, etc. (M.

Patel, Virparia, & Patel, 2012).

Case-based ES is also called case-based reasoning (CBR). It takes on the solutions which

are used earlier to solve the previous problem and utilize them to solve the newer one. In

this approach, depictions of past experiences of human experts, known as cases, are put

away in a database for later recovery when the user experiences another case with

comparable parameters. This framework looks for stored cases with problem attributes

similar to new ones, finds the closest fit, and applies the solution of an old case to a new

case. The new successful solution will also store in the knowledge base along with the old

case. Unsuccessful solutions are also stored in the knowledge base along with the reason

whythey failed. Some of the applications of CBR as manufacturing process design,

medical planning, e-learning, knowledge modeling, etc.(Kolodner, 1997).

Ontology-based ES utilizes the concept of ontology. Ontology is a system of vocabulary.

It is used as a fundamental thought for describing the task/domain data to be known. This

vocabulary is employed as a means of communicating between domain specialists and data

engineers. Consequently, a reusable task/domain model can be characterized and computer


22

code is generated, where ontology can perform data acquisition, reuse, and heuristic

learning. Some common application of ontology includes medical decision support,

preventive control, landscape assessment, chess heuristic planning, etc.(Matkar&Parab,

2011).

2.4.3 Expert system in Medicine

The expert system in the medical field is primarily used for diagnosis, monitoring, tutoring,

and therapeutic purpose. Diagnostic expert systems help doctors to determine possible

diseases. Quick decision making in the diagnosis and selection of proper treatment in a

short time is the most prominent feature of experts systems used in medicine.India, as a

developing country with the second largest population on earth and randomly distributed

population across the country, efficient health care delivery is the most critical issue. In

addition to this, the unavailability of trained professional and field experts, this issue

becomes more critical. Expert systems can play a vital role in solving this problem to a

certain extent. ES can work as an assistive system for untrained medical staff especially in

emergency medicine, which required immediate and effective treatment to the patient. The

expert systems in the medical field are also called Clinical Decision Support Systems

(CDSS)(Miller, McNeil, Challinor, Masarie Jr, & Myers, 1986; Saba, Al-Zahrani, &

Rehman, 2012).

Caduceus (aka The Internist)(Miller, First &Soffer 1986): This system was developed in

the 1970s for CDSSwith an aim of utilizing the reasoning model based on artificial

intelligence, with the main aim of using a “hypothetico-deductive” method for medical

diagnosis. It uses a probabilistic method for ranking diagnosis is the main highlight of this

proposed system. Based on the patient symptoms, the system search through its database

for the most relevant disease using statistics of existing patient data along with predefined

symptoms. But the accuracy of caduceus was not good. As per the investigation performed

and published the report in the year 1981, the Internist failed to cope with the expectations

of experts in terms of accuracy real-world experts. This is due to its limitation in terms of

the available knowledge base and a minimum set of algorithms designed for facilitating the

diagnostic process. This puts limitation over the acceptability of the system in the medical

field.The 1980s has begun the new era of Caduceus by converting into an advanced form

of QMR (Quick Medical Reference). QMR has offered more flexibility with that of

Caduceus. Caduceus was mainly developed to facilitate the process of diagnosis and


23

assistin the consultation procedure.The system has a feature of allowing modification in

probable diagnostic suggestion with the facility of allowing access to its extensive

knowledge base to prove their hypothesis with respect to the therapy of certain intricate

cases.The knowledge base of the QMR needs to be updated as and when new disease was

found. As per the study conducted in 1994 which explored the comparison of three DSS, it

has given fewer correct results than that of the other three systems by a group of

physicians. Because of these reasons,QMR was thrownaway from the market with the aim

of promoting some of the precise and less complex CDSS.

MYCIN(Shortliffe, 2012): MYCIN was developed by Edward Shortlife, from Stanford

University, known as a pioneer of the use of artificial intelligence in medicine. He was the

principaldeveloper of the very first clinical expert system known as MYCIN. In the 1970s,

this system was developed with the aim of developing a system suitable for identification

of infectious diseases and advising appropriatetreatment with some predefined set of

antibiotics. Artificial intelligence (AI) was one of the unique aspects of MYCIN. This AI-

based system was constructed usingmore than 200 predefined rules in order to reach its

knowledge-base and to make the whole rule-base system.The process of identifying the

possible diagnosis accesses the internal decision tree of MYCIN and explores its different

branches to reach the most probable diagnostic option.This system is flexible and

adaptable in very own sense by allowing the clinicians to insert new rules or modifying the

existing ones as per the changing medical demands. Because of this characteristic, MYCIN

was considered as an expert system.Unfortunately, MYCIN was suffering from several

limitations. The first limitation is its sluggish performance as it takes more than 30

minutesfor analysis. Secondly, it was a great concern that whether the clinicians are ready

to risk of relying on computerized results at the cost of their experiencedskills and opinion.

The third thing to consider is the accountability of the system. There was a serious issue of

liability of this system in case of erroneous diagnosis. MYCIN was developed much more

before the era of desktop computing and the existence of the internet, so the system was

based on a rather dated model for computer interaction (Berner, 2007). However, the effect

of this system was far-reaching and can be felt this day, with many existing systems either

merging it with other expert systems (Shyster-MYCIN (O’callaghan, Popple, & McCreath,

2003)) or using it as an influence on the development of new systems (GUIDON(Crowley

& Medvedeva, 2006)).


24

DXplain(Barnett, Cimino, Hupp, & Hoffer, 1987): In 1980s, This is an attempt of devising

a web-based diagnosis system was made by the American Medical Association known as

DZplain. One of the unique characteristics of this system is its simplicity: Clinicians feeds

the patient information based on their own medical terminology and system displays the

list of probable diagnosis from its extensive database of millions of diseases. The system

also offers the choice of potential relevance from the suggested list of diseases.This makes

it possible to use this system for a clinician with less computer acquaintance. DXplain has

proven its reliability in the academic environment with a special feature of cost-efficiency.

In 2010, a group of general medicine residents from Massachusetts has conducted a study

of over 500 various cases of diagnosis. It was concluded from this study that the various

medical charges and service expenses are drastically reduced while using DXplain for

recommendation purpose (Elkin et al., 2010).In addition to this, DXplain has also offered

reasonable diagnostic accuracy for most of the cases. As per the comparative analysis of

four CDSS performed by Lehigh University in2012, it was found that this system was

second in the list in terms of accuracy.

Iliad(Lincoln et al., 1991): In the early 90s, the team of researchers from the University of

Utah, department of medical informatics, under the guidance of Lincoln has developed

another “expert” CDSS known as Iliad. It can be operated in three modes: Consultation

mode, Simulation mode, and Simulation-Test mode. In Consultation mode, clinician

requires to feed actual patient reading in this framework. After that Iliad investigates this

information, based on its matching probabilityit gathers a list of probable diseases. One of

the important characteristicsof the Iliad is its ability to handle "gaps" in patient

information. It means it is capable of suggesting completion methods for inadequate

patient information and ultimately to assist the clinician to work on a possible diagnosis. In

another mode of simulation,the system takes on the role of a complaining patient. It

presents most of the usualcomplaint of the real-life patient and then requests other

information (input, testing, etc.) from the physician. The decisions and responses of the

clinicians will be assessed by Iliad, with the proper response provided once an

investigation is finished. Lastly, in Simulation-Test mode, Iliad presents simulation

identical to the real-life patient, except response is not presented to the physician. Iliad

sends the executed assessment results to another client. Because of its instructive feature,

Iliad is much more effective in training aspiring medical professionals for acquiring real-

life practice.


25

VisualDx(Tleyjeh, Nada, &Baddour, 2006): The attempt of accessing the applicability of

open-source computer technology based on JAVA platform was explored by the team of

researcher from Mayo clinic development, Minnesota. Tleyeh et al. have developed

VisualDx, JAVA-based clinical decision support system. This system is generally used as a

visual aid for assisting healthcare personnelin the process of diagnosis. This is specifically

of importance for surface-level diseases where doctorrequires some visual representations

of these diseases to support the diagnosis. VisualDx is organized by symptoms and other

visual clues instead of being organized by a specific disease. The matching process of

comparing the patient’s image with a pre-existing image from an extensive database is the

uniqueness offered by VisualDX. Based on the outcomes of these comparisons, the system

also suggests relevant therapy. In addition to a huge database of the image, the system also

includes written short outlines about each image.

Isabel(Graber & Mathew, 2008): It is considered to be one of the most comprehensive

CDSS in today’s era. This is also a web-based system developed from the physician

perspective like DXplain. Originally it was designed for pediatric usage but later on its

usage extended for adults as well. Isabel has incorporated two subsystems utility:

Diagnostic checklist and knowledge mobilizing. The first tool of diagnosis checklist seeks

basic and clinical information of the patient and presents of suggested diseases to the

clinicians. In order to get additional information for the suggested diseases, the second

utility tool of knowledge mobilizing can be used. Isabel has performed exceptionally well

in terms of accurate diagnosis in major cases.The study conducted by the Lehigh

University, it was proved to be the most accurate systems amongst the other tested five

systems. In 2003,the School of Medicine of Imperial college has also performed a study

and also confirmed the accuracy of this system(Ramnarayan et al., 2003). As Isabel is a

relatively new CDSS, more extensive validation and testing have to be performed in order

to assure its reliability.


26

2.4.4 Few examples of ES used in India

Maize AgriDaksh developed by the Indian Agricultural Statistical Research

Institute (IASRI), New Delhi, India with the help of AgriDaksh tool(“Agri Daksh,”

n.d.).

It is an agricultural expert system for maize crop, rice crop, jute crop, and

mushroom.

AgriDaksh, a utility tool developed to facilitate the process to build an online

expert system. The important feature of this tool is its simplicity which allows

domain experts to develop their own crop-based online expert system with their

limited set of computer skills and minimal intervention of knowledge system

developer or programmer. This makes it possible to develop an online expert

system for any of the crops in less time with limited resources. It also has the

ability to transfer location specific technology, so makes it suitable to give efficient

and effective suggestion to the farmer. This reduces the losses because of diseases

and pests infestation. Ultimately, this helps to optimize theproduction of the crop

and supports the farmers to earn higher income. Maize AgriDaksh is the first expert

system developed using AgriDaksh tool.

Indian Institute of Horticultural Research Institute, Bangalore: The team of

researcher from IIHRT has developed an expert system which assists the grape

cultivators. Later on, this system was extended for mushroom cultivators, which

performed remarkably well, and became popular. The Institute has alsoextended the

effortsto develop a comprehensive package ofmore than 148 horticulture crops for

cultivation in the four southern states including Kerala, Tamilnadu, Karnataka, and

Andhra Pradesh.

AGREX:A center for Informatics Research and Advancement, Kerala has

developed an Expert System called AGREX. This system is used to facilitate the

people related to the agriculture field and also to give appropriate advice to the

farmers. This was developed with an aim of serving in the field of fertilizer

application, irrigation scheduling, crop protection, and diagnosis of diseases in

paddy and post-harvest technology of fruits and vegetables(Amanpreet Kaur,

2017).

Farm Advisory System: An initiative taken by the Punjab Agricultural University,

Ludhiana, to develop a system to support agri-business management is known as


27

farm advisory system. Based on the responses gathered from user for the pre-

defined set of questions, the system gives recommendations on the farm

management. The inputs are encouraging and it was getting wide acceptance in the

farmer’s community.

TDP Technologies Pvt. Ltd. In Chennai: They were using the MYCIN system for

diagnosing blood disorders.

Tata Memorial Hospital in Mumbai: The utilized PUFF system for the diagnosis

of respiratory conditions.

2.5 Summary

Emergency medicine is considered as the first line of defense for saving the patient's life in

any unintended situation arises from acute diseased condition or any accidental condition.

India, as a developing country with the geographically scattered second largest population,

faces a great challenge while providing quality emergency health care delivery system.

Unavailability of trained paramedic staff in a rural region is the key factor limits the

overall efficacy of emergency care.

Expert systems are in existence for a long time in various applications. With the

development of technology, the Expert system has increased its reach almost in every

possible domain of application. To make the system more effective and automotive, the

Medical field has also identified its suitable role at various levels. MYCIN was first the

expert system developed in the medical field. Later on, a group of researchers with active

collaboration amongst them developed a bigger medical expert system for facilitating the

process of diagnosis and treatment. In the last few years, India has also found the

application of ES in various fields i.e. agriculture, medical, horticulture, etc.

2.6 Definition of the Problem

Providing timely primary care to the patient in an emergency is the most important aspect

of saving the patient's life. Particularly in a developing country, i.e. India with this

geographically widely spread population;it demands equal, effective and quality

emergency health care systems in an urban and rural area as well. But due to the lack of

sufficient medical experts and trained para-medical staff in the rural sector, there is a


28

requirement of assistance from the computer-based expert system. There isadefinite

requirement of the generalized and upgradable expert system in the emergency health care

sector to facilitate the EM-paramedic staff. This system needs to be generic and upgradable

by utilizing the modern state of art open-source technology.

The main goal of the present work is to develop the expert system and the

framework/architecture of the overall system to assist the Emergency Medicine

Practitioner to perform the risk level stratification of a patient. In an emergency health care

service, it is very important to take a timely decision and to initiate therapy as soon as

possible to reduce the further deterioration of a patient's health. The secondary objective of

the developed system is to suggest the list of probable disease and to provide the guidance

of treatment procedure as per the suggested disease. The main objectives are:

To perform the risk level stratification of a patient based on physiological

parameters frequently monitored as a part of ambulatory care as a part of the

emergency care procedure.

To suggest the probable disease based on few physiological parameters and

minimum visually observable parameters of the patient.

To show the interactive treatment guidelines of suggested disease.

Patient Assessment Tools in Emergency Medicine

29

Chapter 3


This chapter includes the system known as early warning scoring system. This system can

be considered as a very important tool for assessing the patient’s risk condition. Here, we

have taken a few models amongst ICU, ED, and PH scoring system. For ED-based, we

have considered modified early warning scoring system (MEWS) and national early

warning scoring system (NEWS). For ICU - based, we included an overview of APACHE-

II and APACHE-III scoring system. For PH, we have included PHEWS scoring system.

The second section includes other primary assessment tools which are used for assessing

the patient’s condition based on minimum visually observable parameters and based on

other frequently monitored vital sign parameters in ambulatory care.

3.1 Introduction

In order to provide immediate and urgent care to the critically ill patient, there is an

overwhelming need for efficient and well-prepared guidelines for the emergency service

provider. For the purpose of initial risk level stratification, they proposed various early

warning scoring (EWS) systems. These EWS systems rely onan individual score calculated

from the range of several physiological parameters, parametric values obtained from lab

test and other visually observable parameters. The calculation of the aggregate score based

on prescribed parameters gives the estimation of the current health status of the patient.

