Content uploaded by Gamal Attiya
Author content
All content in this area was uploaded by Gamal Attiya on Feb 10, 2017
Content may be subject to copyright.
International Journal of Computer Applications (0975 – 8887)
Volume 123 – No.10, August 2015
20
A Survey of Multi-Agent based Intelligent Decision
Support System for Medical Classification Problems
Hanaa Salem
Communications & Computer
Dept.
Faculty of Engineering
Delta University, Egypt
Gamal Attiya
Computer Science &
Engineering Dept.
Faculty of Electronic
Engineering
Menoufia University, Egypt
Nawal El-Fishawy
Computer Science &
Engineering Dept.
Faculty of Electronic
Engineering
Menoufia University, Egypt
ABSTRACT
There has been growing on big data since last decade for
discovering useful trends or patterns that are used in diagnosis
and decision making. Intelligent decision support system an
automated judgment that supports decision making is composed
of human and computer interaction to help in decision making
accuracy. Also multi-agent systems (MAS) are collections of
independent intelligent entities that collaborate in the joint
resolution of a complex problem. Multi-agent intelligent decision
support systems can be used to solve large-scale convention
problem. This paper is a survey of the recent research in multi-
agent and intelligent decision support systems to support for
classification problems. The purpose of the survey described in
this paper is the development of a novel large-scale hybrid
medical diagnosis system based on Multi-agent Intelligent
Decision Support System (IDSS) for distributed database.
Keywords
Multi-Agent; Intelligent Decision Support Systems; Diagnosis;
Feature Selection
1. INTRODUCTION
Medical diagnosis is known to be subjective for several reasons:
Firstly it relies on the physician making the diagnosis. Secondly,
and most essentially, the measure of information that should be
analyzed to make a good prediction is usually huge and at times
uncontrollable. Machine learning can be utilized to automatically
conclude diagnostic rules from descriptions of past, successfully
treated patients, and help experts and specialists make the
diagnostic process more objective and more reliable. An
Intelligent decision support systems are defined as interactive
computer systems used to help decision making to use data sets
and models in order to find problems, solve problems and take
decisions so dataset and models are designed to assist in decision
making through a semi-structured and unstructured decision
making modules that support for decision making [1]. A
definition for the using of the computer technologies in assistant
diagnose, which is: ‘Medical artificial intelligence is primarily
concerned with the construction of AI programs that perform
diagnosis and make therapy recommendations. Unlike medical
applications built on other programming approaches, such as
morally statistical and probabilistic approaches, medical AI
programs are based on representative models of disease units and
their relationship to patient factors. Recently, many studies focus
on intelligent decision support system rather than the Artificial
Intelligence in Medicine (AIM) systems [2].
Many studies in the use of the agents in medical classification
approaches represent the latest research direction, which intends
to disregard disadvantages of earlier developed medical
classification problems computational systems (medical expert
systems for example), that usually consist in limited autonomy,
interaction ability with the environment and intelligence in the
problem solving. Important applications of the agents and multi-
agent systems are represented in diverse problem solving that
appear in medical classification problems.
Examples of medical problems that can be resolved by artificial
agents, medical diagnostics elaboration, medical data gatherings
about patients, medical knowledge search, medical decisions
support, pro-active assistance of the physicians [3]. Agent-based
intelligent decision support systems (IDSS) to support decision
making is significant within the medical classification problems
because they let doctors and nurses to rapidly gather information
and process it in numerous ways in order to support with making
diagnosis and treatment decisions. These frameworks deigned to
support mainly in medical areas that is varied from the sorting
and recovery of medical records, storing and recovery of key
substances in medicines, analysis of real-time information
accumulated from screens, analysis of X-Rays, analysis of patient
history for the purposes of diagnosis, study of family history (for
cardiac conditions for example), and in many other areas [4].
These frameworks condescended to bolster for the most part in
therapeutic ranges that is assorted from the putting away and
recovery of restorative records, putting away and recovery of key
substances in pharmaceuticals, investigation of ongoing
information accumulated from screens, examination of X-Rays,
investigation of patient history for the reasons of analysis,
investigation of family history (for cardiovascular conditions for
instance), and in numerous different zones.
