Clinical decision support system (CDSS) is a computerized program which analyses the patient's physiological data (e.g.
ECG, Heart beat, body temperature etc.) in order to find out symptoms of any
abnormality. These symptoms are used by the CDSS to estimate the current health
situation of the patient. The decision support system is also capable of making
decisions based on the diagnosis of estimated health situation. In the
architecture under discussion, we propose to create a hybrid of model-driven
decision support system and knowledge-driven decision support system.
Model-driven decision support system makes decisions based on the statistical
model of the patient's data. Knowledge-driven decision support system provides
specialized problem solving expertise stored as facts, rules, procedures, or in
similar structures. A hybrid system will augment the knowledge base with the
statistical model to make an improved decision. Thus, the resulting system will
be less vulnerable to 'false alarms'. A decision support system (DSS) can be 'passive' (only makes suggestions for diagnosis),
'active' (formulates diagnosis
and takes decisions) or 'cooperative' (formulates a diagnosis but it needs to be
verified by a consultant). We propose a cooperative DSS which will reach the
diagnosis through this hybrid decision making model. Then, it will present the
diagnosis as well as proposed decisions/actions to the medical consultant who
will verify the situation and decide whether or not the alarm is true. There are
many open-source as well as commercial clinical decision support systems
available in the market. The open source decision support systems will be
preferable to proprietary systems. Among the open source systems under
consideration, EGADSS (Evidence based Guideline and Decision Support System -
http://egadss.sourceforge.net/) is currently being reviewed in great detail to
provide a basis for the development of CDSS. The proposed CDSS consists of four
major components: (a) the data management system, (b) the model management
system, (c) the knowledge engine, and (d) the user interface.
The data management system is an important part of the CDSS. It consists mainly
of a database system used to store the patient's physiological data so as to
retrieve it when required. The database can be created using any specialized
open source or commercial database system available in the market. The data can
be retrieved using SQL queries to the database.
The Knowledge Engine contains the facts, rules, structures and procedures which
are based on expert knowledge. Simply stated, it is a rule base which contains
the decision making rules based on the previous experiences and the expertise of
the physicians. The structural development and accuracy of the rules contained
in the knowledge engine determine the accuracy of the diagnosis made by the
The Model management system is the brain of the CDSS. It is responsible for
creating a health model and comparing the data in the database with the model to
formulate a diagnosis. It works in conjunction with the knowledge engine to
diagnose the patient and to take decisions. patient and to take decisions.
The interaction of CDSS with physicians is only possible through a user
interface. The user interface allows the medical consultant to verify the
correctness of the diagnosis and decisions of the CDSS.