Medicine

• Authors: Shusaku Tsumoto
• Venue: Handbook of data mining and knowledge discovery
• Year: 2002

The development of computer systems has contributed to both medical research and practice and their contribution is now entering into a new phase. Since the early 1980s, there has been a rapid growth in hospital information systems (HISs) leading to a large proportion of laboratory examinations being stored as a huge database. Other types of data, including medical images, will be stored in HISs within the twenty-first century. Thus, it is expected that data mining methods will find interesting patterns from databases of such stored data and will be important for medical research and practice because human beings cannot deal with such a huge amount of data. This chapter provides a practical introduction to knowledge discovery and data mining in medical databases, especially focusing on the following points: (1) the kind of problems medical people want to solve, (2) characteristics of medical data, (3) problems with medical data mining, especially the importance of preprocessing, and (4) an overview of existing research. Discussions show that medical data mining is still in its early days and many problems are still to be solved, even with existing data mining techniques. This suggests that this field should be a hot research topic in medical informatics in the twenty-first century and is awaiting further contributions.