[A classification study of ovarian mucinous cystadenoma using fuzzy mathematics]

Ovarian mucinous cystadenomas (OMCA) are classified into three groups: benign, borderline and malignant. During diagnosis, pathologists often feel puzzled. Quantitative or semiquantitative computer-aided diagnosis is also defective. By using fuzzy mathematics, 129 cases (134 tumors) were studied. At first, they were classified objectively by fuzzy cluster analysis, and the results were used as the basis of study and the judicial criteria of methods. Using these data, an equation of fuzzy synthetical judgement was established. Then these cases were diagnosed again by the equation. The conforming rate was 96.27% (129/134). For comparison, from these data pluralistic concept regression equation was acquired by the computer, then these cases were diagnosed again by the equation, and the conforming rate was 94.03% (126/134). The difficulties in classification of OMCA are analysed and their causes are discussed. Compared with other methods, the merits and demerits of this new method are emphasized.