会議名: 第10回バイオメディカル・ファジィ・システム学会
回次: 10
開催地: 倉敷
開催日: 1997/10/16 - 1997/10/17
p. 19-20
The present study describes the results of analysis performed in an attempt to predict cervical lymph node metastasis in carcinoma of the tongue. In previous study, 137 patients with squamous cell carcinoma of the tongue who were treated at our institution from 1971 to 1990 were investigated. Statistically, the quantification theory Type II of Hayashi (1954) was employed for discriminatory analysis. Ninety-four patients out of 137 patients (75.9%) could be discriminated accurately. When the method of this discriminatory analysis was applied to new cases which were treated from 1991 to 1992, 21 cases (65.6%) were pointouted accurately. The accurate rate of predicting cases among NO was 72.0%. From the results above, the accurate rate of this discriminant analysis was not satisfactorily high. Some histopathological items are subjective factors and sensitive items to the categorical scores. Therefore, the diagnostic logic using Fuzzy theory is being examined to the accurate prediction of cervical lymph node metastasis.