Text Mining Medical Device Guidelines to Explore Classification of Technologies

Regulation is an essential element for the industrialization of technologies, and the first and most basic step to facilitate the regularity of regulations is to classify the regulations. Classification of regulations helps both the regulated and the regulating parties. Regulatory agencies will be able to structure regulations for the appropriate users or challenges; the regulated citizens or companies will be able to know the appropriate rules and regulations they must abide by. This research aims to detect the relationship between technology and regulation in the medical field with text-analysis of medical device guidelines. In the past, we conducted a simple dichotomization according to whether they contained disease-related terms (e.g. disease name, body part) or not. The dichotomization was experimented on a set of medical device guidelines in Japan and technologies listed in the preceding technology forecast which lead to the patterning of the relationship between technology forecasts and guideline formulation. In this research, we utilize the classification we explored in our past research, and use available text mining tools on medical device guidelines in Japan to detect the relationship between technology and regulation within guideline documents.