This paper is directed to highlight the current research in IDSS
and multi-agent. This will involve looking at multi-agents,
intelligent decision support systems (IDSS) and current Multi–
agent both inside and outside the classification medical data. The
information gathered in this research will help with identifying
potential issues with multi-agent based IDSS particularly with our
intended future work on Multi-Agent based IDSS for medical
dataset classification.
2. RELATED WORK
2.1 Overview of Intelligent Decision
Support Systems
In medical industry, the use of intelligent decision support
systems (IDSS) to support decision making is necessary and
important because they allow doctors and nurses to rapidly collect
information and process it in numerous ways in order to support
with making diagnosis and treatment decisions.
Intelligent decision support system (IDSS) is a result of
combining decision support system (DSS) and artificial
intelligent (AI). Its basic design thinking is to combine the
International Journal of Computer Applications (0975 – 8887)
Volume 123 – No.10, August 2015
21
knowledge reasoning techniques of AI and the basic function
models of DSS [5]. The benefits of using intelligent components
with DSSs as opposed to plain IDSSs to improve decisions
making timeliness, consistency in decisions, explanations and
justifications for specific recommendations, enhanced
management of uncertainty, and formalization of organizational
knowledge. The greatest useful of these advantages is the
enhanced explanations and justifications which is very much
indeed useful feature particularly in field like medicine, where it
helps if the real expert can authenticate the machines reasoning
[6]. Artificial Intelligence approaches have been used in
intelligent decision support systems to: 1) aid physicians in their
diagnosis techniques; 2) make decisions more exact, accurate and
effective, 3) reducing medical errors, and 4) improving patient
safety and decreasing costs. Artificial intelligence has become
commonly used in health-related decision support systems [7],
numerous artificial intelligence techniques such as artificial neural
networks [8].
In order to increase the diagnostic accuracy (DA) and decrease
the cost, it is necessary to optimize the check combinations and
achievement the check values fully. Till this point, with focusing
on colorectal cancer (CRC), an artificial intelligent algorithm
named DS-STM (diagnosis strategy of serum tumor makers) is
still under development [9]. Association Rule Mining [10].
Genetic algorithms [11], the decision tree classification algorithm
is a widespread cross-domain classification approach, used in
problem domains such as marketing and customer preservation,
fraud detection and medical diagnosis [12]. The ability of
machine learning tools to detect significant features from
complex datasets detects their importance. A variety of such
techniques, including Artificial Neural Networks (ANNs),
Bayesian Networks (BNs), Support Vector Machines (SVMs)
and Decision Trees (DTs) have been applied in a wide manner in
cancer studies for the development of predictive models, resulting
in effective and accurate decision making [13], a low cost
Smartphone based intelligent system incorporated with
microscopic lens that lets patients in remote and isolated areas for
uniform eye examinations and disease diagnosis. Mobile
diagnosis system have been used an artificial Neural Network
algorithm to analyze the retinal images taken by the microscopic
lens to recognize retinal disease conditions [14].
2.1.1 Some New types of IDSS
With a view to adjust different situations, people improve the
IDSS from various aspects and in different ways. This can be
summarized as follows:
2.1.1.1 Group Decision Support System (GDSS)
It backings gathering or aggregate decision-making: it combines
communication techniques, computer techniques and decision
support systems, artificial intelligent techniques (AI) and thinking
procedures, decision-making model methods, structuralization
collective decision methods [15].
2.1.1.2 Distributed Decision Support System
(DDSS)
This kind of backing is arranged not just toward individual
decision-makers or any decision-making gathering that remain for
the same organization, but also toward decision-making
organizations that have separate character and are joined at the
same time. It is impossible or in convenient to carry out some of
the large-scale decision-making activity. These activities include
decision-makers that have diverse responsibilities [16].
The data or important decision factors required for the decision-
making procedure are distributed among a bigger area and this
decision-making is a sort of association decision making or
distributive decision-making.
2.1.1.3 Intelligent, Interactive and Integrated
Decision Support System (3IDSS)
With the extension of the realistic field of DSS and the increasing
of the application level, DSS has gone into the decision-making
movement of regional medicinal and social advancement
technique and the manufacturing and working decision-making of
great-scale enterprises. Not just do these decision-making
activities includes all the sides of medical activities and all the
levels of management and administration, but also all the factors
are associated with each other and the decision-making
environment is problematical. Under this environment, a new
comprehensive decision-making. DSS initiated existence, which
is arranged toward decision-makers and the decision-making
process. That is Intelligent, Interactive and Integrated DSS,
abridged as 3IDSS [17].
