Fuzzy Logic for the Diagnosis of Surgical Site Infection by the Weighted Membership Functions

抄録

The medical diagnosis plays an important role in patient care. However, it is difficult to make the diagnostic mechanism clear because for the diagnosis, the physician uses many kinds of information obtained from the patient which is largely subjective resulting in the diagnostic logic being fuzzy. For examples, to prevent postoperative surgical site infection, the surgeon's decisions is baqsed on several items such as fever, the complains of patients, WBC counts and so on. In this paper, we propose a diagostic system for surgical site infection based on multiple items and time series data, by bthe new method offuzzy inference. The application of this system to surgical cases produced results as reliable as that of expert surgeons.