Based on the calculated score the clinical person can take the appropriate decision for

initiating the primary treatment. EWS plays a vital role almost at each and every place in

the clinical service. Timely prediction of the deterioration in health can be a life saving for

the ICU patient. An early decision of initiating therapy in an emergency is considered as

one of the key parameters. In addition to the Intensive Care Unit (ICU) and Emergency

Department (ED) scoring system, Pre-Hospitalization (PH) is another sector which is also

equally important in improving health care service in an emergency situation. In this

Early Warning Scoring System

30

section, it includes some of the widely used ICU-based, ED-based and Ambulatory (PH)

scoring systems and their parametric comparison.

3.2 Early warning scoring system

Clinician and researcher require a robust method for prediction in a critically ill patient.

There are varieties of scoring system developed for this purpose in the Intensive care unit

(ICU), emergency department (ED), and pre-hospitalization (PH). ED-based scoring

system considers lesser parameters which are readily available from the patient, while ICU

scoring system includes more parameters which are generally available from a patient

admitted in ICU. Pre-hospitalization based scoring system is also designed in line with the

ED scoring system(Smith et al., 2014).

Generally, it is believed that the scoring system with a larger number of parameters should

perform better than lesser parameter based scoring system. But in some cases scoring

system with a lesser number of parameters performs far better than the system with a

higher number of parameters if the population is well defined. Furthermore, some of the

ED-based scoring systems perform even better in ICU as compared to ICU scoring system.

The pre-hospitalization scoring system also includes lesser parameters and used to define

the level of criticality in the patient as a part of the primary decision-making process.

The most widely used ICU based scoring system includes APACHE II and APACHE II.

While ED-based scoring system includes: MEWS and NEWS. PHEWS is the early

warning scoring system used as a part of prehospital emergency care.

3.2.1 APACHE II

The Acute Physiology and Chronic Health Evaluation (APACHE) scoring system is

widely used to assess the patient’s severity of illness in ICU.APACHE II is the upgraded

and modified version of the original APACHE. The original APACHE system uses

weighing of 34 potential physiologic measures, the addition of that results in an acute

physiologic score(APS). This system includes four letters (A, B, C, and D) to designate the

health condition of patient ranging from excellent health (A) to severe chronic organ

system insufficiency (D). The original APACHE system is complex and demands proper

multi-institutional validation. So, they proposed APACHE II system which is a simplified


31

version with higher accuracy and comprehensive patient classification system(Knaus,

Draper, Wagner, & Zimmerman, 1985).

The system developed by Knaus and his team (Knaus et al., 1985), utilizes clinical

judgment and documented physiologic relationship to select variables and assign

appropriate weights. In this system, the total number of physiologic variables has been

reduced from 34 to 12. This reduction was taken by ignoring parameter which is not

measured frequently. Further reductions were achieved by establishing a minimum set of

clinically essential variables and then carefully assessing the role of other physiologic

measurements with regard to their impact on survival. Some thresholds and weights for the

physiological variables have been changed too.

APACHE II includes Temperature, Mean arterial pressure, heart rate, respiratory rate,

oxygen, arterial pH, serum sodium, serum potassium, serum creatinine, hematocrit, WBC,

GCS score. These individual parameters are assigned with a score ranging from 0 to 4

based according to their predefined range. The score from these parameters calculate APS

points and then added to age points and chronic health points to estimate total APACHE II

score. The maximum possible APACHE II score is 71. As APACHE II system is relatively

less complex and independent of therapeutic decisions, offers an advantage over original

APACHE(Waters, Nightingale, & Edwards, 1990).

3.2.2 APACHE III

APACHE III was developed by exploring the relationship existed between acute changes

in a patient’s physiologic balance and immediate risk of death. APACHE III was formed

by upgrading the risk prediction capability of APACHE II by revising the choice of

physiological parameters and their weightin aggregate score calculation.Compared with the

APACHE II systems, the APACHEIII system employs more variables and is more

complex and time-consuming(Hsu et al., 2001; Knaus et al., 1991).

3.2.3 MEWS

MEWS is an acronym stands for Modified Early warning scoring System, proposed by the

heart of England NHS foundation trust(Heart of England NHS Foundation Trust, 2012). It

is considered as a tool used to aid recognition of deteriorating patients and is based on

physiological parameters, which are taken when recording patient observations. The


32

parameters which are included in this scoring system includes temperature, pulse, blood

pressure, and respiratoryrate, with oxygen saturation, level of consciousness and urine

output. Based on the values of the individual parameter, an aggregate score is calculated.

MEWS scoring system is shown in Table 3.1.The threshold for the system is already

defined, whenever reached, it immediately follows an extended path. This path mainly

outlines the steps required for on-time review thus ensuring appropriate treatment for

patients.TABLE 3.1: MEWS scoring system

Score

Parameter

3 2 1 0 1 2 3

Respiration rate 8 orless

9-16 17-20 21-29 30 or more

Pulse 51-100 101-110 111-129 130 or more

O2 Saturation 94% ormore

90-93 % 85-89 % 84% or less

Systolic BloodPressure

70 orless

71-80 101-199

200 ormore

Conscious level(AVPU)

Alert Voice Pain Unresponsive

Temperature (oC) 35 orless

35.1-36

36.1-37.5 37.6-38.1 38.2 ormore

Hourly urine for 2hours

Noconcerns

21-35 1-20 Nil

Based on the value of the aggregate score, the escalation pathway is divided into three

main categories: Low, Medium, and High. These three categories are ultimately pertinent

to a risk level of that patient. Low is for the score range 1 to 3, medium for 4 to 5 and high

for 6 or more. According to this category, the patient should be treated as per the

guidelines listed in the escalation pathway(Heart of England NHS Foundation Trust,

2012).

MEWS is considered to be more effective in the emergency department as it does risk level

stratification based on a minimum number of parameters. MEWS is specifically designed

for an adult patient. So it could not be used for a paediatric patient(Le Onn Ho, Shahidah,

Koh, Sultana, & Ong, 2013).

3.2.4 NEWS

National Early Warning Scoring System developed by the royal college of a

physician(NEWS, 2019). The applicability of this system was evaluated in Indian scenario


33

by a group of researcher from the department of general medicine,

Vishakhapatnam(Vanamali, Sumalatha, & Varma, 2014). While developing the NEWS

system, they consider seven parameters for calculating the score and ultimately to assess

the patient's health status shown in Table 3.2. The parameters and the reason for their

selection are discussed here:

Respiratory Rate:Sudden rise in respiratory rate is one of the most powerful symptoms

pointing to acute illness and respiratory distress. This rapidrise in respiratory rate may also

because of pain and distress, lung sepsis, Central Nervous System (CNS) disturbance, and

metabolic disturbances.If the RR falls below some predefined level it can be served as a

critical sign of CNS depression and narcosis.

Oxygen saturation: Oxygen saturation is routinely measured as a part of clinical

assessment with pulse oximetry method non-invasively. But it was not used much by the

current EWS systems. As of now, the regular monitoring of oxygen saturation is feasible; it

is now incorporated in NEWS considering its importance in risk stratification process.

Oxygen saturation is a critical indicator for assessing jointly pulmonary and cardiac

function. Pulse oximetry is used widely method used for the measurement of oxygen

saturation, due to advancement in technology which makes the instrument portable and

inexpensive.

Temperature: The changes in the normal value of body temperature on either of the

extreme value considered to be one of the key marker indicating acute illness and probable

trouble in physiological events.

Systolic blood pressure: To assess the Cardiovascular System (CVS) related problem,

variation in blood pressure serves as one of the key parameters. The increase in blood

pressure value indicates critical risk to the CVS. In fact, the reduced systolic blood

pressure, termed as Hypotension, is the key factor while evaluating the severity of

acuteillness. Hypotension signifies the effect of medication, CNS depression, cardiac

failure or arrhythmic disturbance in cardiac activity,any problem in the circulatory system

due to sepsis. In this context, diastolic blood pressure is of lesser importance due to the fact

that it does not add value so it is not included in any of the scoring systems for assessing

acute-illness severity.


34

Pulse rate: Heart rate / Pulse rate is another key indicator used to assess the patient health

status.Sudden rise in pulse rate above the nominal value (Tachycardia)may indicate

problems in the circulatory system due to sepsis (or volume depletion), pyrexia, heart

failure, and general distress. Tachycardia may also be due to the arrhythmic nature of

heart, certain disturbance in body metabolism, or toxic effect of certain drugs. Bradycardia

(reduced heart rate) is another significant physiological indicator. Sometime, certain

physical conditioning or an effect of some drug may reduce the heart rate which is quite

normal. Nevertheless, it can be served as an essential vital sign indicates the conditions

such as hypothermia, CNS depression, hypothyroidism or heart block.

Level of consciousness (LoC): Level of consciousness will be assessed by performing an

evaluation based on Alert Voice Pain Unresponsive (AVPU) scale. This scale is used to

assess the patient’s level of consciousness and serves to measure the severity of acute

illness. This AVPU scale used to monitor the response of a patient based on four possible

inputs. This is a sequential event and out of these four events only one will be recorded.

Alert: A patient, who responds by opening eyes, reacts to verbal inputs and performs

certain motor activities. Voice: A patient who responds to verbal inputs in any of the three

mentioned forms (eyes/voice/motor). Pain: A patient who reacts to a pain stimulus.

Unresponsive: It is total unconsciousness. This response is documented when a patient

doesn’t react to verbal or pain stimulus in any of the stated form.

TABLE 3.2: NEWS scoring system

PhysiologicalParameter

3 2 1 0 1 2 3

Respiration Rate ≤8 9-11 12-20 21-24 ≥25

OxygenSaturation

≤91 92-93 94-95 ≥96

AnySupplemental

Oxygen

Yes No

Temperature ≤35.0 35.1-36.0 36.1-38.0 38.1-39.0 ≥39.1

Systolic BP ≤90 91-100 101-110 111-219 ≥220

Heart Rate ≤40 41-100 51-90 91-110 111-130 ≥131

Level ofConsciousness

A V,P or U


35

NEWS Threshold and Trigger

Having defined the scores in NEWS table, the system has also defined the level of

thresholds and triggering of clinical response shown in Table 3.3(NEWS, 2019).

TABLE 3.3: NEWS threshold trigger level

NEWS scores Clinical risk

0 Low

Aggregate 1-4

Aggregate 5-6 Medium

Aggregate 7 or more High

3.2.5 PHEWS

The PHEWS (Pre-Hospital Early Warning Score) scoring system was primarily developed

to differentiate the patients who can be managed in the pre-hospital care service and

patients who require to be transferred and admitted to the hospital (North West Ambulance

Service (NHS trust), 2014). This system depends on the values of certain physiological

parameters observed during the monitoring phase. Minute variations in individual

observationslead to predicting the declining health status, acute-ill patient. This system was

proposed under the guidelines called paramedic pathfinder.

They considered following parameter: Heart rate (bpm), Respiratory rate, Systolic blood

pressure (mmHg), Oxygen saturation, Central nervous system, Temperature tympanic, BM

mmol/l (capillary), Pain score (0-10)

Parametric comparison of various EWS is listed in Table 3.4. This clearly demonstrates the

number of parameters in APACHE-III is more as compared to other EWS systems. Few

basic physiological parameters are essential to consider in every system considered here.


36

TABLE 3.4: Parametric Comparison of various EWS

SystemParameters

APACHE III APACHE II NEWS MEWS PHEWS

Heart rate √ √ √ √ √

Mean BP √ √

Systolic BP √ √ √

Temp √ √ √ √ √

RR √ √ √ √ √

SpO2 √ √

PaO2 √ √

AaDO2 √ √

Blood Glucose √

Age √ √

GCS visual √ √

GCS speech √ √

GCS motor √ √

AVPU score √ √ √

Pain score √

Urine output √ √

WBC √ √

Hematocrit √ √

pH √ √

pCO √

Primary comorbidities √

Serum creatinine(without ARF)

√ √

Serum creatinine (withARF)

√ √

Serum Na √ √

Serum potassium √

Serum albumin √

Serum bilirubin √

Serum glucose √

Serum BUN √


37

3.2.6 Discussion

None of the ICU based available scoring systems appears to be suitable for bedside

assessment of ward patients in a routine fashion. NEWS and MEWS are proven to be a

versatile tool in this context. NEWS and MEWS score can be calculated based on a few

parameters readily measured or recorded.

The ICU based scoring system is based on an assessment of more number of patient’s

physiological parameters and it outperforms in certain cases too. But when a patient is in

the Emergency Department, it is the situation which needs urgent attention and immediate

action. For that reason, the decision should be taken from minimum parameters possible.

So the EWS designed for ED should be slightly different and it should include minimum

parameters which are mostly measured as a part of emergency (Vanamali et al., 2014).

Most of the EWS developed are based on some specific population, so the applicability of

those EWS in other population needs to be assessed properly. In some cases it needs to be

developed some dedicated EWS for some specific population too, and verify its efficacy

and applicability in that specific region. The accuracy of any scoring system is highly

dependent on the quality of the input. Accuracy of a decision taken by the system depends

on an understanding of the definitions of related terms by the user, time of data collection,

rules used for missing data at the time of collection of data, and so on must exactly match

those applied when building the model.

Primary assessment tools in Emergency health care system

38

3.3 Primary assessment tools in Emergency health care system

The primary Assessment of patient is the most critical part of the emergency health care

system. The vital sign monitoring of the patient should be during immediately during the

transportation through ambulance. The frequency of monitoring vital sign mainly depends

on the actual status of the patient. If the patient’s condition gets worse than monitoring

should be done morefrequently.

This assessment is performed either by asking the question to the patient or by observing

the patient situation visually. In addition to this few assessment tools utilizes values of vital

sign parameters also to get the results. These assessment tools help to determine the current

situation of the patient and to predict the probable disease as well. This will guide the

appropriate path to paramedic to initiate the second level of treatment. It makes sure that

treatment can be started before the patient reaches the hospital, which further increases the

chances of a patient’s survival (Pre-hospital emergency care council, 2014).