2.1.1.4 Intelligent Decision Support System based
on Knowledge Discovery (IDSSKD)
It is applied in commerce and finance market with the analysis,
prediction and decision support. Generally speaking: IDSSKD =
DM + WM + KAS + RM
In this formula, DM, namely data mining, mainly represents
traditional structural data mining, in another word, KDD* system;
WM (Web Mining) mainly includes text mining based on web,
user access log mining and web structure mining; KAS
(Knowledge Acquirement System with Intervention) is mainly
used to acquire the knowledge of domain expert through
induction; RM (Reasoning Machine) is mainly used when IDSS
is carrying on reasoning. These four parts are combined through
synthetic knowledge base and web knowledge base [18].
2.2 Overview of Multi – Agent Systems
Multi agent technology is applied by intelligent systems to solve
the problems of analysis of complex systems and intelligent
management activities. Multi agent system technology is a
software paradigm where the concept of an agent denotes to “an
encapsulated computer system that is located in some
environment and that is capable of flexible, autonomous action in
that environment in order to recognize its design objectives” [19].
Intelligent Multi Agent Systems (IMAS) based learning take in
collection of information from their environment, recognition
data, and intelligent classification data and forecast future data,
storage data, delivery data to knowledge management systems
such as Decision Support System (DSS) and Management
Information System (MIS) [20].
A multi-agent system is functioned to automate processing within
each step using agents with diverse abilities, and provides the
parallelism of data processing using multi-agent system. The
power of system is in its capability to support concurrent
processing of gene data and the modular structural design that
tailors to biologists’ demands, also dispute the possibility of
applying Machine Learning technique to anatomize gene
expression data in a multi-agent environment [21]. Implement a
homeopathic medical aid platform; first have to figure out how
the general homeopathic medical diagnosis was done by a
homeopathic practitioner. At the first stage the patient declared
his/her infection (symptoms of illness) and also produced his
former prescription or clinical or lab report if any. Next the initial
diagnosis part where the signs of disease and several diagnoses
were done. At the last stage (the final diagnosis part) based on the
initial diagnosis and lab reports, the final diagnosis was done.
After final diagnosis it was determined whether to generate
International Journal of Computer Applications (0975 – 8887)
Volume 123 – No.10, August 2015
22
formula for the patients or recognize the patient to a hospital, if
his/her condition was critical [22]. The proposed hierarchical
contains of service mobile agents as upper layer agent,
coordinator agent as middle layer agent and initial agent as lowest
layer agent. Coordinator agent helps as matchmaker agent that
usages Naïve Bayesian learning method for gain general
information and selects the best service supplier agent using
matchmaking mechanism. Therefore this system can moderate
the communication overhead between agents for sending
messages and transferring data and can avoid sending the
problem to unrelated agents [23].
2.3 Review of Multi Agent Based on IDSS
Recently, there is the absence of one united framework for
integration of the two applicable flows of intelligent multi agent
technology and IDSS in real environment. An overview over
these agent-based methods and distributes them into five sections:
Medical data management, decision support systems, scheduling
and resource allocation, remote care as well as complex systems.
Here, we focus on decision support systems that essentially aim at
assisting professionals in their daily routine [24]. Agent is a
procedure that can exchange the user to carry out a specific task
individually and automatically. It can support the decision making
and problem solving in various stages and also can enhance the
functions of old-fashioned decision support systems. Multi-agent
is loosely coupled network and has good dexterity and
capability and it is used to form applications in a variety of areas
and it can solve large-scale difficult problems through reciprocal
scheduling and mutual cooperation among agents [25].
Application of evolutionary Multi-agent system (EMAS) to
analysis of gene expression data to find considerable
classification genes using simple classifier that can be used by
agent when searching through the gene expression database
search space, that agents improve their distinct performance
through evolution and cooperation with other agents. Simple
nearest neighbor classifiers are used in combination with
evolutionary multi-agent based system [26].
