Development of a Medical Inferencing System Using Data Clustering

Abstract

Now a day with the advance technology data collection is far easier than before. Data of relevant subjects are collected and stored for future analysis. The analysis can be online or offline. Data can be collected from different sources like industrial plant, wireless sensors, stock market, banking and financial institutions, and medical parameters of the patient. The data are collected with the help of different sensors or transducers and stored in digital format in different storage media. Data mining or knowledge discovery from raw data is very much important and a lot of research has been going on in this field. Knowledge discovery gives a better edge to a person to take better action. Medical diagnosis is a very important area of research where with the help of engineering techniques diagnosis is made. A new approach of medical diagnosis uses the day to day monitoring of the patient’s medical data to determine the type of disease, degree of seriousness of the disease. First of all the day to day medical monitoring data is taken and clustered, so that it can be arranged in a better way. Different data mining or soft computing techniques are implemented in the clustered data to discover hidden trends, knowledge in the data which helps to gain an insight view of the subject. This thesis tries to develop a medical diagnosis system using the day to day patient monitoring data. First of all the data are acquired and clustered or grouped. Fuzzy based inferencing techniques are used to discover knowledge from the clustered data.