Database Reference
In-Depth Information
7. Knowledge Discovery and Data Mining in
Medicine
Takumi Ichimura, 1 Shinichi Oeda, 2 Machi Suka, 3 Akira Hara, 1
Kenneth J. Mackin, 4 and Katsumi Yoshida 3
1
Faculty of Information Sciences, Hiroshima City University, 3-4-1,
Ozuka-higashi, Asaminami-ku, Hiroshima 731-3194, Japan;
email: {ichimura, ahara}@its.hiroshima-cu.ac.jp
2
Department of Information and Computer Engineering, Kisarazu National
College of Technology, 2-11-1, Kiyomidai-higashi, Kisarazu, Chiba
292-0041, Japan; email: oeda@j.kisarazu.ac.jp
3
Department of Preventive Medicine, St. Marianna University School of
Medicine, 2-16-1, Sugao, Miyamae-ku, Kawasaki 216-8511, Japan;
email: {suka,k2yosida}@marianna-u.ac.jp
4
Department of Information Systems, Tokyo University of Information
Sciences, 1200-2, Yatohcho, Wakaba-ku, Chiba 265-8501, Japan;
email: mackin@rsch.tuis.ac.jp
Medical databases store diagnostic information based on patients' medical records. Because
of deficits in patients' medical records, medical databases do not provide all the required
information for learning algorithms. Moreover, we may meet some contradictory cases, in
which the pattern of input signals is the same, but the pattern of output signals is different.
Learning algorithms cannot correctly classify such cases. Even medical doctors require
more information to make the final diagnosis. In this chapter, we describe three methods of
classifying medical databases based on neural networks and genetic programming (GP). To
verify the effectiveness of our proposed methods, we apply them to real medical databases
and prove their high classification capability. We also introduce techniques for extracting
If-Then rules from the trained networks .
7.1 Introduction
Medical databases store diagnostic information based on patients' medical records.
Medical information such as laboratory tests, lifestyles, and chief complaint is
often ambiguous, and it is difficult to distinguish between normal and pathological
values.
Because of deficits in patients' medical records, medical databases do not
provide all the required information for learning algorithms. There may be some
contradictory cases, in which the pattern of input signals is the same, but the
pattern of output signals is different. Learning algorithms cannot correctly classify
such cases. Even medical doctors require more information to make the final
diagnosis.
Many algorithms for neural networks have been developed. Different
algorithms focus on different facets of learning, such as finding the optimal
 
Search WWH ::




Custom Search