3.3.1 Perfusion Status Assessment

“The ability of the cardiovascular system to provide tissues with an adequate oxygenated

blood supply to meet their functional demands at that time and to effectively remove the

associated metabolic waste products” is called perfusion. Perfusion assessment is carried

out from the Table 3.5(Victoria, 2018):

TABLE 3.5: Perfusion status assessment

Status

Para

Adequate Borderline Inadequate Extremelypoor

No perfusion

Skin Warm,Pink,Dry

Cool,Pale,Clammy

Cool,Pale,Clammy

Cool,Pale,Clammy

Cool,Pale,Clammy

Pulse 60-100 50-100 <50 or >100 <50 or >100 No PalpablePulse

BP >100 80-100 60-80 <60 Unrecordable

Conscious State Alert &Oriented totime and place

Alert &Oriented totime and place

Either Alert &Oriented to timeand place oraltered

Altered orUnconscious

Unconscious


39

3.3.2 Respiratory Status Assessment

The respiratory status gives the indication of the breathing pattern and breathing capability

of the patient. Respiratory status assessment is possible from the Table 3.6(Victoria, 2018):

TABLE 3.6: Respiratory Status Assessment

Status

Para

Normal Mild distress Moderate distress Severe distress (life threat)

GeneralAppearance

Calm, quite Calm or mildlyanxious

Distressed oranxious

Distressed, anxious, fightingto breathe, exhausted,catatonic

Speech Clear and steady Full sentences Short phrases Words or unable to speak

Breath Sound Quiet Able to cough Able to cough Unable to cough

Wheeze No Wheeze Mild expiratorywheeze

Expiratory wheeze,+/-inspiratorywheeze

Expiratory wheeze, +/-inspiratory wheeze or nobreath sounds (late)

Crackles No crackles orscattered finebasal crackles

Crackles at base Crackles at bases tomid-zone

Fine crackles-full field, withpossible wheeze

Respirationrate

12-16 16-20 >20 >20 or <8

Respirationrhythm

Regular evencycles

May have slightlyprolongedexpiratory phase

May have slightlyprolongedexpiratory phase

Prolonged expiratory phase

Work ofbreathing

Normal chestmovement

Slight increase inchest movement

Marked chestmovement +/- useof accessorymuscle

Marked chest movement withaccessory muscle use,intercostal retraction, +/-tracheal tugging

HR 60-100 60-100 100-120 >120 or Bradycardia late sign

Skin Normal Normal Pale and sweaty Pale and sweaty, +/- cyanosis

Consciousstate

Alert Alert May be altered Altered or unconscious

Contribution to this research work

40

3.3.3 Conscious State Assessment (Glassgow Coma Scale)

The GCS is the neurological scale which is used to measurethe level of consciousness. The

score is actually calculated by examining the response of the patient to varieties of input.

The response is recorded in the form of eye, verbal and motor activities as per Table 3.6.

The main thing here is the patient should get the maximum score in an individual category

based on the recorded response. The higher score indicates that the patient’s higher state of

consciousness. The entire process is performed by the medical professional even in case of

the pain stimulus (Victoria, 2018).TABLE 3.7: GCS score

Parameter Response Score

Eye Opening None 1

To pain 2

To voice 3

Spontaneous 4

Motor Response None 1

Abnormal extension to pain 2

Abnormal flexion to pain 3

Withdraws from pain 4

Localizes to pain 5

Obeys command 6

Verbal Response None 1

Incomprehensible sounds 2

Inappropriate words 3

Confused 4

Oriented 5

3.4 Contribution to this research work:

In the proposed system, we have selected NEWS scoring system considering its

effectiveness and viability for the assistance of paramedic in emergency health care

service. A team of researcher from the department of general medicine, Khammam,


41

Andhra Pradesh, performed an assessment of NEWS scoring system in the Indian scenario.

The team has proved and recommended that this system could perform equally well in the

Indian scenario and should be considered as an aid to clinical assessment and judgment

(Vanamali et al., 2014). As per GVK EMRI EM Care report(GVK Emergency

Management and Research Institute, 2016), the parameters suggested by the NEWS

scoring system are also the parameters which paramedics are recording as a part of

ambulatory care and monitoring. The differential diagnosis requires the usage of other

patient assessment tools which are listed above. These tools utilize the current value of the

physiological parameter along with some visually observable symptoms. This method

ultimately brings out some valuable conclusion, which helps the paramedic to decide the

course of treatment.

3.5 Summary

Early warning scoring systems are an essential part of primary risk level stratification of a

patient. In this chapter, we have discussed various ICU, ED, and PH based scoring system

and their effectiveness as a risk assessment tool. Other than that, Patient assessment tools

which are used for the purpose of differential diagnosis are also included in the latter part

of the chapter. These tools are used to assess patient’s perfusion, respiratory and

consciousness status.

Tools of Knowledge-Based System

42

Chapter 4


This chapter includes different tools for effective designing and development of a

knowledge base system. The chapter starts with one of the most widely used knowledge

management and structuring system is known as CommonKADS. The details of different

types of CommonKADS framework models are discussed in detail in that section. The

models are designed for the particular emergency medicine expert system considering all

domain related concepts and their interrelation. The latter part of the chapter includes

Introduction about semantic web and ontology for generic and interoperable machine

understandable system. In the latter part of the chapter, it will discuss different ontology

languages and ontology editors available in the market with their comparison. The chapter

ends with the discussion and summary.

4.1 CommonKADS: AModeling Approach for Knowledge Engineering

4.1.1 Introduction

Conventionally, the branch of knowledge engineering was considered as a process of

“eliciting” or “mining from the expert’s head” and transferring it in a machine in

computational form. Knowledge engineering has evolved from the late 1970s

onwards(Kingston, Shadbolt, Tate, & others, 1996). This view is considered to be

rudimentary and relatively immature. Today, Knowledge engineering is more of a

modeling activity. A model is a purposeful abstraction of some part of reality. Modeling is

a process of developing a good description of only certain important aspects of knowledge

and ignoring the rest. In CommonKADS, a knowledge project includes the construction of

models which are a significant part of the products conveyed by the project (Milton,

Shadbolt, Cottam, & Hammersley, 1999; Schreiber, Wielinga, de Hoog, Akkermans, &

Van de Velde, 1994; Weih, Schu, & Calmet, 1994).

CommonKADS: A Modeling Approach for Knowledge Engineering

43

4.1.2 CommonKADS Modelling Framework

This CommonKADS framework includes six models to construct the knowledge-based

system. These models are Organizational, Task, Agent, Communication, Knowledge and

Design model. This particular framework is proposed to develop an expert system for the

emergency health care system with the ability to provide a decision on clinical risk to the

patient and disease probability with treatment options. This System in this context will be

called a decision support system for emergency medicine (DSS-EM). The modeling

approach of Knowledge Base Systems (KBS) offers modular architecture, which helps to

fragment the entire problem of knowledge engineering in smaller possible tasks. This

makes this approach very much popular and acceptable in the Knowledge Engineering

(KE) communities(J. Patel & Bhatt, 2014; Wielinga, Schreiber, & Breuker, 1992). The

subsequent section discusses these models in detail.

4.1.2.1 Organizational Model

This model offers an analysis of the socio-organizational environment in which the KBS

will have to function. In this context, the model proposed here includes the major

contributors helping to develop this emergency medicine DSS. As shown in Fig.4.1, the

knowledge engineer gets the knowledge from the knowledge provider which is the main

source of information for this particular system. In this case doctors, EM experts and

scientists are the knowledge providers. The gathered knowledge is structured and

organized in a systematic way by knowledge engineer and with the help of knowledge

system developer, the computer-based EM-DSS will be designed and implemented. DSS

will assist the end-user, in this case, they are EM-staff, to take necessary, relevant and

quick decision based on the information available with them and ultimately to facilitate the

patient (J. Patel & Bhatt, 2014).


44

FIGURE 4.1: Organizational Model for Emergency Medicine Expert System

4.1.2.2 Task Model

This model specifies how the function of the system can be achieved by performing a

number of tasks. An identified task can be decomposed into sub-tasks. Individual task

listed is presented with input/output specification. Here, the input represents the

information utilized for getting the output and the output symbolizes the goal which needs

to be achieved.

For the given system as shown in Fig.4.2, three main tasks are identified: i) Clinical risk

detection (risk level stratification) ii) EM score-based disease forecasting (differential

diagnosis) iii) Treatment suggestion. Clinical risk detection is based on a score calculated

from early warning scoring system. In this case, NEWS scoring system is preferred

because of its clinical efficacy and applicability in the prehospital health care system. EWS

is a triage-tool for taking a primary decision, in an emergency department or in the

intensive care unit, based on the values of major vital parameters of a patient. This

individual score helps to calculate the total aggregate score. Based on the total score the

risk level of an individual patient can be accessed. This risk detection task also takes the

information from the priorities and disease classification defined by emergency medicine


45

guidelines. This particular task is performed by risk level ontology. A second important

task is forecasting of probable disease from risk level ontology (based on the calculated

score) and decision ontology.The differential diagnosis takes help of different primary

assessment tools which requires minimum visually observable parameters along with

values of few physiological parameters. This ontology is developed by the knowledge

engineer by eliciting knowledge from the domain experts. The third task will be carried out

by rule-based expert system designed from the data of clinical risk, disease and other

conditions of a patient.

4.1.2.3 Agent Model

Agent model constitutesall agents which are included in the process of problem-solving.

So, this model specifies who does the tasks specified in a task model. It illustrates the

characteristics of agents. An agent in CommonKADS can be a human, a robot or a

software program.

As depicted in Fig.4.2, a total of four agents is needed to achieve the tasks listed in the task

model. The first task needs two agents of and second & third task needs one agent

respectively. The first agent forclinical risk detection, a score calculated by the emergency

medicine based EWS (NEWS-scoring system) plays a main role. The second agent called

EM disease classification will also help to assess the priorities involved in clinical risk. For

disease prediction, EM-based decision ontology along with disease forecasting agent (third

agent) helps to gather disease probability. Fourth agent, rule-based expert system, helps to

identify the possible treatment based on the disease condition and risk levels of an

individual patient.

4.1.2.4 Communication Model

In a knowledge-based system, the communication model becomes more important than in

the normal expert system.This model indicates communication between agents. In Fig.4.2,

the third agent requiresdata from the second agent in order to do disease forecasting. In

addition to this, agents require database from the external world that is also indicated in

fig. The communication model basically gives the direction of flow of information

amongst the agents and with the outside world too.


46

FIGURE 4.2: Agent, Task and Communication Model for Emergency Medicine Expert System

Where, AVPU status = Alert/Verbal/Pain/Unresponsive

EM = Emergency Medicine

MNEWS = Modified National Early Warning Scoring System

RT score = Real-time score


47

4.1.2.5 Knowledge Model

This is probably the most important model amongst other models of commonKADS. This

model also refers to as expertise model. It contains three knowledge categories: Domain

layer, Inference layer, and Task layer.

Domain layer represents, modeling of domain-specific knowledgewhich is required to

perform the task. It includes conceptualization of domain in the domain ontology.

Inference layer describes the most basic reasoning steps. Task layer specifies the goals of

the reasoning process and the strategies to achieve these goals.

In order to construct the Knowledge Model, it requires three stages to be followed: i)

Knowledge Identification ii) Knowledge Specification iii) Knowledge Refinement(Yang,

Tong, Ye, & Wu, 2006).

This model particularly depends on the knowledge available with the experts. It needs to

get as much as information available from the domain expert and convert that knowledge

into an appropriate form. For that purpose, it proposes a knowledge elicitation form (Table

4.2) for different diseases of Emergency Medicine department. The knowledge Engineer

has to acquire all this information from the expert. Table 4.3 also lists the most commonly

practiced treatment in EM department as a part of the primary treatment of the patient.

While Table 4.1 helps to stratify the clinical risk detection based on the score calculated by

the EWS.

TABLE 4.1: NEWS table for score calculation and clinical risk determination

NEWS scores Clinical risk

0 Low

Aggregate 1-4

RED score

(Individual parameter scoring 3)

Medium

Aggregate 5-6

Aggregate 7 or more High


48

TABLE 4.2: Knowledge elicitation form for diagnosis of EM disease

Disease Name: Chronic obstructive pulmonary disease

Physiological Parameter (Abnormal Value): HR: 97 BPM

RR: 17 per minute

BP: 160-120 mmHg

BT: 35.5oC

SPO2: 92.4 %

Age dependencies: Yes

Prior History of a patient and its relevance tothe disease:

Yes (Social circumstances, quality of life, currenttreatments, smoking

EM category (Priority) High

TABLE 4.3: Knowledge elicitation form for possible treatment of EM disease

Primary treatment

List of possible theory IV lines: Hydrocortisone 200mg IV may be given initially if the oralroute isnot appropriate

Medication: Antibiotics (Amoxicillin 500 mg oral tds)

Defibrillation: Not required

Airway management: Bronchodilators (Nebulised salbutamol 5mgand ipratropium bromide 500NIPPV (Positive Ventilation)

microgrammes should be given on arrival and repeated 4-6 hourly.)Any other primary care:PEFR and start PEFR chart.FBC, U&Es.Sputum and blood cultures.12 lead ECG.

Age dependencies: Yes

Prior History of patient’s therapy andits relevance

Reaction or allergy to some specific medicines

4.1.2.6 Design Model

This model is used to map the abstract study carried out by the knowledge and

communication models to preciseexecution. It states the most appropriate software and

targeted hardware needed for the implementation purpose. This model also enlists the

technical and functional specifications, different software incorporated in the intended


49

system, the various concepts and relevant components found during the initial phase of

analysis.

4.1.3 Discussion

Web-based EM DSS is primarily designed for the purpose of assisting the emergency

medical staff in order to initiate timely treatment of the patients. The nature of the problem

which the EM staff faces in their day to day work is very complex. CommonKADS helps

to fragment this entire complex system into a smaller and modular architecture which

ultimately simplifies the development process. The CommonKADS methodology is used

to implement many of the design thoughts by addressing the knowledge modeling

paradigm. It also offers consolidated architecture to execute knowledge engineering and

management projects. Each of these models was created by meticulous exercise and

incredible exertion. The organizational model of CommonKADS is the most important

model. This model is the actual backbone of the system. Agent model along with

communication model helps to form the problem-solving architecture for a particular task.

The task model is developed by considering the three most important tasks which this

system has to perform. They are Risk level stratification, Differential diagnosis, and

Treatment guideline. Each task is performed either by one or more agents, which is

specified in the agent model. These agents are communicating with each other described in

the communication model. The different model developed through this approach would

streamline the process of developing the final design of the system. Knowledge model is

the most important model amongst all, used to elicit the knowledge needed to develop the

system through knowledge elicitation tool. Development of commmonKADS framework

models for this expert system is the most valuable and effective accomplishment of this

work. These models are proven to be actual pillars needed to construct the web-based EM

DSS.


50

4.2 Introduction to Ontology

One of the most important challenges that need to be solved by the IT field is“to provide

the rightinformation to the right person at the right time”. In order to achieve this goal, it

requires flawlessinteractions between people, software agents and othertypes of IT

systems. This kind of association is needed to assist the evolution of vibrant communities

that can exchange and utilize effectively the full range of data, information, knowledge,

and wisdom. Tim Berners-Lee has suggested the solution for this challenge in the form of

Semantic Web(Uschold&Gruninger, 2004). The Semantic Web has the ability to express

web content in a form which can be read, recognize and utilized by software agents. The

semantic web is actually the extensionof the existing World WideWeb (WWW) which

provides an integration of information amongst various software programs easily. Thus,

they can exchange, share and integrate information with ease(Karoui, Aufaure, &Bennacer,

2004).