Health Agents, European Commission (EC)-funded research
project to improve the classification of brain tumors through
multi-agent decision support in extra of a distributed network of
local databases or Data Marts. Define a method to assess the
quality and availability of a new candidate local database
containing an arrangement of new cases, based on a compatibility
score [27]. Ecological-medical situation assessment is using
agent-based decision support system (ADSS). The system
receives statistical information in form of direct and indirect
pollution pointer values. The ultimate goal of the exhibited multi-
agent system (MAS) is to evaluate the power of the exposure to
pollutants in population health [28]. A medical diagnosis system
that combines the benefits of multi-agent system technologies and
neural networks in order to realize a greatly reliable, adaptive,
scalable, flexible, and robust diagnosis system for diseases. The
medical diagnosis system consists of a structured alliance of
medical experts - realized by agents - that cooperate in order to
provide a sustainable medical diagnosis. Each agent has a certain
responsibility. Reactive pattern-based matching process is the
main support for the agents [29].
Modern research areas in medical that include intelligent decision
support services, expert medical services and autonomous
management are based on multi-agent systems. The assistance of
these software agents offers efficient monitoring, analyzing, and
managing the data of patient where abnormal patterns are
detected to have an improvement treatment and avoid loss of life.
The research presented a framework for ubiquitous healthcare
based on multi-agent, proposes a mobile agent for diagnosis
support in global healthcare. The expert mobile agent (EMA)
categorize the data of patient by using neuro-fuzzy algorithm for
consultation report. A pre-processing technique rely on the profile
of an expert is used to as a filter of the patient history data.
Results of neuro-fuzzy from cross-validation test shows a great
accuracy in data classification compared to other highly accurate
classifiers [30].
A multi-agent system developed in order to enhance gene
expression analysis with the computerization of tasks about
identification of genes involved in a cancer, and classification of
tumors according to molecular biology. An agent that helps in
system integrity, carry out reading files of data amount of genes
from microarrays, information pre-processing, and with machine
learning approaches make groups of genes involved in the
process of a disease as well as the classification of samples that
could propose novel subtypes of tumors hard to identify based on
their morphology. Studies proved that the multi-agent system
needs a minimal intervention of user, and the agents generate
knowledge that decrease the time and complication of the work
of preventing and diagnosis, and thus let a more effective
treatment of tumors [31]. A multi agents system (MAS), which
distributes the diagnosis on three agents. Each agent is a
professional able to resolve and communicate with the other
agents [32].
In vivo magnetic resonance spectroscopy (MRS) consolidated
with in vitro multi-agent system (MAS) and gene expression
promised to enhance the classification of brain tumours and
produce novel biomarkers for prognosis. Health Agents was
tended this issue by building a distributed system of databases
centered on the clients and managed by agents. Thus, Health
Agents proposes an exclusive blend of state-of-the-art
technologies to advanced novel clinical tools for the diagnosis,
management and understanding of brain tumours [33]. Decision
algorithm is designed of decision tree and rough set. The method
is applicable for cooperative decision among Agents and can help
to whole complex works in Multi-Agent System. In the system,
Agents can complete mutual decision practice of algorithm about
cooperation and decision. As the Agent's own uncertainty, the
effectual coordination requests to be solved between Agents for
multi-Agent system [34].
An intelligent agent looks for circumstances that require
informing healthcare professionals about events. These events can
include giving shots, vaccinations, surgeries, follow-up checks
and other essential events. Analyses of administrative data are the
creator for these events as different to clinical data. The case
study existing was tested on determining pneumonia vaccinations
provider however it can be used in much greater areas [35].
Present an agent-based IDSS which is used for diagnosis includes
three levels. The deepest level is called agents, which contains
several kinds of agents including user interface agents, data
mining agents and problem solver agents. The second level is
communicators which act as translators among the task managers
and the agents to ensure that every agent gets the information in
their desired format. The upper level is task-managers, which
break the task up into sub problems that are given to agents [36].
