17-03-2014, 06:54 PM
Abstract—Over the years, health institutions have collected
significant amount of medical data. By applying adequate
methods, collected medical data can be observed as a
knowledge base and used to support decision making.
Developments of medical knowledge-based systems, which
apply Case-based Reasoning (CBR) techniques to provide
decision support, have become very interesting research area.
Making a diagnosis is one of the most important areas in which
contemporary medical systems should give support to
physicians. Decision support in making diagnoses can be
achieved by filtering knowledge and experience, or in other
words, by using previous problem solutions which are the most
similar to the current problem. Thus, this paper describes
research results of applying CBR techniques for supporting
infective disease diagnoses.
Keywords-case based reasoning; decision support to
diagnostics; mobile decision support
I. INTRODUCTION
Case Based Reasoning (CBR) is a problem solving
paradigm that in many respects is fundamentally different
from other major AI approaches [1]. Basic aim of this
paradigm is to provide usable knowledge and experience
collected in previous problem solving. Nowadays, CBR
represents an important component of systems which are
based on knowledge and used in different areas of human
activity [2].
Medicine branch, as the area where decisions should be
extremely adequate and based on knowledge and experience,
has an indispensable need for efficient decision support. This
especially refers to more specialized area of medicine as
cardiology, infective diseases, dermatology etc. Complexity
and differences between areas of medicine make it difficult
to create efficient unique model for decision support
implementation.
Human health condition can be deteriorated by the
influence of many different factors. As a result, there are a
lot of mutually similar but also different cases. Great
similarity/difference complicates the process of finding
adequate mechanisms for case identification and adjustment.
Also, most of the values which characterize a patient’s
condition don’t have the same importance in all medical
branches. For example, in cardiology the important values
are heart beat and heart rhythm; in dermatology skin
appearance is an important value; in the area of infective
diseases the attention is on the values like temperature,
headache etc.
However, it is important to mention another project
which initially started with intentions to create some basic,
core system (framework) that can produce decision support
in different domains. As a direct result of these intentions
was “CaBaGe” (Case Base Generator). The main
characteristic of this system is that it is domain independent.
The input for the system is the description of the database
and the database from any domain. Based on those data,
system creates Case Retrieval Net (CRN) and it is capable to
solve new problems (or to propose solutions) from a domain
of the input database. The system reads the data from two
input files. In the first input file (“Case Pattern File”), the
description of the case is stored, while the second file (“Case
Base File”) contains the list of the already solved cases [5].
This paper describes a research with a main goal to
analyze requests and to implement decision support in
diagnosing infective diseases. An infective disease is a
clinically evident disease resulting from the presence of
pathogenic microbial agents, including pathogenic viruses,
pathogenic bacteria, fungi etc. Infective pathologies are
usually qualified as contagious diseases (also called
communicable diseases) due to their potentiality of
transmission from one person or species to another. Based on
the fact that infective diseases kill more people (worldwide)
than any other single cause, it is very important to provide
efficient decision support in diagnosing and treating infective
diseases [9].
significant amount of medical data. By applying adequate
methods, collected medical data can be observed as a
knowledge base and used to support decision making.
Developments of medical knowledge-based systems, which
apply Case-based Reasoning (CBR) techniques to provide
decision support, have become very interesting research area.
Making a diagnosis is one of the most important areas in which
contemporary medical systems should give support to
physicians. Decision support in making diagnoses can be
achieved by filtering knowledge and experience, or in other
words, by using previous problem solutions which are the most
similar to the current problem. Thus, this paper describes
research results of applying CBR techniques for supporting
infective disease diagnoses.
Keywords-case based reasoning; decision support to
diagnostics; mobile decision support
I. INTRODUCTION
Case Based Reasoning (CBR) is a problem solving
paradigm that in many respects is fundamentally different
from other major AI approaches [1]. Basic aim of this
paradigm is to provide usable knowledge and experience
collected in previous problem solving. Nowadays, CBR
represents an important component of systems which are
based on knowledge and used in different areas of human
activity [2].
Medicine branch, as the area where decisions should be
extremely adequate and based on knowledge and experience,
has an indispensable need for efficient decision support. This
especially refers to more specialized area of medicine as
cardiology, infective diseases, dermatology etc. Complexity
and differences between areas of medicine make it difficult
to create efficient unique model for decision support
implementation.
Human health condition can be deteriorated by the
influence of many different factors. As a result, there are a
lot of mutually similar but also different cases. Great
similarity/difference complicates the process of finding
adequate mechanisms for case identification and adjustment.
Also, most of the values which characterize a patient’s
condition don’t have the same importance in all medical
branches. For example, in cardiology the important values
are heart beat and heart rhythm; in dermatology skin
appearance is an important value; in the area of infective
diseases the attention is on the values like temperature,
headache etc.
However, it is important to mention another project
which initially started with intentions to create some basic,
core system (framework) that can produce decision support
in different domains. As a direct result of these intentions
was “CaBaGe” (Case Base Generator). The main
characteristic of this system is that it is domain independent.
The input for the system is the description of the database
and the database from any domain. Based on those data,
system creates Case Retrieval Net (CRN) and it is capable to
solve new problems (or to propose solutions) from a domain
of the input database. The system reads the data from two
input files. In the first input file (“Case Pattern File”), the
description of the case is stored, while the second file (“Case
Base File”) contains the list of the already solved cases [5].
This paper describes a research with a main goal to
analyze requests and to implement decision support in
diagnosing infective diseases. An infective disease is a
clinically evident disease resulting from the presence of
pathogenic microbial agents, including pathogenic viruses,
pathogenic bacteria, fungi etc. Infective pathologies are
usually qualified as contagious diseases (also called
communicable diseases) due to their potentiality of
transmission from one person or species to another. Based on
the fact that infective diseases kill more people (worldwide)
than any other single cause, it is very important to provide
efficient decision support in diagnosing and treating infective
diseases [9].