The rapid development in ICT makes the existing web an enormous source of data and

services. Thus, it becomes essential to share this massive source information amongst

people as well as applications. Ontology is an integral part of the semantic web. Ontology

defined the data set which characterizes the concepts and in the prescribed domain and

their interrelation. In other sense, Ontology is a hierarchicalstructure of the most important

concepts related to a particular domain,the relationship between classes and their

properties. The sharing of information between the existing web and semantic web is

facilitated with the help of ontology. It plays a significant role by linking the existing web

information to the semantic web with semantic interoperability. Ontology also gives a

consensus understanding of a particular domain which makes the information exchange

possible amongst people along with diversified and distributed systems.

The semantic web was proposed by berners-lee. It offers access to information

automatically based on the semantics of data which is machine-processable. It also uses

heuristics of metadata for automated information access. Using domain theories

(ontologies) with this clear illustration of the semantics of data makes a web to offer an

entirely innovative way of knowledge. It has the ability to knit together this large network

of human knowledge with machine processability. Ontologies have been used widely in

Introduction to Ontology

51

the field of computer science and specifically in the domain of semantic web for getting a

comprehensive and machine interpretable common understanding(Berners-Lee, Hendler,

Lassila, & others, 2001).

Ontologies organize the domain semantics by stating their components; thus they contain

the concepts which defined the inner attributes of the stated concepts and the properties to

describe their interrelationship. Ontologies are developed from the common vocabularies

shared and agreed amongst the knowledge developers.These characteristics of the ontology

make it suitable to utilize in various tasks of the diversified field of

research(Maedche&Staab, 2001).

4.2.1 Ontology

Ontology definition given by Gruber(Gruber, 1993) defined is as “a formal, explicit

specification of a shared conceptualization”.Here, Gruber gave more stress on the

formalizing the specification of concepts and their interrelations; ultimately it permits

representation of the knowledge & sharing that amongst different agents. Later Studer and

his fellow researcher (Studer, Benjamins, & Fensel, 1998)investigated thisdescription and

perceived that there are four main concepts; formal, explicit, shared and conceptualization.

Formal means ontology should bemachine understandable; explicit states that all concepts,

properties, relations, functions, constraints, and axioms used are defined explicitly; shared

implies the ontology should confine accepted and agreed knowledge in the communities;

and conceptualization presents a conceptual model and a simplified view of some

phenomenon in the world that we want to represent.Guarino(N. Guarino, 1998) has also

given another definition of ontology: “a set of logical axioms designed to account for the

intended meaning of a vocabulary”. Where, Guarino focused on the role of logic theory as

a way of representing an ontology(Corcho, Fernández-López, & Gómez-Pérez, 2003; A.

Gómez-Pérez & Corcho, 2002; Liu & Zsu, 2009).

In general, Knowledge in the ontology can be described using five basic components:

classes, instances, relations, functions, axioms, and instances.

Classes or concepts: In a wide sense, they are agroup of individuals sharing

common attributes. A concept can be an explanation of a task, function, action,

policy, reasoning process, etc. Most of the ontology languages use the definition of

concepts as per these characteristics.


52

Relations and functions:It defines the way in which concepts and attributes are

related to each other in the stated domain. In other words, it represents a type of

communication between concepts in a similar domain.

Axioms: Axioms are an important component of ontology. It represents a logically

formulated assertion that includes the core knowledge which ontology wants to

illustrate in particular domain. Strictly speaking, axioms are utilized for modeling

the sentences which are generally true. The classification of types of axioms is as

pertheir semantic meaning.

Instances: It is also known as individuals. These individuals are used to model

concrete objects and signifies the core components of an ontology.

4.2.2 Advantages of using ontology

• It gives a common semantics.

• Ontologies are machine interpretable andthey can be shared amongst several

people.

• People can work in collaboration without any uncertainty or loss of information.

• It is generic and reusable.

• In order to develop a big ontology,it is possible to integrate a few small size

existing ontologies pertaining to a specific part of the field.

• It is possible to reuse Top-level Ontology and extend it to depict a particular area of

interest.

• It offers easy maintenance with minimal cost.

• Because of the characteristic called "linked data" of ontologies, the inference

process becomes easy and automatic.

4.2.3 Principles for the Design of Ontologies

The principles of design of Ontologies are listed below (J. M. Gómez-Pérez & Ruiz, 2010):


53

Clarity: Necessary to converse on the exact meaning of defined terms.In order to

communicate ontology effectively, it is necessary to define the meaning of different

clearly. A definition should be stated on formal axioms, should be objective, and

distinctly conveyed by necessary andsufficient conditions, or defined by only

necessary or sufficientconditions.

Coherence: To allow inferences those are consistent with definitions.It was stated

that “AnOntology should be coherent: that is, it should sanction inferences thatare

consistent with the definitions. If a sentence that can be inferred from the axioms

contradicts a definition or example given informally,then the Ontology is

incoherent” (Mellouli, Bouslama, & Akande, 2010).

Extendibility: It is necessary for anticipating the utilization of the shared

vocabulary.It was stated that“there should be provision to introduce new terms

based on available vocabulary,in a way which does not need the modification of the

available definitions” (Calero, Ruiz, & Piattini, 2006).

Minimal Encoding Bias: It makes the ontology independent at the symbolic

level.It was said that “The conceptualization should be stated at the knowledge

level avoiding its dependence on symbol-level encoding” (Khosrow-Pour, 2006).

Minimal Ontological Commitments: It is required to propose minimum possible

claims about the world.It was said that “Since ontological commitment is based on

the consistent use ofthe vocabulary, ontological commitment can be minimized by

specifyingthe weakest theory and defining only those terms that are essential tothe

communication of knowledge consistent with the theory”(Gomez-Perez,

Fernández-López, &Corcho, 2006).

The representation of disjoint and exhaustive knowledge:If the set of subclasses

of a concept are disjoint, it can be defined in a category of a disjointdecomposition.

The decomposition is exhaustive if it defines thesuperconcept completely.

To improve the understandability and reusability of the Ontology:The

implementation of the given ontology should be such that it tries toreduce the

syntacticdistance among sibling concepts.

The standardization of names:The identical naming standardsshould be

maintained in ontology for making the effortless understanding of ontology.


54

4.2.4 Types of Ontology

Ontologies can be classified into three given categories (Nicola Guarino, 1998):

4.2.4.1 Top-level / Upper / Foundation Ontologies

The aim of proposing top-level ontology is to have a number of Ontologies accessible

under the roof of top-level ontology. It is used to express very general concepts lie matter,

time, event, etc., which are not dependent on a specific domain. This makes it possible to

integrate top-level ontologies for other users. It means upper ontology is used to express

very common concepts that are almost identical across all knowledge domains.

Few examples of Top-level Ontology:

WordNet: English language lexical database. It can be accessed as a dictionary

too. It includes somewhere around 80,000 concepts.

SUMO (Suggested Upper Merged Ontology): It is extended withmany domain

Ontologies and a complete set of links to WordNet.

BFO: Top level ontology in the biomedical domain. It has 36 classes and they are

related via “is_a relation”.

Cyc: It is developed for Network RiskAssessment, Natural Language Processing,

and Terrorism Management. It includes 3,00,000concepts.

GFO: It is developed for the biomedical domain. It includes 79 classes.

4.2.4.2. Domain Ontologies / Task Ontologies /Domain-Specific Ontology

As the name indicates, it is used to illustrate the terminology associated to a generic

domain like automobiles, medicine or biology; or any basic task or activity like acquiring

or diagnosing or suggesting by special phrase included in the upper ontology. For example,

PEN word has two distinct meanings.Ontology developed for the university or student

domain "ball pen"meaning of the word; on the other hand, ontology about the computer

hardware domain would have "Pen drive" meanings.

4.2.4.3. Application Ontologies

It is used to develop the concepts related to a specific domain and task. For e.g.Agricultural

Ontology which describes different concepts related to Agricultural domain. In the


55

application domain, it is possible to integrate Top-level Ontology aswell as Domain

Ontology with Application Ontology.

4.2.5 Ontology Languages

There are various formal languages available in computer science to create ontologies.

These languages serve the purpose of knowledge encoding of an ontology in the easiest,

formal and human-understandable way. These languages are declarativelanguages and the

major changes in them are their degree of expressivenessand the inference engine ofthe

language. The comparison of these ontology languages is possible based on the featuresand

elements which are provided for identifying the ontology knowledge(A. Gómez-Pérez &

Corcho, 2002).

The classification of ontology languages can be done in two parts (Corcho & Gómez-

Pérez, 2000): Traditional ontology language and web-based semantic ontology language.

4.2.5.1 Traditional Ontology Languages

Even before the introduction of the semantic web, these languages were in existence. In the

1990s, these languages were introduced. They are mainly created on the basis of first-order

logic and on a certain combination of it with frames. For example KIF (Knowledge

Interface Format),OCML (Operational Conceptual Modeling Language), F-Logic (Frame

Logic).

4.2.5.2 Web-based semantic ontology languages

These languages were introduced after the evolution of the Semantic Web. Most of these

languages are following the XML syntax and based on their structure they can be further

classified in different sub-types.

The below section discusses a few most popular types of ontology mark-up languages.

XOL (Ontology Exchange Language): This ontology language was earlier designed

for the exchange of bioinformaticsontologies; later on these XML based-languages

are used for creating an ontology for other domain.

SHOE (Simple HTML Ontology Extension): This was the extension of HTML

utilized to incorporate knowledge inside web pages. Afterward, this language was

converted to XML base structure.


56

RDF (Resource Description Framework) (“RDF/XML Syntax Specification

(Revised),” n.d.): This was the first language proposed bythe W3C (the World

Wide Web Consortium) for depicting resources ofthe web. RDF expression follows

thearchitecture which comprises of triples, including a subject, a predicative and

object. These triples make a set and that set is called RDF graph. These elements

are recognized by an identifier known as URI (Uniform ResourceIdentifier).

RDF Schema: this is an extension of RDF and its combination with RDF which in

this case knownas RDF(S). This language is usedto integratewith others, a way of

expressing associationbetween classes andproperties.

DAML-OIL: This language is developed with the help of RDF triples and created

by combining features from two languages i.e. DAML (DARPA AgentMarkup

Language) and OIL (Ontology Inference Layer).

OWL (Ontology Web Language) (“OWL Web Ontology Language Guide,” n.d.):

OWL language is a vocabularyextension of RDF. The structure of this language

defines a particular domain in terms of classes and properties. It also offers a set of

axioms to state assumptions or correspondence with respect to classes or defined

properties.

4.2.5.3 Selection of ontology language

There are different ontology languages available for the development of the ontology.

Some of them are discussed in the previous section. The comparison of some popular

ontology languages is shown in Fig. 4.3(Cardoso, 2007). It is evident from the figure, that

OWL is the most widely accepted ontology language. So, in this particular expert system,

the ontology is developed in the OWL language.


57

FIGURE 4.3: Percentage of ontology languages currently used by a user

4.2.5.4 OWL language:

OWL is one of the most trusted and widely used standard ontology languages today. It

follows the structure of XML format and it is believed to be a semantic upgrade of RDF.

RDF is an XML-based framework used to describe web information. Based on the degree

of their expressiveness, OWL languages are classified in three sublanguages: OWL Lite,

OWL DL and OWL Full. OWL Lite supports those users primarily needing a classification

hierarchy and simple constraints. OWL DL supports those users who want the maximum

expressiveness while retaining computational completeness and decidability. OWL full is

meant for users who want maximum expressiveness. OWL contains classes, properties,

relations and individuals it supports a reasoner-based inference (“OWL Web Ontology

Language Guide,” n.d.).

OWL is an object-oriented ontology language. For describing ontology in OWL, the

following basic elements are required:

Classes: They form the base for the description of the domain. These classes signify

various entities or concepts related to a particular domain.

75.9%64.9%

17.0%12.0%11.8%

3.7%2.6%2.6%2.2%1.9%1.9%1.7%

0.9%0.9%0.9%

11.8%

OWLRDF(S)

Desciption LogicDAML+OIL

FlogicWSML

Ontolingua/KIFCommon LogicSemantic Net

SHOEOKBC

CyclXOL

OCMLLOOMOther


58

Individuals:Individuals are the elements of the class. A class isa collection of

properties used to define the instances in the form of individuals. Individuals with

common characteristic are classified in a single group with the help of classes.

Properties: OWL has two types of properties:

Object Properties:Therelationships between individuals of twoclasses are

described thought the object properties.

Data Properties: It is used to provideconnection of individuals with a particular

type of data values.

In addition to this, OWL also provides annotation property used to describe annotations.

These annotations are provided on classes, properties,individuals and ontology headers.

4.2.6 Reasoning

The ontology offers a unique advantage over other existing technologies of data

representation by enabling knowledge inference from the data. Reasoning plays an

important role in knowledge extraction from the ontology, even if it is not clearly stated in

the actual knowledge base. Therefore, the process of reasoning can be stated as a sequence

of processes that allows the finding of hidden facts in the ontology specified through

explicitlydefined facts. A reasoner is generally should be able to perform these tasks:

Ontology consistency, Satisfiability of a concept, Instance checking: Retrieval and

Realization of individuals Classification, Conjunctive Query Answering:

4.2.7 SPARQL

SPARQL performs a very significant role while using ontology. Ontologies are developed

in such a way that it enables the extraction of information by applying some queries. These

queries are used for ontology evaluation and validation. Queries are used to check the

response of ontology to various questions.

As discussed earlier, RDF is used to represent distributed data and used as a standard

model information exchange over the web. It follows a triple format to illustrate the

interrelationship among two things. SPARQL is used to access RDF data. SPARQL is an

abbreviation for SPARQL Protocol And RDF Query Language (“SPARQL Query

Language for RDF,” n.d.).


59

SPARQL has standard syntax containing two main keywords: They are SELECT and

WHERE. A SELECT query has two sections: a set of question words and a question

pattern. While the keyword WHERE is used to define the pattern of selection, it is to be

written in brackets. A SELECT keyword is followed by the symbol of “?” and question

wordand WHERE keyword in brackets followed with some information and relationships

of the question word which are required to query.

4.2.8 Ontology Editors

There are various tools available for creating an ontology. Some of the popular ontology

editors are listed here:

4.2.8.1 Protégé

Protégé, developed at Stanford University, is the most popular ontology development tool

today. Protégé is an open-source editor and it is supported bya large community of active

users. Various domain experts working in medicine and manufacturing domain, utilize this

tool for modeling domainand for developing knowledgebase systems. The add-ons

(ontology visualization, project management, software engineering, and other modeling

related tasks) provided with protégé and its intuitiveness makes it one of the most widely

accepted editor in the community(“protégé,” n.d.).

4.2.8.2 OntoEdit

The Knowledge Management Group of the AIFB Institute has developed the editor known

as OntoEdit. This particular environment is suitable for creating an ontology. It also offers

creating, browsing, maintaining and managing ontologies.The client/server architecture of

this editor makes it a collaborative development environment forontologies. This

architecture supports the management of ontology located in central server while clients

can access these ontologies from different locations. OntoStudio is the current descendant

of OntoEdit. It is developed as a marketable product based on Integrated Development

Environment (IDE) Eclipse(“DAML Tools,” n.d.).