Studies utilized multi-agent IDSS about cancer gene study the
system searches for information on several databases like
PubMed keeping in mind the end goal to gather the most up and
coming information as possible [37]. The EMG investigation
framework is a data mining system that is utilized to mine
electromyography (EMG) data the system contains three different
agent types. These are the task agent which is the client interface
agent, the sub-agent which has data around a specific part of
anatomy and the data agent which is used to mine one around a
specific field from one particular table [38]. The Clinical-HINTS
International Journal of Computer Applications (0975 – 8887)
Volume 123 – No.10, August 2015
23
system is DSS aimed to meet the requests for intelligent real-time
clinical management in critical care medical environments for
example, intensive care units (ICU). The system is a single agent
based system where the agent is utilized to assess the data and
stimulate any alarms that are required [39]. Intelligent agent
proposed structure depends on neural network classification
approach to solve the gap problem among two applicable flows of
intelligent/ multi agent technology and the methodology of
learning from real environment [20].
3. COMPARISION BETWEEN MULTI-
AGENT BASED IDSS
Determining key practical necessities for later improvement of
the multi-agent based IDSS, based on current multi-agents based
IDSS research applied within the classification medical diseases
was the purpose of this comparison framework, used to compare
the research introduced in the previous section, was advanced
using broad IDSS features, which include what category of
decisions the systems support, making the given IDSS for the
purpose of either decision speed or quality and decisions time
sensitivity. The data subsystem was compared upon data
medication features that at least one of the system tested
supported. The data distributed or not there support for distributed
data mining. The following features that were mentioned at
minimum one of the examined systems can be compared to the
agent subsystems. What are the languages of the agents coded in?
The structure of the agents which is either a single agent or
functional or layered multi-agent structures.
The user interface (UI) subsystem can be compared on numerous
features, the utility of the user interface which is that it is either
request/response where each request results is a reaction which
can then drive to another request or an interactive UI where the
person can justify and amend his requests at any time and the
computer can make clarification request if required, what was the
devices that UI available through.
This section presents the comparison of the study. The framework
is first presented, and then the comparison of the research based
on the framework. A discussion of the results of the comparison
is then presented. The motivation of this comparison is to
recognize open research questions, common system components,
and the common structure of multi-agent based IDSS to further
our research into multi-agent based IDSS for use cancer
classification. The comparison of the study is shown in table 1.
Table 1. The Comparison between Frameworks
Paper
No.
What does it
support?
Is the aim to improve
decision speed or
decision quality?
Is the data
distributed?
What
Languages
were the agents
coded in?
What is the
Agent
organizational
structure?
What sort
of UI is it?
Diseases
[26]
Diagnosis
Decision Quality
No
Not Specified
Functional
Not
Defined
Colon
Cancer
[27]
Diagnosis/
Prognosis
Decision Quality
Yes
Not Specified
Layered
Not
Defined
Brain
Tumors
[28]
Medical
Ecological
Assessment
Decision Quality
Yes
Not Specified
Layered
Interactive
Not
Defined
[29]
Diagnosis
Decision Quality
Yes
Not Specified
Single
Not
Defined
Any
Diseases
[30]
Diagnosis
Decision Quality
Yes
Java
Mobile
Not
Defined
Heart
disease
[31]
Diagnosis
Decision Quality
Yes
Java
Functional
Request/
Response
Cancers
[32]
Diagnosis
Decision Quality
No
Not Specified
Functional
Interactive
Not
Defined
[33]
Diagnosis
Decision Quality
No
Not Specified
Layered
Not
Defined
Brain
Tumors
[35]
Treatment
Decision Quality
Yes
Not Specified
Functional
Request/
Response
Not
Defined
[36]
Diagnosis
Decision Quality
Yes
Not Specified
Layered
Request/
Response
Not
Defined
[37]
Research
Decision Quality
Yes
Not Specified
Single Agent
Request/
Response
Not
Defined
[38]
Diagnosis
Decision Quality
No
Java
Layered
Request/
Response
Not
Defined
[39]
Diagnosis
Decision Quality
No
C++
Single Agent
Request/
Response
Not
Defined
4. DISCUSSION
The agent-based IDSSs studied cover the full scope of potential
uses in treatment, diagnosis and research. The greater parts of the
systems are deepened on enhancing decision quality instead of
speed as they are either time insensitive. None of the agent-based
IDSSs give feedback info about the data precision used to create
decisions. Just few of the systems use distributed data but of those
that do all of them are capable of distributed data mining.
