Akinori Abe, Norihiro Hagita, Michiko Furutani, Yoshiyuki Furutani, and Rumiko Matsuoka [Abstract]

[Abstract]

Categorized and Integrated Data Mining of Medical Data from the Viewpoint of Chance Discovery

Akinori Abe¹, Norihiro Hagita², Michiko Furutani³, Yoshiyuki Furutani³, and Rumiko Matsuoka⁴
1. NTT Communication Science Laboratories, JAPAN
2. ATR Intelligent Robotics and Communication Laboratories, JAPAN
3. Department of Pediatric Cardiology, Tokyo Women's Medical University, JAPAN
4. Wakamatsu Kawada Clinic, JAPAN

In this paper, we analyze the procedure of computational medical diagnosis based on collected medical data. Especially, we focus on features or factors which interfere with sufficient medical diagnoses. In order to reduce the data complexity, we introduce medical data categorization. Data are categorized into six categories to be analyzed and to generate rule sets for medical diagnosis. We analyze the relationships among categorized data sets within the context of chance discovery, where hidden or potential relationships lead to improved medical diagnosis. We then suggest the possibility of integrating rule sets derived from categorized data for improving the accuracy of medical diagnosis.