4.2.8.3 DOE

DOE (Differential Ontology Editor) is a simple ontology editor and was developed by the

INA (Institut National de l’Audiovisuel - France). DOE has a classical formal specification


60

process. DOE only allows the specification part of the process of structuring ontology.

DOE is rather a complement of others editors(“Differential Ontology Editor,” n.d.).

4.2.8.4 WebODE

A group of ontological engineersfrom the Technical University of Madrid has developed

WebODE. It is considered to be an ontological engineering workbench which performs

different ontology related tasks. This tool also has almost all types of functions to facilitate

the process of ontology development and other ontology related tasks such as ontology

edition, navigation, documentation, merge, reasoning, etc.(“WebODE,” n.d.).

4.2.8.5 pOWL

pOWL is another open source ontology management tool based on PHP technology.It also

offers parsing, storing, querying, manipulating, versioning, serving and serialization of

RDFS and OWL-based knowledge collaborative Web-enabled environment(“pOWL - A

Web-Based Platform for Collaborative Semantic Web Development,” n.d.).

4.2.8.6 SWOOP

SWOOP is the first ontology editor which offers a web-based environment for ontology

management. It supports the management of multiple ontology environment with the

ability of OWL validation. In addition to this, it has a facility to look for various views of

OWL presentation.Thus, it makes this platform suitable to compare and merge ontologies

with the facility of editing. SWOOP’s interfaceoffers easy & simple navigation and it

doesn't follow a strict procedural method for the development of ontology (“SWOOP -

Semantic Web Standards,” n.d.).

4.2.8.7 Selection of Ontology Development Tool

Amongst all the editors listed above to develop the ontology, protégé is most widely used

by researchers, professionals, programmers, developers and other people associated with

this community. As per the above discussion, there are a lot of ontology editors available.

As per the source, there are more than seventy different ontology editors available. Some

of them are discussed in the previous section.

The comparison of protégé with some of other popular ontology editor tool is given in Fig.

4.4, It is evident from this statistical chart that protégé has the highest market share


61

(Cardoso, 2007). And because of that reason, protégé is selected in this work for

developing ontology.

FIGURE 4.4: Percentage usage of ontology editors by respondents

4.2.9 Steps for creating ontology in Protégé(Noy, McGuinness, & others, 2001):

1) Determine the domain and scope of the ontology: This is the first step where one

has to define the domain and scope of application for which ontology needs to be

designed.

2) Consider reusing existing Ontologies: It is advisable to consider the usage of

existing ontology in our specific domain with minor refinement it needed. Reusing

existing ontologies may be a required if system demands interaction with other

applications which are already dedicated to particular Ontologies.

3) Enumerate important terms in the ontology: It is necessary to pin down a list of

all terms using that we will make statements or we may covey it to a user.

4) Define the classes and the class hierarchy: There are three popular approaches:

First is a top-down development process which initiates by defining the most

common concepts in the domain and following specialization of the concepts. The

second is bottom-up development process which starts by defining the most

explicit classes, the leaves of the hierarchy, following grouping of these classes into

68.2%13.6%

12.2%10.3%10.3%

9.1%7.3%

5.5%4.9%

3.7%3.7%

2.8%1.8%1.6%

ProtégéSWOOP

OntoEditText editor

Altova SemanticWorks 2006OtherOilEd

OntoStudioIsaViz

webODEOntoBuilder

WSMO StudioTop Braid Composer

pOWL


62

more common concepts. The third approach is acombination development process

which combines the top-down and bottom-up approaches: In this, it starts by

defining the most significant concepts first and then generalize andspecialize them

suitably.

5) Define the properties of classes—slots: it includes assigning the different

properties of slots.

6) Define the facets of classes—slots: it includes defining slot value type and its

Domain and range.

7) Create instances:it includes creating individuals under the defined classes.

4.2.10 Discussion:

The above sections have highlighted the concept of ontology and the reason for

incorporating ontology in this expert system. The development of the risk level ontology

and disease ontology is the first major task of this emergency medicine based expert

system. The initial step in the process of developing an ontology requires concepts,

attributes, relationship, and axioms as a part of the analysis phase. After identifying the

scope of the existing ontology and discovering all possible concepts related to the initial

phase, it is the next important thing is to identify the most suitable ontology language to

formalize the ontology. After satisfying the above requirement, the last and important

phase is to identify and select the most suitable ontology editor. For this system, a protégé

tool is selected to develop the ontology with OWL as an ontology language. During the

development of the ontology, there were series of sessions conducted with experts in

emergency medicinewith some predefined questionnaires and reviewed a number of papers

and guidelines/manuals for the purpose of information gathering and knowledge

extraction.

4.3 Summary:

In this chapter, the first section discussed the concepts of CommonKADS, a system for

knowledge management in a most structured way. Various models used to represent

CommonKADS framework for emergency medicine expert system are presented and

discussed in detail in that section. These models are considered to be an essential part of

knowledge engineering. This approach of model construction required rigorous efforts for

extracting the valuable information needed for constructing the expert system. The


63

framework of commonKADS for EM-DSS has proven to be very effective in the actual

implementation of the system. The second section described the basics of semantic web

and ontology. In that section, the various components ontology and ontology development

process are discussed in detail. The third section includes an introduction to various

ontology languages and their comparison. It is followed by various ontology editors and

their comparative market share.

Architecture and Development of Expert System - Meditrace

64

CHAPTER 5

Architecture and Development of Expert System -

“Meditrace”

This chapter includes the overall architecture of a developed system. The first section

discusses the block diagram of the developed system and discusses individual blocks in

detail. The next section includes the ontology development process with the help of

protégé. That section also includes each of the steps required for developing an ontology.

The process and of linking the developed ontology in java is also discussed. The system is

developed in a web-based environment with Java, the later section includes activity and

use-class diagram in order to understand the flow of event and nature of responsibilities of

individual users. The developed system is registered with a domain name called

“Meditrace”. The next section includes different pages or screens incorporated in designing

the “Meditrace” along with details to be filled by the users. The last section includes

discussion about diseases included in differential diagnosis with an introduction about the

treatment guidelines considered in this system.

5.1 Overall System Architecture

The architecture of this developed system is based on the client-server technology

following MVC architecture. As per Fig.5.1, the architecture has different sections: the

client layer, the application logic level (JSP), the semantic web framework (JENA) and

database layer. The client layer initiates a request by accessing the front end designed for a

web browser through JSP technology. This request is being served by server running on a

machine where the controller is designed to serve the desired task using servlets. Ontology

is created by protégé, an open source ontology editor developed by Stanford University,

for storing the knowledge base. OWL file is stored in either local directory or it can be

available online for assessing the ontology from anywhere across the web through JENA

Overall System Architecture

65

API for inferring the effective information. JENA is a Java framework for building the

semantic web application. It provides a programming environment for RDF, RDFS, and

OWL. Additional information required to support the application is stored in MYSQL

database relating to user management, treatment guidelines, and disease information. The

details of an individual block is discussed hereafter.

FIGURE 5.1: Architectural framework of the developed system

JSP Technology and JAVA Servlets

JSP technology is used to develop the platform-independent web-based application; thus

enables the developers to cope up with the fast-growing web technologies. The main

feature of this JSP technology is, it splits the business logic from the designing part of the

user interface. As a result, it allows the developer to modify the user side page design

without modifying the business logic. Specific tags are used to entrap the logic of page

contents in JSP files. While the Java Beans components are used on the server side to hold


66

the application business logic, it can be accessed by JSP page tags(“JavaServer Pages

Technology,” n.d.).

Java Servlets are the elements residing on the server side. They are platform independent

components which are used to extend the Web server capabilities with minimal

maintenance and operating cost. By combining these two powerful technologies, i.e. the

JSP technology and Java Servlets, it is possible to develop platform-independent enterprise

applications with the enhanced performance and detached user side page design with

application business logic (“Java Servlet Technology,” n.d.).The basic architecture of JSP

and Java Servlets technologies are depicted in Fig.5.2.

FIGURE 5.2: Simplified architecture of JSP and Java servlet technology

Semantic Web Framework: Apache Jena

Apache Jena (Jena) is a java based framework offered for developing semantic web-based

applications. It is an open source framework consists of various APIs acting together to

handle RDF data. It provides a developing environment forRDF, RDFS,

and OWL, SPARQL and offers a rule-based inference engine. Jena is a Java API which

can be used to make and manipulate RDF graphs. Jena utilizes the object classes to

represent graphs, resources, properties, and literals. In Jena, a model is a term used for a

graph, while the model interface is used to represent a graph.

Overall System Architecture

67

Additionally, Jena provides several reasoners such as the RDFS reasoner, OWL reasoner,

Transitive reasoner and provides the general-purpose rule-based reasoner that performs

reasoning on both RDFS and OWL knowledge base and is available for general

use(Ameen, Khan, & Rani, 2014; “Apache Jena -,” n.d.; Carroll et al., 2004).

As discussed in chapter 4, there exists various programming language for the

representation of ontology information over the semantic web.The collection includes the

most significant and meaningful language of the OWL Full and even up to the weakest,

RDFS. With the help of ontology API, Jena offers a reliable and uniform development

environment for creating ontology application irrespective of the ontology language(Yu,

2011).

Database Server:

This is one of the important elements of the given system. It stores the information related

to a patient, stores the registration details of EM-staff and doctor, stores login credential of

different users and also stores the treatment chart. The admin user has control and access

to various functionalities of the system. This database also contains patient record and

treatment profile which can be shared with other organization for further analysis purpose.

MySQL database server is used to serve this purpose.

Knowledge Base:

The knowledgebase is a central component of the developed expert system. It is the actual

database which has all information and data related to risk level stratification and

differential diagnosis. It contains diseases (i.e. various cardiac and respiratory emergency),

patient, vital sign, visually observable symptoms, allergies, and other patients related

information. In the developed system, the decisions are asserted based on ontology level

mapping based on predefined rules already in the owl file. This owl file may be available

in various formats such as turtle, RDF/XML, OWL/XML and so on.


68

User Interface

It serves as a communication platform between EM-staff (Paramedic) and EMES

application. Paramedic staff will interact with the giver system through this

interfacedesigned for user’s perspective.EM staff should be given different interfaces

depending on the various tasks that need to be performed. Actually, EM staff will utilize

this interface to enter patient data and other patient-related information and will get the

result as an output through this interface.

5.2 Ontology Development Phase

In this developed expert system, as per concepts discussed in chapter 3, it incorporates

NEWS based scoring system for risk level stratification and various primary assessment

tools for differential diagnosis purpose in emergency health care service. As explained in

chapter 4, Protégé is used to develop the ontology for emergency medicine purpose.

Ontology is developed by the knowledge extracted from the knowledge providers in the

emergency medicine sector.

This ontology is developed using a tool developed by Stanford University called protégé.

Protégé allows the user to create an ontology based on knowledge base gathered by the

knowledge system developer from the experts and from other sources. The step by step

process for achieving this task is listed below with relevant screenshots.

5.2.1 Define class and class hierarchy

This is the first and most important step in developing ontology. Initially, a patient class is

created with the relevant characteristics of the patient as its subclass. The risk level

ontology includes another class as a vital sign and under that few subclasses included i.e.

body temperature, heart rate, respiration rate, etc. Fig. 5.3 shows the overview of classes

created in an ontology. Fig. 5.4 shows the part of some the concepts included in protégé

with the help of class editor available in the editor.

Ontology Development Phase

69

FIGURE 5.3: An overview of the classes of the ontology


70

FIGURE 5.4: Part of some concepts in a developed ontology

5.2.2 Define Object and Datatype property

Property is used to define a binary relationship between individuals or between individuals

to the XML data type. OWL has two main types of properties: First is object property,

which gives a relationship between individuals from two classes. Second is datatype

property, which is used to assign the relationship between an individual to a data value.


71

This developedontology contains several data and object properties. Fig. 5.5 shows some

of the properties included in the developed ontology.

FIGURE 5.5: Object and Datatype property

Object properties are shown in Fig on the left side. These properties are used to relate to

individuals. For instance, Patient is related to PatientID through object property called

“hasID”. Every patient has a specific vital sign indication related through object property

“hasVitalSign”.

The data type properties shown in Fig are used to assign a specific data type to individuals.

While the patient has PatientID which has ID in integer format defined through data type

property called “hasDID”. Patient class has a specific gender in String form related through

data property called “hasGender”.


72

5.2.3 Defining facet, range, and domain of the property

Facets are characteristic of property which is used to put a certain restriction on property

value and value type of property. In addition to this, the property also has range and

domain. For the agiven case, Object property hasPulse1 has a domain of Perfusion_Status

and range extended to a class Pulse. SimilarlyhasVitalSign has a domain Patient and range

as a class VitalSign as shown in Fig. 5.6

FIGURE 5.6: Facet, domain, and range of properties

Table 5.1 shows different object properties used in the ontology and their specified domain

and range. While Table 5.2 shows datatype properties used in the ontology along with their

domain and range with specifying the data type.

TABLE 5.1: List of object properties with their domain and range

Sr No Property Name Domain Range

1 hasAppearance1 Repiratory_Status Gen_Appear

2 hasBP1 Perfusion_Status BP

3 hasBreath1 Respiratory_Status Breath_Sound

4 hasConscious1 Perfusion_Status Conscious_State


73

5 hasCrackels1 Respiratory_Status Crackles

6 hasPerfusion Diagnosis Perfusion_Status

7 hasPulse1 Perfusion_Status Pulse

8 hasRbreathing Respiratory_Status Workof_Breathing

9 hasRconscious Respiratory_Status ConsciousState

10 hasRespiration1 Respiratory_Status Respiration_Rate

11 hasRespiratory Diagnosis Respiratory_Status

12 hasRHR Respiratory_Status Heart_Rate

13 hasRrhythm Respiratory_Status Respiratory_Rhythm

14 hasRskin Respiratory_Status Skin_R

15 hasSkin1 Perfusion_Status Skin

16 hasSpeech1 Respiratory_Status Speech

17 hasTotalScoreN Diagnosis Risk_Level_Score

18 hasVitalSign Patient VitalSign

19 hasWheez1 Respiratory_Status Wheeze

20 hasID Patient

TABLE 5.2: List of Data type properties with their domain and range

Sr NoProperty Name Domain Range

1 has Age Patient Integer

2 hasBP BP String

3 hasConscious ConsciousState String

4 hasDecription Risk_Level_Score String

5 hasDID Patient Integer

6 hasEye Eye String

7 hasGender Patient String

8 hasHR HeartRate String

9 hasLOC LevelofConsciousness String


74

10 hasLocation Patient String

11 hasMotor Motor String

12 hasO2S OxygenSaturation String

13 hasO2Suppl SupplementalOxygen String

14 hasOption Respiratory_Assessment String

15 hasPulse Pulse String

16 hasRR Respiratoin_Rate String

17 hasSBP SystolicBloodPressure String

18 hasScore Risk_Level_Score String

19 hasSkin Skin_R String

20 hasTemp BodyTemperature String

21 hasTotalGCSScore GCS_Assessment String

22 hasTotalScore1 VitalSign String

23 hasVerbal Verbal String

5.2.4 Creating Individuals

Individuals are the concrete object of classes. These instances were created in protégé tool.