Just four of the system descriptions contain what language was
utilized to program the agents, which were C++ and JAVA, this
while not significant data is honestly significant for maintenance
and upgrade purposes. For the most part, a considerable lot of the
systems investigated had agent types specialized to work at
different layers or for different purposes in the functional
structured system situation. Two kinds of multi-agent IDSS from
systems popular are layered as functional opposing. This is done
International Journal of Computer Applications (0975 – 8887)
Volume 123 – No.10, August 2015
24
to simplify the system and to make adjusting and improving the
agents or the framework simpler.
Most of the systems overviewed utilize a request/response
system, which is where the system is given guidelines from the
user and then gathers the necessary information and responds to
the user and expects further commands. Just two uses an
interacting system which is where the system might ask for or
automatically provide clarification in a real-time style. None of
the systems has any stand by forms of UI such as PDA, etc. Most
of the systems examined dealt with closed MAS. These resources
that the agents were intended to work cooperatively together
simply and with few problems. So no conflict handling
mechanisms were included, nor were any additional cooperation
instruments for example auctions or negotiations included.
5. CONCLUSION AND FUTURE WORK
In this survey paper we have presented how multi-agent
intelligent decision support system (MAS-IDSS) has been
integrated in medical classification Problems to resolve various
medical diagnosis and related problems. Still there are a few
preferences, advantages and disadvantages of using MAS-IDSS
in medical classification Problems. This survey paper aimed at
providing the basic information and related work using multi-
agent, IDSS and MAS-IDSS system. No one of the designed
techniques examined controlled time critical decisions or made
try to reduce the decision time only to attempt to increase the
decision accuracy. Most of the agent-based IDSSs examined
checked support for distributed database except four systems not
support distributed database mining. Only two of the systems
collect data from any real-time data streams.
As future work, more agents will be designed and implemented
for a multi-agent IDSS system to run in different machines,
located in several diverse places and attempts to reduce the
decision time and implementing more learning and statistical
approaches to enhance and improve gene identification and
cancer classification. Each of the agents may implicate of the
learning required to solve the problem and add or delete these
agents without affecting in the operation. Multi-agent IDSS
systems offer a natural way of attacking distributed problems.
Agents are able to communicate among themselves, using some
type of agent communication language, in order that interchange
any kind of information.
6. REFERENCES
[1] D. S. Kumar, G. Sathyadevi and S. Sivanesh,” Decision
Support System for Medical Diagnosis Using Data
Mining”, IJCSI International Journal of Computer
Science Issues, Vol. 8, issue 3, No. 1, Pages147-153,
May 2011.
[2] A. B. AL-Badareen, M. H. Selamat, M. Samat, Y.
Nazira and O. Akkanat,” A Review on Clinical Decision
Support Systems in Healthcare” Journal of Convergence
Information Technology (JCIT), vol. 9, no. 2, pages125-
135, March 2014.
[3] B. L. Iantovics,” Agent-Based Medical Diagnosis
Systems”, Computing and Informatics, vol. 27, no. 4,
pages 593–625, 2008.
[4] D. Foster, C. McGregor and S. El-Masri,” Survey of
Agent-Based Intelligent Decision Support Systems to
Support Clinical Management and Research”,
International Conference on Autonomous Agents and
Multi-agent Systems, ACM, 2005.
[5] A. Kaklauskas, “Biometric and Intelligent Decision
Making Support, chapter 2”, Intelligent Systems
Reference Library, © Springer International Publishing
Switzerland, Pages 31-85, 2015.
[6] S. Chakraborty, S. Gupta, “Medical Application Using
Multi Agent System - A Literature Survey”, Sougata
Chakraborty et al Int. Journal of Engineering Research
and Applications, Vol. 4, No. 1, pages 528-546, 2014.
[7] D. Bassen, S. Nayak, X. Chong Li and M. Sam,” Clinical
Decision Support System (CDSS) for the Classification
of Atypical Cells in Pleural Effusions”, Procedia
Computer Science, Elsevier, Vol. 20, No. 2, Pages 379–
384, 2013.
[8] R.R.Janghel, A. Shukla, R. Tiwari and R. Kala,
“Intelligent Decision Support System for Breast Cancer”,
Proceedings of the International Conference on Swarm
Intelligence, Springer Lecture Notes in Computer
Science, Beijing, China, Vol. 6146, No. 10, Pages351-
358, 2010.
