Like data analysis, pattern recognition and data mining, fuzzy clustering also has been applied widely, and successful applications have been reported. In this paper we aim to employ the technique of fuzzy c-means (FCM) cluster to prognosticate the operation possibility on gastric cancer patients. Our purpose is to partition some clinical data in two fuzzy clusters. One of them considers patients who have a chance for successful surgery whereas the other cluster contains the patients without a view for surgery. Each patient is given by characteristic biological markers. The initial values of membership degrees taking place in the partition matrix are usually determined randomly. In this work we will use particularly designed membership functions to calculate the degrees of membership.