These individuals are used to form a rule based on provided range values/symptoms or

disease to the user. Fig. 5.7 shows an example of individuals created in an ontology.

Ontology Model Creation

75

FIGURE 5.7: Example of instances in ontology

5.3 Ontology Model Creation

5.3.1 Loading Ontology file

In order to load the ontology file, we have created one class. That class contains one

method responsible for reading the model. Using ModelFactory class, we have created an

empty ontology model (model uses OWL full profile, in memory storage) which holds all

asserted statement in our ontology file. The Figure shows the java code for loading and

reading the ontology file and creating a model. The system was developed using one of the

most popular IDE known as NetBeans(“Welcome to NetBeans,” n.d.).


76

FIGURE 5.8: JAVA code for loading and saving an ontology model

5.3.2 SPARQL query to retrieve the information from the ontology

SPARQL is used to search and query triples stored in the inferred ontology model. There

are different queries included in the system to retrieve different responses from the inferred

model. Fig shows the query used to get the total score of the individual patient based on the

values of the different physiological parameter and their individual score(Alves, Damásio,

& Correia, 2015; Magesh & Thangaraj, 2011; “SPARQL Query Language for RDF,” n.d.).

FIGURE 5.9: Example of SPARQL query to retrieve risk level of patient


77

5.3.3 MySQL Database server

A large amount of database in any web application requires the assistance from MySQL

Server which includes transactional data dictionary for storing databaseas shown in Table

5.3. These database objects are protected and can be accessed only in debug builds of

MySQL. Nevertheless, it is possible to access the stored data from the data tables using

INFORMATION_SCHEMA tables and SHOW statements. MySQL data dictionary offers

several advantages: Uniformity in stored dictionary data due to simplicity of a centralized

data dictionary schema, Removal of file-based metadata storage, Transactional crash-safe

storage of dictionary data, Uniform and centralized caching for dictionary objects, a

simpler and improved implementation for some INFORMATION_SCHEMA tables

(“MySQL,” n.d.).

TABLE 5.3: Data dictionary in the database

Table Name Field Name1 Login -Loginid

-Username-Password

2 Registration -Regid (FK login)-Fields as per EM registration form-Additional specialization id

3 Doc_spl_mst -Role_id (FK login)-Spl_id-Spl_name

4 Role_mst -Role_id-Role_name

5 Drug_mst -Drug_id-Drug_name-Drug_property (as per sheet)

6 Disease_mst -Disease_id-Disease_name-Disease_desc-Spl_id (FK Spl_id)

7 Patient_mst -Patient_id-Patient additional information (as per form)

8 Risk_level_para_Trn -Trn_id-P_id-Risk level parameter (as per form)-Score (Updated from ontology)

9 Treatment_action_mst -Trn_id-Disease_id (FK disease_id)-Treatment_title-Treatment_action-Sub_level_id


78

5.3.4 Use case for Developed Expert system

Use case diagram is used to model high-level functions, the scope of the system and show

a relationship with its intended users referred as actors so that users can understand the

features of the system before implementation and serves as a foundation for other diagrams

such as activity diagram shown in Fig. 5.10 (“Use Case Diagrams - Use Case Diagrams

Online, Examples, and Tools,” n.d.).

FIGURE 5.10: Use case diagram of developed system


79

FIGURE 5.11: Activity Diagram

Start

Login for EM staff

Enter Basic Information of Patient

Enter Patients Vital Parameters(Respiration Rate, Heart Rate etc..)

Displays Risk Level ofPatient

Going forDifferentialDiagnosis

Transport tonearest Hospital

Immediately

Do Further Assessmentof Patient

Enter additional visually observablesigns/symptoms of Patient

Result Display1. Perfusion Status Assessment

2. Respiratory status Assessment3. GCS Score

4. Probable list of Disease

Going forTreatmentsuggestion

Displays relevant treatmentsteps as per guidelines

Is suggested TreatmentApplied? (Y/N)

END

System activity

EM-staff activity

Start/End


80

5.3.5 Description of Activity diagram

Activity

No

Description

1 First EM-staff needs to do registration on the form prescribed. After receiving

login credential, EM-staff needs to login in order to access the functionality of

the system. The system checks the login and password which user has already

created and stored in database of the system.

2 After login, Paramedic has to enter patients basic information like his name,

age, gender, etc. and the system saves this information in an internal database

and assigns a unique id to each patient.

3 After entering the basic information of the patient, EM-staff has to enter the

most important and frequently monitored physiological parameters. This

information will be saved again in the database.

4 The values of selected vital sign values will be parsed to a knowledge base

stored on the server side. Depending on the values, individual scores are

retrieved and calculated. The calculated aggregate score will be displayed on the

user screen along with risk level.

5 In this step, Based on the risk level, EM-staff can either go for differential

diagnosis or transfer the patient immediately to the nearest hospital. This

decision will be taken by EM-staff considering individual risk and patient’s

onsite condition.

6 Then EM-staff needs to do further assessment of a patient in order to judge the

existing condition of the patient and to prevent further deterioration.

7 In order to perform a differential diagnosis of a patient, Paramedic has to enter

additional information based on visually observable signs and symptoms of the

patient.

8 Based on the selection of additional symptoms, The system again parses this

information to the knowledge base stored in an ontology. This additional

information is used for the perfusion, respiratory and GCS score assessment.

This assessment results along with some other information and predefined rules

in knowledge used to display the results of the diagnosis.

Implementation of Ontology based Expert System - Meditrace

81

9 The next step is to decide whether paramedic wants to go for a treatment

suggestion or decides to transfer a patient to the nearest hospital immediately

depending on the patient’s existing condition

10 After opting for treatment, User has to select the particular diagnosed disease.

After clicking, the system retrieves steps of treatment guidelines stored in the

database server. This guideline is an interactive type.

11 The system also logs whether paramedic has applied the treatment or not.

5.4 Implementation of Ontology based Expert system for Emergency

Medicine – Meditrace

5.4.1 Emergency Staff registration page:

As shown in Fig. 5.12, this page is used to enter all the basic information on emergency

staff. This screen generates a login for that paramedic. The system also demands the name

of the hospital and unique ID number provided by the hospital. This stores email id of the

patient as login and default password will be given to that person.

FIGURE 5.12: Emergency Staff registration page


82

5.4.2 Login page

This page is designed to access the system through already assigned or generated login

credentials as shown in Fig.5.13. With the help of this screen, EM-staff, as well as Admin,

can access the system and its functionalities.

FIGURE 5.13: Login screen

5.4.3 Emergency Risk Level Assessment

This screen is designed to take appropriate input from paramedic staff. As shown in Fig.

5.14, this page includes two sections. The first section is designed to take basic information

of the patient including name, age, and gender. The second section demands information

about the values of physiological parameters.


83

FIGURE 5.14: Emergency assessment screen

5.4.4 Display page of Risk level score

This screen is used to display the Risk assessment score and its level as shown in Fig.5.15.

This page will fetch the result from the knowledge base based on the inputs provided

earlier by the paramedic staff. This screen has the selection on the bottom section, where

EM-staff can decide whether to go for differential diagnosis or to transport the patient to

the nearest hospital.

FIGURE 5.15: Risk assessment score page


84

5.4.5 Differential Diagnosis

This screen is used to get the additional input from the paramedic staff. This page is

divided into three main sections. As per Fig. 5.16, the first section is for perfusion

assessment, which takes a few visually observable inputs along with values of the

physiological parameter. While the second section includes respiratory assessment, it asks

additional visually observable parameters along with the value of few vital sign

parameters. The third section demands additional information for GCS score calculation

and consciousness assessment.

FIGURE 5.16: Differential diagnosis assessment screen


85

5.4.6 Primary Assessment Result screen

This is the display screen designed to display primary assessment result and list of

probable diseases. As shown in Fig.5.17, this page includes the status of perfusion,

respiratory and GCS score. This page also displays suggested disease and seeks the

paramedic input whether to go for treatment of transport the patient to the nearest hospital.

FIGURE 5.17: Assessment result screen

5.4.7 Treatment

This page is designed to fetch the treatment chart for the particular disease. This process

follows a sequence of title and action as shown in Fig. 5.18. The treatment may include

single or more than one step as shown in Fig. 5.19 & 5.20. Depending upon the response of

the patient to a given treatment, the paramedic needs to select an appropriate option.

Accordingly, the treatment chart displays the next step of treatment. As shown in Fig. 5.21,

the last page of the treatment asks the paramedic for selecting the option about whether the

treatment has been applied as per the suggestion or not. On every screen, there is an option

for transporting the patient to hospital by leaving further steps of treatment.


86

FIGURE 5.18: Treatment screen step-1




87

FIGURE 5.21: Treatment screen last step

5.4.8. Emergency Assessment Patient Report

This page is only available to Admin login. As shown in Fig. 5.22, this page shows the

report of all patients assessed with the developed system. As per Fig. 5.23, by clicking on

view report admin can see the details of all parameters fed earlier during the assessment

cycle. This report also stores results of all successive assessments of a patient under single

unique patient ID. This helps to see the effectiveness of the treatment and trends the

patient’s condition with time.

FIGURE 5.22: Admin screen for patient assessment report


88

FIGURE 5.23: Screen showing an assessment of one patient

5.5 Emergencies included in developed expert system

For this expert system, amongst many of health emergency reported at the emergency

medicine department, we have included two of the most frequently encountered

emergencies. These emergencies are included as a part of the primary phase

implementation of this system(Darlene Ellchuk, 2005; Wardrope& MacKenzie, 2004;

Woollard & Greaves, 2004). They are discussed here:

5.5.1 Cardiac Emergency:

Inadequate perfusion: Perfusion defined as the ability of the cardiovascular

system to supply enough oxygenated blood to the tissues and to collect unwanted

metabolic waste products. Inadequate or poor perfusion results in creating

disturbance in normal human activity.

Emergencies included in the developed system

89

Cardiac Arrest: The condition, in which cardiac activity ceases completely, is

known as cardiac arrest. In this condition,the patient loses cardiac function,

breathing, and consciousness. This is typically due to an electrical disturbance in

the patient’s heart which interrupts the normal pumping action of the heart,

ultimately stops the blood supply to the body.

Bradycardia:Bradycardia is a condition in which heart rate becomes slower than

normal. The normal heart rate of an adult ranges from 60 to 100 beats per minute in

resting state. The situation when the heart beats at a rate lesser than 60 bpm is

called Bradycardia. It creates a severe problem if the heart doesn't supply enough

oxygen-rich blood to the body.

Tachycardia (Narrow and Broad complex): Tachycardia is a condition in which

the heart starts to beat at a rate faster than 100 bpm.

Acute coronary syndrome: This term is used to cover a set of conditions under

which the blood flow to the heart reduces, suddenly. A heart attack is one such

condition in which a cell dies due to damaged heart tissues. Sometimes, the acute

coronary syndrome does not cause cell death, but it affects the normal functionality

of heart due to reduced blood flow. This condition may indicate the possible risk of

a heart attack. Another symptom of this condition includes severe chest pain or

discomfort. This condition comes under an emergency situation which needs

immediate diagnosis and appropriate treatment.This treatment is provided to

facilitate the blood flow, treating complication and avoiding potential future risks.

5.5.2 Respiratory Emergency:

Asthma (Conscious / Unconscious / No cardiac output): The constriction and

swelling of airways, which results in the production of extra mucus. This condition

is called Asthma. This condition makes difficult breathing and triggers coughing,

wheezing and shortness of breath. Asthma can be a major problem that interferes

with daily activities and may lead to a life-threatening asthma attack.

Chronic Obstructive Pulmonary Syndrome (COPD): COPD is a chronic

inflammatory lung disease which results in the obstruction in the airflow from the

lungs. The most common symptoms of COPD include difficulty in breathing,

cough, extra production of mucus and wheezing. This condition may arise due to

exposure to some annoying gases, mostly due to cigarette smoking. Individuals


90

suffering from COPD are at greater risk of developing cardiac problems, lung

cancer,and various other conditions.

Pulmonary edema: The condition which accumulates excessive fluid in the lungs

is called pulmonary edema. As the lungs contain various air sacs, this fluid

accumulated in these sacs, makes it difficult to breathe. The main reason for

pulmonary edema is heart-related issues. In addition to this, it may also occur due

to pneumonia, rapid contact with certain toxic substances and medications, chest

wall trauma, and staying at high altitude. Acute pulmonary edema requires

emergency medical care as if it can be fatal sometimes if not treated on time.

5.5.3 Treatment Guidelines

The emergency medicine treatment guideline is used to provide patient management in

critical condition to treat the patient. These guidelines are provided by the experts who deal

with these emergencies daily. These guidelines are reviewed by the specialists from time to

time and also accepted as a standard. It also lists out the primary assessment tools and

other patient management strategy. Sometimes, separate guidelines are provided for EMT-

Basic, EMT-Paramedic and EMT-Advance(Laura Hand &Yasmini Mawji, 2012; Pre-

hospital emergency care council, 2017a; Pre-hospital emergency care council, 2017b;

Victoria, 2018). Fig.5.34 indicates the treatment chart for one of the respiratory emergency

i.e. COPD.

Emergencies included in the developed system

91

FIGURE 5.24: Treatment steps for COPD

5.6 Summary

This chapter includes an architectural framework of the developed system. The main

blocks of this system are discussed in the first section of this chapter. After that, the next

section has included the system development steps. The web-based java technology is used

to develop this system called “Meditrace”. The last section has included all the

pages/screens included in this system along with their importance and screenshot. The

differential diagnosis has included two major emergencies and treatment chart followed by

paramedics as a part of prehospital emergency care. That is the last section of this chapter.

Results & Validation

92

CHAPTER 6


This expert system is developed for facilitating the paramedic staff in order to assist them

for effective emergency health care delivery. This system takes the inputs from the

paramedic in terms of few frequently monitored vital sign parameters and some visually

observable parameters. This chapter includes the testing dataset selected and varieties

incorporated for the validation purpose of this system. The second section consists of the

validation results of the system and its effectiveness in the diagnosis of different

emergencies.

6.1 Testing Dataset

The proposed Meditrace system tested and evaluated on localhost using apache tomcat

server. This system is also available online www.meditrace.in. As for finding the

effectiveness of his system, it needs to be tested with the actual patient database. Shree

GirirajMultispecialty Hospital is one of the well-known hospitals located in Rajkot city.