[9] J. Shia, Q. Sud, C. Zhangb, G. Huangb and Y. Zhuc,” An
intelligent decision support algorithm for diagnosis of
colorectal cancer through serum tumor markers”,
computer methods and programs in biomedicine,
Elsevier, Vol. 100, Pages 97-107, 2010.
[10] C. W. Cheng, Ni. Chanani, J. Venugopalan, K. Maher,
and M. D. WANG,” icuARM _ An ICU Clinical
Decision Support System Using Association Rule
Mining”, Translational Engineering in Health and
Medicine, IEEE, Vol.1, No. 2, Pages 8-17, 2013.
[11] Aishwarya S and Anto S, “A Medical Decision Support
System based on Genetic Algorithm and Least Square
Support Vector Machine for Diabetes Disease
Diagnosis”, International Journal of Engineering
Sciences & Research Technology, IJESRT, Vol. 3, No.
4, Pages 4042-4046, April 2014.
[12] Y. Zhou, Y. Tan, H. Li and H. Gu, ”A Multi-Classifier
Combined Decision Tree Hierarchical Classification
Method”, International Symposium on Image and Data
Fusion (ISIDF), IEEE, Pages1–3, 2011.
[13] K. Kourou, T. P. Exarchos, K. P. Exarchos, M. V.
Karamouzis and D. I. Fotiadis, ”Machine learning
applications in cancer prognosis and prediction”,
Computational and Structural Biotechnology Journal,
Elsevier, Vol. 13, No. 6, Pages 8–17, 2015.
[14] A. Bourouis, M. Feham, M.A. Hossain and L. Zhang,”
An intelligent mobile based decision support system for
retinal disease diagnosis”, Decision Support Systems,
Elsevier, Vol. 59, No. 7, Pages 341–350, 2014.
[15] P. Gray, “The Nature of Group Decision Support
Systems”, Handbook on Decision Support Systems 1.
International Handbooks Information System, pages 371-
389, 2008.
[16] A. Gachet,” A New Vision for Distributed Decision
Support Systems” Decision Making and Decision
Support in the Internet Age, proceedings of the DSIage
Conference, Oak Tree Press, Cork, Ireland, Pages 341-
352, 2002.
[17] Z. Mo, S. Feng, and C. Tang, “A Study on Integrated
Model of Decision Support System”, Journal of System s
Science and System s Engineering, Vol. 11, No. 3, Pages
328-332, 2002.
International Journal of Computer Applications (0975 – 8887)
Volume 123 – No.10, August 2015
25
[18] F. Zhou, B. Yang, L. Li and Z. Chen,” Overview of the
New Types of Intelligent Decision Support System”, the
3rd Intetnational Conference on Innovative Computing
Information and Control (ICICIC'08), IEEE, Pages 267-
272, 2008.
[19] N. Jennings, “On agent-based software engineering”,
Artificial Intelligence, Vol. 117, No. 2, Pages277–296,
2000.
[20] R. Asadi, N. Mustapha, N. Sulaiman, “A Framework For
Intelligent Multi Agent System Based Neural Network
Classification Model”, (IJCSIS) International Journal of
Computer Science and Information Security, Vol. 5, No.
1, Pages168-174, 2009.
[21] H.C. Lam, M. Vazquez Garcia, B. Juneja, S. Fahrenkrug
and D. Boley,” Gene Expression Analysis in Multi-Agent
Environment”,
http://citeseerx.ist.psu.edu/viewdoc/summary?doi=10.1.1
.109.2453.
[22] S. Chakraborty and S. Gupta,”A novel Multi Agent
System based Homeopathic Medical Diagnosis System”,
CALCON 2014, IEEE, Pages1-4, 2014.
[23] E. E. Chaw,” Naïve Bayesian Learning based Multi
Agent Architecture for Telemedicine”, International
Journal of Innovation and Applied Studies, Vol. 2, No. 4,
Pages 412-422, Apr. 2013.
[24] D. Isern, D. Sánchez, and A. Moreno, “Agents applied in
health care: A review“, International Journal of Medical
Informatics, Elsevier, Ireland, Vol. 79, No. 3, Pages145-
166, 2010.
[25] S. Yongyong,” decision algorithms for Multi-agent
Intelligent Decision Support System based on
blackboard”, Information Technology Journal, Vol. 12,
No. 21, Pages 6235-6240, 2013.
