2004 Fiscal Year Final Research Report Summary
Extraction of Prediction rules for Infective Diseases from Medical Inspection Data
| Project/Area Number |
13558036
|
|---|
| Research Category |
Grant-in-Aid for Scientific Research (B)
|
| Allocation Type | Single-year Grants |
|---|
| Section | 展開研究 |
|---|
| Research Field |
Intelligent informatics
|
|---|
| Research Institution | Kyusyu Institute of Technology |
|---|
Principal Investigator |
HARAO Masatretu Kyusyu Institute of Technology, Faculty of Computer Science and Systems Engineering, Professor, 情報工学部, 教授 (00006272)
|
|---|
| Co-Investigator(Kenkyū-buntansha) |
HIRATA Kouichi Kyusyu Institute of Technology, Faculty of Computer Science and Systems Engineering, Associate Professor, 情報工学部, 助教授 (20274558)
YAMADA Keizo Kyusyu Institute of Technology, Faculty of Computer Science and Systems Engineering, Research Assistant, 情報工学部, 助手 (60325579)
YOKOYAMA Shigeki Koden Industry Company, Infomedical, Chief Researcher, インフォメディカル事業部, 研究主任
|
|---|
| Project Period (FY) |
2001 – 2004
|
| Keywords | data mining / association rule / infective disease / hospital-acquired disease / drug resistance / decision support system / MRSA / server & client system |
|---|
| Research Abstract |
The purpose of this research is to establish a method of extracting prediction rules on outbreak of hospital-acquired diseases from medical test data-base which are stored in hospitals, and to construct a decision support system for preventing hospital-acquired diseases. We have performed the research by setting the following sub-themes :
(1)Establishing a method of extracting rules from medical test data-bases.
(2)Construction of a decision support system for preventing hospital-acquired diseases
For the former, we investigated two methods to extract rules of the style "if_then_else" based on the data-mining approach, one is the decision tree construction and the other is the association rule extraction. Especially, we discussed the prediction rules concerning to MRSA which is serious in present hospital-acquired infection. Furthermore, it is important for making clear the relation between medical treatment and hospital-acquired infection to analyze the characteristics such as the change
… More
of drug resistance. From this viewpoint, we have also discussed the sequential pattern mining. We used the medical test data which are stored in a certain hospital from 1994 to 1999 (8MB) as the object data. Medical staffs examined the obtained results and evaluated that some valuable rules are included though the majority were predictable ones.
For the latter, we investigated the construction of decision support system for preventing hospital-acquired diseases for persons engaging to medical institutions from the data warehouse viewpoint. We have developed the system in the server & client style by using JAVA language to use it in the network environment. The developed system consists of the rule extraction tool and the rule examination tool. The rule extraction tool works under SQL circumstance. The rule examination rule has the function to judge the significance of any given rules, and to display the analyzed statistical data in the forms of graphs or tables so that medical staffs can understand the situation. Less
|