This hospital has specialization in critical care with a team of energetic and experienced

medical professional. The developed system was tested with the database provided by this

hospital. For effective testing of the proposed system, it would be a critical process of

selecting suitable patient dataset. The data collection method is specifically concentrated

inthe Emergency Department.

The database is selected considering all possibilities and variations. This includes gender,

age variation (Adults) and patients with past history. In order to check the efficacy of the

system and to check its validity in a clinical environment, it is necessary to cover all

possibilities arises in emergency health care sector. As shown in Fig.6.1, it shows the

patient database selected with variation in gender. It shows the selected database includes

67 male patients while 31 female patients. While Fig. 6.2 shows the variation of age group

Testing Dataset

93

consider for thevalidation purpose of this system. It is evident from the figure that the

majority of the patient considered for this case are above 40 years. In Fig. 6.3, it indicates

that around 55 patients are having a past history. These past history data include:

Hypertension, Diabetes, Major surgeries or Ischemic heart disease.

FIGURE 6.1: Patient database with Gender variation

FIGURE 6.2: Patient database with Age variation


94

FIGURE 6.3: Patient database with variation in Past history

FIGURE 6.4: Patients NEWS score and its range variation

Validation Results

95

6.2 Validation Results

As per the chart is shown in Fig 6.4, this patient database is given to Meditrace and their

scores are stored in a database. The analysis of that database shows that more than 50

patients are having low score and rest of the 50 patients having scored in either medium or

in a higher range. In high-risk patient, majority cases required immediate intervention by

the paramedic staff. Majority of the patient died during their in-hospital treatment, The

NEWS score of those patients are more than 10. This clearly indicates that the NEWS

score of individual patient states the potential risk to an individual’s life.

Fig. 6.5 indicates patient percentage with a cardiac emergency. Patient database of around

27% includes a score of more than 7, while 27% data includes patients with a medium

score, while the rest of 45% of patients are having a low score. This wide variety of

scoring and having various diseases serves as valuable data for the validation of Meditrace

system. While Fig.6.6 revels percentage patient with a respiratory emergency. This chart

shows patients with more than 7 scores are in the higher side of the population. While

patients have scored in the medium and lower range are sharing 30 percent of the total

population. With this various possible input database, the Meditrace system was tested and

validated. The Risk level stratification performed by the Meditrace based on the calculated

total score is proven to be very effective and accurate for time-critical cases.

FIGURE 6.5: Patient database with cardiac emergency variation


96

FIGURE 6.6: Patient database with respiratory emergency

As depicted from Fig.6.7, when this database is utilized for validating the system, its

disease prediction probability shows a success rate of around 75%. This means the

developed system is capable of forecasting the probable disease with an accuracy of

around 75%. This clearly indicates its effectiveness and usefulness of the system in the

emergency health care scenario for assisting paramedics. While as per Fig.6.8, It shows the

success rate of individual diseases as per the patient database used for the validation

purpose. For some of the diseases, the system has proven to be 100% accurate with precise

disease forecasting proficiency. While for the rest of the diseases the accuracy of the

system varies from 20% to 80%.

Validation Results

97

FIGURE 6.7 Disease prediction probability

FIGURE 6.8: Success rate for different diseases

Conclusion & Future Scope

98

CHAPTER 7


Urgent and immediate medical intervention is essential in emergency health care sector to

stabilize the patient and to prevent further health deterioration. Meditrace is an approach of

web-based expert emergency medicine decision support system developed for facilitating

paramedic in the emergency health care system. The first section of this chapter includes

the concluding remarks of the research work and summarizes the contribution. The second

section includes the possible future scope of the presented research work, including

probabilities of different approaches needed to extend the research beyond this point.

7.1 Conclusion

The Meditrace system is developed for assisting the paramedic staff in case of emergency

health care sector in the absence of an expert medical professional. This web-based system

is developed for risk level stratification, differential diagnosis and treatment procedure.

The risk level stratification helps to prioritize the course of treatment and works as an

important triage tool for assigning the degree of urgency. The web-based system assists the

paramedic staff to diagnose the probable disease based on a minimum set of visually

observable symptoms. The system also has the support to show step-by-step guidelines of

treatment for listed diseases. This helps the inexperienced or untrained paramedic staff to

choose the most appropriate path of action. For making the system approachable, the

developed system is a launched over the web platform.

Risk level stratification is performed by well-designed and widely accepted scoring

system. This NEWS scoring system categorizes the risk level of patient based on the

aggregate score calculated from the individual score of most frequently monitored

physiological parameters. NEWS scoring system is designed to serve effectively in

emergency medicine. In the Indian scenario, the effectiveness of this scoring system is

Future Scope

99

reasonably acceptable. The results are also replicating the same phenomenon, indicating

higher the score higher the risk to an individual’s health. On the other hand, differential

diagnosis task is performed by various patient assessment tools. These tools are accessing

the patient’s condition based on some visually observable symptoms along with few

physiological parameters. The results achieved from the patient assessment tools are used

by the rules defined for differential diagnosis purpose. The final outcome of this developed

system is the list of probable disease along with an option to choose the most appropriate

one. Both of these tasks are performed by the ontologies designed and developed by

extracting the knowledge from the domain experts. The system also offers a treatment

suggestion for a specific disease. This facility is incorporated in the developed system by

considering the existing emergency guidelines available and accepted by the EM

community. This makes the system to be helpful while providing training to EM-staff.

The development of an expert system from a semantic web based ontological framework is

one of the key aspects of Meditrace. Ontologies are used in semantic web domain for

getting information in a comprehensive and machine-understandable format. Ontologies

are generic, reusable and can be shared between people. This architecture enables the

system to build its knowledge base using available ontology either stored locally or over

web. This architecture allows the domain expert to maintain the knowledge base in the

ontology without making any changes in the overall system. This is the modular approach

accepted for developing the expert system which makes the system scalable.

7.2 Future Scope

At present, Meditrace supports the differential diagnosis of a few cardiovascular diseases

and respiratory diseases. This can be extended for other diseases from the same categories

as well. In addition to this, the existing system can also be extended for other emergency

conditions as well. The existing system takes values of body vitals from the client screen

designed for getting the inputs from paramedic manually. It can be possible to acquire

these body vital parameters using integrated sensors connected to the patient’s body. This

can make the decision making faster and reduces the additional burden of manual data

entry of paramedic. It also helps to assess the patient’s condition continuously, facilitates

uninterrupted monitoring. It is also possible to incorporate uasage of the current


100

IoT(Internet of Things) technology to communicate with body sensors and other devices

effectively without external intervention.

Currently, Meditrace uses a dedicated ontology developed for risk level stratification and

disease forecasting. It is possible to integrate other already developed and tested ontology

with the existing system. This helps to enrich the knowledge base of the existing system

and helps to incorporate other necessary information. The existing system requires

knowledge upgradation by the domain experts by updating its ontology through some

dedicated user interface. This is necessary for making the system scalable and upgradable.

The existing system requires paramedic to feed the patient data manually and assigns

system generated a unique ID. Instead of this, it would be desirable if the patient data can

be taken from some unique ID already assigned by the government (ex. AADHAR No)

and this can be linked to patient EHR (Electronic Health Record). This helps to fetch the

patient’s basic information, past medical history, and allergies, which helps in saving the

time of entering patient information, refining diagnostic algorithm, and deciding the course

of treatment. By incorporating most of the above suggestion, it is possible to develop full-

scale decision support system for all emergencies arises in emergency health care delivery

system.

List of References

101

List of References

Abo-Zahhad, M., Ahmed, S. M., & Elnahas, O. (2014). A wireless emergency

telemedicine system for patients monitoring and diagnosis. International Journal of

Telemedicine and Applications, 2014, 4.

Agri Daksh. (n.d.). Retrieved June 11, 2019, from http://agridaksh.iasri.res.in/home.jsp

Ahmed, A., Ishaque, A., & Nawaz, T. (2014). Information and communication technology

introducing efficiency in Emergency Medical Services. 2014 IEEE International

Conference on Management of Innovation and Technology, 211–215. IEEE.

Alves, M. B., Damásio, C. V., & Correia, N. (2015). SPARQL commands in Jena rules.

International Conference on Knowledge Engineering and the Semantic Web, 253–262.

Springer.

Amanpreet Kaur. (2017). Case studies of expert system & research aspects in artificial

intelligence. International Conference of Recent Innovations in Engineering, Science,

Humanities and Management, 778–783. Punjab, India.

Ameen, A., Khan, K. U. R., & Rani, B. P. (2014). Extracting knowledge from ontology

using Jena for semantic web. International Conference for Convergence for Technology-

2014, 1–5. IEEE.

Apache Jena -. (n.d.). Retrieved June 11, 2019, from http://jena.apache.org/

Ariani, A., Koesoema, A. P., & Soegijoko, S. (2017). Innovative healthcare applications of

ICT for developing countries. In Innovative Healthcare Systems for the 21st Century (pp.

15–70). Springer.

Badr, N. G. (2016). Could ICT Be Harnessed for Prehospital Emergency Medical

Services? Proceedings of the International Joint Conference on Biomedical Engineering

Systems and Technologies, 269–276. SCITEPRESS-Science and Technology Publications,

Lda.

Barnett, G. O., Cimino, J. J., Hupp, J. A., & Hoffer, E. P. (1987). DXplain: an evolving

diagnostic decision-support system. Jama, 258(1), 67–74.

Berner, E. S. (2007). Clinical decision support systems (Vol. 233). Springer.

List of References

102

Berners-Lee, T., Hendler, J., Lassila, O., & others. (2001). The semantic web. Scientific

American, 284(5), 28–37.

Bhat, R. (1996). Regulation of the private health sector in India. The International Journal

of Health Planning and Management, 11(3), 253–274.

Calero, C., Ruiz, F., & Piattini, M. (2006). Ontologies for software engineering and

software technology. Springer Science & Business Media.

Cardoso, J. (2007). The semantic web vision: Where are we? IEEE Intelligent Systems,

22(5), 84–88.

Carroll, J. J., Dickinson, I., Dollin, C., Reynolds, D., Seaborne, A., & Wilkinson, K.

(2004). Jena: implementing the semantic web recommendations. Proceedings of the 13th

International World Wide Web Conference on Alternate Track Papers & Posters, 74–83.

ACM.

Chandwani, R. K., & Dwivedi, Y. K. (2015). Telemedicine in India: current state,

challenges and opportunities. Transforming Government: People, Process and Policy,

9(4), 393–400. https://doi.org/10.1108/TG-07-2015-0029

Corcho, O., Fernández-López, M., & Gómez-Pérez, A. (2003). Methodologies, tools and

languages for building ontologies. Where is their meeting point? Data & Knowledge

Engineering, 46(1), 41–64.

Corcho, O., & Gómez-Pérez, A. (2000). A roadmap to ontology specification languages.

International Conference on Knowledge Engineering and Knowledge Management, 80–96.

Springer.

Crowley, R. S., & Medvedeva, O. (2006). An intelligent tutoring system for visual

classification problem solving. Artificial Intelligence in Medicine, 36(1), 85–117.

DAML Tools. (n.d.). Retrieved June 11, 2019, from http://www.daml.org/tools/

Darlene Ellchuk. (2005). Patient Assessment - Self Assessment Tool (p. 75). Alberta,

Canada: College of Licensed Practical Nurses of BC.

Differential Ontology Editor. (n.d.). Retrieved June 11, 2019, from

http://www.eurecom.fr/~troncy/DOE/

List of References

103

e-Health & Telemedicine | Ministry of Health and Family Welfare | GOI. (n.d.). Retrieved

June 11, 2019, from https://mohfw.gov.in/about-us/departments/departments-health-and-

family-welfare/e-Health%20%26%20Telemedicine

Elkin, P. L., Liebow, M., Bauer, B. A., Chaliki, S., Wahner-Roedler, D., Bundrick, J., …

others. (2010). The introduction of a diagnostic decision support system (DXplainTM) into

the workflow of a teaching hospital service can decrease the cost of service for

diagnostically challenging Diagnostic Related Groups (DRGs). International Journal of

Medical Informatics, 79(11), 772–777.

Gómez-Pérez, A., & Corcho, O. (2002). Ontology languages for the semantic web. IEEE

Intelligent Systems, 17(1), 54–60.

Gomez-Perez, A., Fernández-López, M., & Corcho, O. (2006). Ontological Engineering:

with examples from the areas of Knowledge Management, e-Commerce and the Semantic

Web. Springer Science & Business Media.

Gómez-Pérez, J. M., & Ruiz, C. (2010). Ontological Engineering and the Semantic Web.

In J. D. Velásquez & L. C. Jain (Eds.), Advanced Techniques in Web Intelligence - I (pp.

191–224). https://doi.org/10.1007/978-3-642-14461-5_8

Graber, M. L., & Mathew, A. (2008). Performance of a web-based clinical diagnosis

support system for internists. Journal of General Internal Medicine, 23(1), 37–40.

Gruber, T. R. (1993). A Translation Approach to Portable Ontology Specifications. Knowl.

Acquis., 5(2), 199–220. https://doi.org/10.1006/knac.1993.1008

Guarino, N. (1998). Formal Ontology in Information Systems: Proceedings of the 1st

International Conference June 6-8, 1998, Trento, Italy (1st ed.). Amsterdam, The

Netherlands, The Netherlands: IOS Press.

Guarino, Nicola. (1998). Formal ontology in information systems: Proceedings of the first

international conference (FOIS’98), June 6-8, Trento, Italy (Vol. 46). IOS press.

Gupta, P., Pol, S., Rahatekar, D., & Patil, A. (2016). Smart ambulance system.

International Journal of Computer Applications, 6, 23–26.

GVK Emergency Management and Research Institute. (2016). EM CARE towards saving

lives and bringing hope (p. 36). Telangana, INDIA.

List of References

104

Health Information and Quality Authority. (2014). Review of pre-hospital emergency care

services to ensure high quality in the assessment, diagnosis, clinical management and

transporting of acutely ill patients to appropriate healthcare facilities. (p. 192). Retrieved

from https://www.hiqa.ie

Heart of England NHS foundation Trust. (2012). Adult Modified Early Warning Score

(MEWS) (Policy & Escalatation Pathway No. Version 3; p. 18). England.

Henriquez-Camacho, C., Losa, J., Miranda, J. J., & Cheyne, N. E. (2014). Addressing

healthy aging populations in developing countries: unlocking the opportunity of eHealth

and mHealth. Emerging Themes in Epidemiology, 11(1), 136.

Hofman, K., Primack, A., Keusch, G., & Hrynkow, S. (2005). Addressing the growing

burden of trauma and injury in low-and middle-income countries. American Journal of

Public Health, 95(1), 13–17.

Hsu, C.-W., Wann, S.-R., Chiang, H.-T., Lin, C.-H., Kung, M.-H., Lin, S.-L., & others.

(2001). Comparison of the APACHE II and APACHE III scoring systems in patients with

respiratory failure in a medical intensive care unit. Journal of the Formosan Medical

Association, 100(7), 437–442.