[26] G. Stiglic, P. Kokol, “Colon cancer prediction with
genetics profiles using evolutionary techniques”, Expert
Systems with Applications: An International Journal,
Vol. 38, Issue 3, ACM, Pages 2752-2757, 2011.
[27] H. Gonzalez-Velez, M. Mier, C. Arus, B. Celda, S. V.
Huffel, P. Lewis, A. Peet and M. Robles,” Agent-Based
Distributed Decision Support System for Brain Tumour
Diagnosis and Prognosis” , Applied Intelligence, Spring
Link, Vol. 30, No. 3, Pages191-202, June 2009.
[28] M. V. Sokolova and A. F. Caballero, “An Agent-Based
Decision Support System for Ecological-Medical
Situation Analysis”, Nature Inspired Problem-Solving
Methods in Knowledge Engineering Lecture Notes in
Computer Science, Vol. 4528, Pages 511-520, 2007.
[29] C. Klüver, J. Klüver and R. Unland,” A Medical
Diagnosis System based on MAS Technology and Neural
Networks”, BPSC, vol. 147, Pages179-191, 2009.
[30] R. M. A. Mateo, L. F. Cervantes, H. Yang and J. Lee,“
Mobile Agents using Data mining for Diagnosis Support
in Ubiquitous Healthcare”, Agent and Multi-Agent
Systems: Technologies and Applications Lecture Notes
in Computer Science, Springer link, Vol. 4496,
Pages795-804, 2007.
[31] E. Marquez, J. Savage, C. Lemaitre, A. L. Laureano-
Cruces3, J. Berumen, A. Espinosa, R. Leder and A.
Weitzenfeld,” A Decision Support System Based on
Multi-Agent Technology for Gene Expression Analysis”,
International Journal of Intelligence Science, Vol. 5,
Pages158-172, 2015.
[32] O. Kazar, Z. Sahnoun and L. Frecon,” Multi-agents
system for medical diagnosis”, International Conference
on Intelligent System and Knowledge Engineering, Vol.
1, Pages1265 – 1270, 2008.
[33] C. Arus, B. Celda, S. Dasmahaptra and D. Dupplaw,” On
the Design of a Web-Based Decision Support System for
Brain Tumour Diagnosis Using Distributed Agents”,
International Conference, Web Intelligence and
Intelligent Agent Technology Workshops,
IEEE/WIC/ACM, Pages 208 – 211, 2006.
[34] S, Yongyong,” The Realization on Cooperative Decision
in Multi-Agent Intelligent Decision Support System”,
International Journal of Advancements in Computing
Technology(IJACT), Volume 5, No. 6, Pages 998-1007,
March 2013.
[35] C. Coffin, C. Saunders, C. Thomas, A. Loewen, N.
Campbell and William Ghali,” Using Intelligent Agents
to Repurpose Administrative Data in Fostering Disease
Prevention in an Outpatient Context: The Case of
Pneumococcal Vaccination”, Proceedings of the 37th
Hawaii International Conference on System Sciences,
IEEE, Pages1-7, 2004.
[36] M. E. Cohen, D. L. Hudson,” Meta Neural Networks as
Intelligent Agents for Diagnosis”, Neural Networks,
IJCNN '02. Proceedings of the International Joint, IEEE,
Vol. 1, No. 3, Pages 233 – 238, 2002.
[37] S. Z. H. Zaid, S. S. R. Abidi and S. Manickam,”
Leveraging Intelligent Agents for Knowledge Discovery
from Heterogeneous Healthcare Data Repositories”, US
National Library of Medicine National Institutes of
Health, , Vol. 90, Pages 335-340, 2002.
[38] J. Balter, A. Labarre-Vila, D. Ziebelin, C. Garbay,” A
knowledge-driven agent-centred framework for data
mining in EMG”, US National Library of Medicine
National Institutes of Health, C R Biol, Vol. 325, No. 4,
Pages375-82, 2002.
[39] D. Kalogeropoulos, E.R. Carson and P.D. Collinson,”
Clinical-HINTS: Integrated Intelligent ICU Patient
Monitoring and Information Management System”, US
National Library of Medicine National Institutes of
Health, Stud Health Technol Inform, Vol. 43, 1997,
Pages906-1010.
IJCATM : www.ijcaonline.org