テキストマイニングを用いた医療機器分野の研究動向分析

抄録

To quantitatively investigate the history of medical device research in Japan, we analyzed Japanese titles of “Annual conference of Japanese Society of Medical Instrumentation (JSMI)” during last 45 years using text mining. All general and topics presentation titles (4379 presentations) between 1975 and 2022 were collected from article databases, and the text of titles was analyzed using the KH Coder, free software for quantitative text analysis of the Japanese language. Several research categories were identified through multivariate analysis of frequently appearing words. Also, coding rules were created to assign presentation titles to these research categories, and the categorization results were analyzed quantitatively. Sterilization and infection control regarding surgery were principal categories in each age, with the availability and evaluation as the major area of research. In the 1970’s and 80’s, research and development of artificial organs and biomedical measurements were actively reported in JSMI, but there were notably less in recent years. Medical device safety area such as clinical engineering, maintenance management and traceability were significantly increased since 2000’s. Our findings suggest that there has been a shift in the dominant research area of medical devices in Japan.