Hudli, A. V., Palakal, M. J., & Zoran, M. J. (1991). A neural network based expert system

model. [Proceedings] Third International Conference on Tools for Artificial Intelligence-

TAI 91, 145–149. IEEE.

Java Servlet Technology. (n.d.). Retrieved June 11, 2019, from

https://www.oracle.com/technetwork/java/index-jsp-135475.html

JavaServer Pages Technology. (n.d.). Retrieved June 11, 2019, from

https://www.oracle.com/technetwork/java/index-jsp-138231.html

Joshipura, M., Hyder, A. A., & Rehmani, R. (2004). Emergency care in South Asia:

challenges and opportunities. Journal of the College of Physicians and Surgeons–Pakistan:

JCPSP, 14(12), 731–735.

Karlsten, R., & Sjoqvist, B. (2000). Telemedicine and decision support in emergency

ambulances in Uppsala. Journal of Telemedicine and Telecare, 6(1), 1–7.

List of References

105

Karoui, L., Aufaure, M.-A., & Bennacer, N. (2004). Ontology discovery from web pages:

Application to tourism. In the Workshop of Knowledge Discovery and Ontologies.

Citeseer.

Khosrow-Pour, M. (2006). Emerging trends and challenges in information technology

management. Idea Group.

Kingston, J., Shadbolt, N., Tate, A., & others. (1996). CommonKADS models for

knowledge based planning. AAAI/IAAI, Vol. 1, 477–482.

Knaus, W. A., Draper, E. A., Wagner, D. P., & Zimmerman, J. E. (1985). APACHE II: a

severity of disease classification system. Critical Care Medicine, 13(10), 818–829.

Knaus, W. A., Wagner, D. P., Draper, E. A., Zimmerman, J. E., Bergner, M., Bastos, P. G.,

… others. (1991). The APACHE III prognostic system: risk prediction of hospital

mortality for critically III hospitalized adults. Chest, 100(6), 1619–1636.

Koceska, N., Komadina, R., Simjanoska, M., Koteska, B., Strahovnik, A., Jošt, A., …

others. (2019). Mobile wireless monitoring system for prehospital emergency care.

European Journal of Trauma and Emergency Surgery, 1–8.

Kolodner, J. L. (1997). Educational implications of analogy: A view from case-based

reasoning. American Psychologist, 52(1), 57.

Krishnan, G., Aditi, R., & others. (2009). Telemedicine in India: the Apollo story.

Telemedicine and E-Health, 15(6), 576–585.

Kyriacou, E., Pavlopoulos, S., Berler, A., Neophytou, M., Bourka, A., Georgoulas, A., …

others. (2003). Multi-purpose HealthCare Telemedicine Systems with mobile

communication link support. Biomedical Engineering Online, 2(1), 7.

Laura Hand, & Yasmini Mawji (Eds.). (2012). The ABCs of Emergency Medicine

(Twelth). University of Toronto.

Le Onn Ho, H. L., Shahidah, N., Koh, Z. X., Sultana, P., & Ong, M. E. H. (2013). Poor

performance of the modified early warning score for predicting mortality in critically ill

patients presenting to an emergency department. World Journal of Emergency Medicine,

4(4), 273.

List of References

106

Lincoln, M. J., Turner, C. W., Haug, P. J., Warner, H. R., Williamson, J. W., Bouhaddou,

O., … Grant, M. (1991). Iliad training enhances medical students’ diagnostic skills.

Journal of Medical Systems, 15(1), 93–110.

Liu, L., & Zsu, M. T. (2009). Encyclopedia of Database Systems (1st ed.). Springer

Publishing Company, Incorporated.

Maedche, A., & Staab, S. (2001). Ontology learning for the semantic web. IEEE Intelligent

Systems, 16(2), 72–79.

Magesh, N., & Thangaraj, P. (2011). Semantic image retrieval based on ontology and

SPARQL query. International Conference on Advanced Computer Technology (ICACT).

Marchevsky, A., Truong, H., & Tolmachoff, T. (1997). A rule-based expert system for the

automatic classification of DNA “ploidy” histograms measured by the CAS 200 image

analysis system. Cytometry: The Journal of the International Society for Analytical

Cytology, 30(1), 39–46.

Matkar, R., & Parab, A. (2011). Ontology based expert systems–replication of human

learning. In Thinkquest∼ 2010 (pp. 43–47). Springer.

Mellouli, S., Bouslama, F., & Akande, A. (2010). An ontology for representing financial

headline news. Web Semantics: Science, Services and Agents on the World Wide Web, 8(2–

3), 203–208.

Miller, R. A., McNeil, M. A., Challinor, S. M., Masarie Jr, F. E., & Myers, J. D. (1986).

The INTERNIST-1/quick medical REFERENCE project—Status report. Western Journal

of Medicine, 145(6), 816.

Milton, N., Shadbolt, N., Cottam, H., & Hammersley, M. (1999). Towards a knowledge

technology for knowledge management. International Journal of Human-Computer

Studies, 51(3), 615–641.

Ministry of Health and Family Welfare Government of India. (2017). National health

policy 2017 (p. 28) [Government Policy]. India.

Mishra, S. K., Kapoor, L., & Singh, I. P. (2009). Telemedicine in India: current scenario

and the future. Telemedicine and E-Health, 15(6), 568–575.

List of References

107

Mishra, S. K., Singh, I. P., & Chand, R. D. (2012). Current status of telemedicine network

in India and future perspective. Proceedings of the Asia-Pacific Advanced Network, 32(1),

151–163.

MySQL. (n.d.). Retrieved June 11, 2019, from https://www.mysql.com/

National Early Warning Score (NEWS) (p. 47). (2019). London.

nmcn - National Telemedicine Portal. (n.d.). Retrieved June 11, 2019, from http://nmcn.in/

North West Ambulance Service (NHS trust). (2014). Paramedic pathfinder and community

care pathways. Bolton.

Noy, N. F., McGuinness, D. L., & others. (2001). Ontology development 101: A guide to

creating your first ontology. Stanford knowledge systems laboratory technical report KSL-

01-05 and ….

O’callaghan, T. A., Popple, J., & McCreath, E. (2003). SHYSTER-MYCIN: a hybrid legal

expert system. Proceedings of the 9th International Conference on Artificial Intelligence

and Law, 103–104. ACM.

OWL Web Ontology Language Guide. (n.d.). Retrieved June 11, 2019, from

https://www.w3.org/TR/owl-guide/

Patel, J., & Bhatt, C. (2014). A CommonKADS Model Framework for Web Based

Agricultural Decision Support System. International Journal on Food System Dynamics,

5(4), 196–203.

Patel, M., Virparia, P., & Patel, D. (2012). Web based fuzzy expert system and its

applications–a survey. International Journal of Applied Information Systems, 1(7), 11–15.

Pavlopoulos, S., Kyriacou, E., Berler, A., Dembeyiotis, S., & Koutsouris, D. (1998). A

novel emergency telemedicine system based on wireless communication technology-

AMBULANCE. IEEE Transactions on Information Technology in Biomedicine, 2(4), 261–

267.

Pinki_CF, M. (n.d.). Introduction_Telemedicine. Retrieved June 11, 2019, from AIIMS

NEW website: https://www.aiims.edu/en/introduction-telemedicine.html

List of References

108

pOWL - A Web Based Platform for Collaborative Semantic Web Development. (n.d.).

Retrieved June 11, 2019, from http://powl.sourceforge.net/overview.php

Pre-hospital emergency care council. (2014). Clinical Practice guidelines (No. Fifth

Edition). Ireland.

Pre-hospital emergency care council. (2017a). Guidelines Clinical Practice - Emergency

Medical Technician (p. 134). Ireland.

Pre-hospital emergency care council. (2017b). Guidelines Clinical Practice - Paramedic

(p. 168). Ireland.

protégé. (n.d.). Retrieved June 11, 2019, from https://protege.stanford.edu/

Ramnarayan, P., Tomlinson, A., Rao, A., Coren, M., Winrow, A., & Britto, J. (2003).

ISABEL: a web-based differential diagnostic aid for paediatrics: results from an initial

performance evaluation. Archives of Disease in Childhood, 88(5), 408–413.

Razzak, J. A., & Kellermann, A. L. (2002). Emergency medical care in developing

countries: is it worthwhile? Bulletin of the World Health Organization, 80, 900–905.

RDF/XML Syntax Specification (Revised). (n.d.). Retrieved June 11, 2019, from

https://www.w3.org/TR/REC-rdf-syntax/

S K Mishra. (2018, April). Telemedicine for Rural Mass-Current Initiatives and Future

Scope. Presented at the Oditelecon2018, Lucknow, India.

Saba, T., Al-Zahrani, S., & Rehman, A. (2012). Expert system for offline clinical

guidelines and treatment. Life Sci J, 9(4), 2639–2658.

Schreiber, G., Wielinga, B., de Hoog, R., Akkermans, H., & Van de Velde, W. (1994).

CommonKADS: A Comprehensive Methodology for KBS Development. IEEE Expert:

Intelligent Systems and Their Applications, 9(6), 28–37. https://doi.org/10.1109/64.363263

Sharma, M., & Brandler, E. S. (2014). Emergency medical services in India: the present

and future. Prehospital and Disaster Medicine, 29(3), 307–310.

Shortliffe, E. (2012). Computer-based medical consultations: MYCIN (Vol. 2). Elsevier.

List of References

109

Smith, M. B., Chiovaro, J. C., O’neil, M., Kansagara, D., Quiñones, A. R., Freeman, M.,

… Slatore, C. G. (2014). Early warning system scores for clinical deterioration in

hospitalized patients: a systematic review. Annals of the American Thoracic Society, 11(9),

1454–1465.

Sonoda, T., & Ishibaei, K. (2015). Project on information-support solution in emergency

medical services. Fujitsu Sci Technol J, 51, 39–49.

SPARQL Query Language for RDF. (n.d.). Retrieved June 11, 2019, from

https://www.w3.org/TR/rdf-sparql-query/

Studer, R., Benjamins, V. R., & Fensel, D. (1998). Knowledge Engineering: Principles and

Methods. Data Knowl. Eng., 25, 161–197. https://doi.org/10.1016/S0169-023X(97)00056-

6

SWOOP - Semantic Web Standards. (n.d.). Retrieved June 11, 2019, from

https://www.w3.org/2001/sw/wiki/SWOOP

The Innovation Working Group (IWG) ASIA Task Force on Telemedicine. (2014).

Roadmap for Telemedicine: key considerations and recommendations. (p. 48). Retrieved

from https://www.ghdonline.org

Tleyjeh, I. M., Nada, H., & Baddour, L. M. (2006). VisualDx: decision-support software

for the diagnosis and management of dermatologic disorders. Clinical Infectious Diseases,

43(9), 1177–1184.

Tripathi, K. (2011). A review on knowledge-based expert system: concept and

architecture. IJCA Special Issue on Artificial Intelligence Techniques-Novel Approaches &

Practical Applications, 4, 19–23.

Uschold, M., & Gruninger, M. (2004). Ontologies and semantics for seamless connectivity.

ACM SIGMod Record, 33(4), 58–64.

Use Case Diagrams - Use Case Diagrams Online, Examples, and Tools. (n.d.). Retrieved

June 11, 2019, from https://www.smartdraw.com/use-case-diagram/

Vanamali, D., Sumalatha, N., & Varma, S. (2014). The role of National Early Warning

Score (NEWS) in medical emergency–patients in Indian scenario: a prospective

List of References

110

observational study. Journal of Evolution of Medical and Dental Sciences, 3(13), 3524–

3529.

Victoria, A. (2018). Ambulance Victoria Clinical practice guidelines for ambulance and

MICA paramedics. Ambulance Victoria.

Wardrope, J., & MacKenzie, R. (2004). The ABC of community emergency care: 2 The

system of assessment and care of the primary survey positive patient. Emergency Medicine

Journal, 21(2), 216–225.

Waters, M., Nightingale, P., & Edwards, J. (1990). A critical study of the APACHE II

scoring system using earlier data collection. Emergency Medicine Journal, 7(1), 16–20.

WebODE. (n.d.). Retrieved June 11, 2019, from http://mayor2.dia.fi.upm.es/oeg-

upm/index.php/en/old-technologies/60-webode/

Weih, H.-P., Schu, J., & Calmet, J. (1994). CommonKADS and cooperating knowledge

based systems. In Proceedings of the 4th KADS User Meeting. Citeseer.

Welcome to NetBeans. (n.d.). Retrieved June 11, 2019, from https://netbeans.org/

WHO World Health Organization (WHO). (1998). The world health report 1998 - Life in

the 21st century: A vision for all (p. 226). Geneva.

WHO World Health Organization (WHO). (2000). The World Health Report 2000.Health

Systems: Improving Performance (p. 215). Geneva.

WHO World Health Organization (WHO). (2010). The World Health Report 2010. Health

Systems Financing: The path to universal coverage. (p. 128). Geneva.

Wielinga, B. J., Schreiber, A. T., & Breuker, J. A. (1992). KADS: A modelling approach

to knowledge engineering. Knowledge Acquisition, 4(1), 5–53.

Woollard, M., & Greaves, I. (2004). 4 Shortness of breath. Emergency Medicine Journal,

21(3), 341–350.

Wright, D. (1998). Telemedicine and developing countries. A report of study group 2 of

the ITU Development Sector. Journal of Telemedicine and Telecare, 4, 1–85.

List of References

111

Yang, D., Tong, L., Ye, Y., & Wu, H. (2006). Applying CommonKADS and semantic web

technologies to ontology-based e-government knowledge systems. Asian Semantic Web

Conference, 336–342. Springer.

Yu, L. (2011). A developer’s guide to the semantic Web. Springer Science & Business

Media.

List of Publications

112

List of Publications

Rutvik K Shukla &DrChetan B Bhatt.Review of Early Warning Scoring System as

a Primary Diagnostic Tool in Critically ill Patient. International Journal of Applied

Engineering Research (IJAER) ISSN 0973-4562 Volume 13, Number 16 (2018) pp.

12783-12787

Rutvik K Shukla &Dr Chetan B Bhatt. CommonKADS Model Framework for Web-

based Emergency Medicine Decision Support System. Global Journal of

Engineering Science and Researches (GJESR) ISSN 2348 – 8034 Volume 5, Issue

9 (2018) pp.306-312

Rutvik K Shukla &Dr Chetan B Bhatt. Emergency Health Care ontology for Risk

Level Detection. Journal of Emerging Technologies and Innovative Research

(JETIR). ISSN 2349 – 5162 Volume 6, Issue 5 (2019) pp. 472-478

Doctor of Philosophy - Amazon S3...Medicine submitted by me for thedegree of Doctor of Philosophy is...

Documents

Transcript of Doctor of Philosophy - Amazon S3...Medicine submitted by me for thedegree of Doctor of Philosophy is...