In the event of design changes in a medical device, the same pathway for approval as the initial released product must be obtained from the authority in accordance with country-specific regulations. In this paper, we collect materials on the change management of medical devices approved by the US Food and Drug Administration (FDA) and analyze the design change data sets from the viewpoint of a digital X-ray device. Based on this analysis, we investigate the association among the items that need a 510(k) to be cleared, and then present a model for assessing the impact of 510(k) clearance on the basis of the related items. From this model, two key contributing causes for decision of change permission are derived, expressing as membership function graphs. Moreover, we will develop a fuzzy-based expert system that predicts the degree of the change registration of an X-ray device through a deductive reasoning in advance for many design change ideas. Finally, we analyze the correlation between the evaluation result of non-experts using this system and the evaluation result depending on the experience of human experts according to FDA 510(k) guidance.
