CHAPTER ONE
1.0 INTRODUCTION
One of the applications of medical information has been the implementation and use of expert system to predict medical diagnosis based on a set of symptoms.
Furthermore, such expert systems serve as aid to laboratory testing routines and effective treatments. An intelligent computer program assisting medical diagnosis offer the physician easy access to a wealth of information from past patient data such resources may help hospitals reduce excessive cost from unnecessary laboratory testing and ineffective patient treatment, while maintaining high quality of medical care.
These develop an improved expert system e.i. computerized system that will build the unfailing method or knowledge base for such a system. On major drawback of conventional medical expert system are, they use static knowledge bases, developed from a limited population size and a limited number of case demographic and geographic location.
The knowledge base is inherently not dynamic and is not undated to keep up with emerging trends such as the appearance of increase prevalence of diagnosis before now unseen.
Thus, after a given period of time this inflexibility limits the use of the knowledge base as it no longer reflects the current characteristics of the population at risk.
Given this point, the development of a knowledge base using this artificial network technology naturally lend itself toward the task of predicting medical diagnosis. In addition, this technology appear to be promising method for recommendation possibly routines.
One to it dynamic nature and online learning capability and web network knowledge-base date, thus, once an initial knowledge-base has been set up it can effectively capture varying ailment trends in a given population while retaining its precious knowledge.
This research will construct a web-base network capable of predicting various medical diagnosis with a plausible degree of accuracy.
The data to be use for this task are:
- Patient demography
- Present vital signs
- Present of specific symptom in patient
- Patient verbal complaints
- The diagnosis by physician
The element of the first four components serve as an input variable to the network and the last (fifth) component serve as the output.
Therefore, the task of the web-base network is to draw a dynamic correlation between the patient’s presentations, using self reported symptoms and vital signs (complaint 1 to 4) and the diagnosis recorded by the physician (complaints).
The patient’s demographics and vital sign are key element in providing dues to an ailment. It was observed that the patient literal complains provides great in sight for predicting medical diagnosis.
1.1 STATEMENT OF PROBLEM
The major problem in training a web-base network to patient diagnosis is to process the data in such a manner as could be interpreted by the neural networks variable transformation modules.
The steps toward achieving this will involve using a software package that will suit such task, this application will help the physician every patient. Problems facing manual system or conventional system in hospital are:
- INSECURITY OF DATA:- Hospitals where manual system is in operation, data are not very secured due to the way patient file are being kept in the hospitals.
- LACK OF FILE ORGANIZATION:- Using manual or conventional system of file/cards organization in hospital causes a tedious problem to the nurses/physician during storing processing and searching of past patient files.
- Manual system does not permit file keeping for a very long time.
- TEDIOUS AND STRESSFUL:- Manual Operating Hospital find jobs stressful and time consuming around them.
- ILLEGIBILITY IN WRITING:- Illegible writing sometimes causes time wasting while searching for patients file. Hence these works will art against those negative vises.
1.2 OBJECTIVES OF THE RESEARCH
The objectives of this project research are:
- To assist physician facilitate reasonable task, treatment and minimize unnecessary laboratory routines so as to reduce operational cost.
- To predict medical diagnosis and possible treatment.
- To construct and train Artificial web-base network (AWBN) that will serve as a dynamic “Look up Table” for accurate classification of medical diagnosis due to patient symptoms.
