Using Association Rules to Comprehensively Scan Diseases Associated With Rheumatoid Arthritis
Student: Yu-Lun Chiu
Advisor: Dr. Jorng-Tzong Horng, Dr. Li-Ching Wu
Institute of Computer Science and Information Engineering,
National Central University, Taiwan
Introduction: Rheumatoid arthritis (RA) is a chronic inflammatory autoimmune disease. The common patients in women and the elderly. The age of onset is 40-60 years old in women. Patients with rheumatoid arthritis, if not in the early diagnosis and treatment, patients will suffer from persistent and permanent bone and joint damage, reduce the quality of life, or even reduce life. The aim of this study uses the association rule analysis to comprehensively scan the association of disease with rheumatoid arthritis.
Method: Based on NHIRD of the registration files and original claim data, the patients 30 years of age with rheumatoid arthritis. After using the International Classification of Diseases (ICD-9-CM) grouping the disease, the relative risk is calculated for each classification of diseases. Then, the data compiled into a WEKA required format, do the association rule to find out the relevance of the disease.
Result: Most of the results are consistent with the association of RA and disease in the known literature. Such as tuberculosis-related diseases, thyroid-related diseases, gout, autoimmune diseases, iron deficiency anemia, temporomandibular Joint disorders are associated with RA. Also found that some diseases associated with RA, but there is no literature to explore a single disease or at the same time suffering from a relationship between the two diseases and RA. Such as dehydration, hypokalemia, leukopenia, and Raynaud′s syndrome and SLE / SS are associated with RA.
Conclusion: We search the relevant research to verify the accuracy of the association rule analysis results and discover the relevance of the disease to the target disease.
Index terms－ Rheumatoid Arthritis, Comprehensive scanning, Association rule||en_US|