论文部分内容阅读
探讨径向基神经网络(RBF-NN)在胶质瘤分级中的应用价值。收集2008年2月~2009年4月我院116例胶质瘤病例,对各级别胶质瘤患者的年龄行Kruskal-Wallis H检验;并先后引入和剔除年龄变量分别对比RBF-NN模型、判别分析和RBF-NN与判别分析联合模型的预测结果。结果发现,不同级别胶质瘤的年龄有差异,且引入年龄的模型预测准确率高于未引入者;RBF-NN模型、联合模型对胶质瘤病理分级的预测准确率和Kappa值均优于常规Bayes判别分析模型。提示,年龄与胶质瘤的病理分级有关;以RBF-NN作为预测模型或辅助其他预测模型对胶质瘤病理分级预测有意义。
To explore the value of radial basis neural network (RBF-NN) in glioma grading. A total of 116 glioma cases from our hospital from February 2008 to April 2009 were collected. Kruskal-Wallis H test was used to evaluate the age of patients with glioma at various stages. RBF-NN model was compared with age variable separately Analysis and RBF-NN and discriminant analysis of the joint model of the forecast results. The results showed that the different grades of gliomas were different in age, and the accuracy of the models for introducing age was higher than that of the non-enrolled ones. The predictive accuracy and Kappa value of RBF-NN model and the combined model for glioma pathological grade were better than Conventional Bayes Discriminant Analysis Model. Prompted, age and glioma pathological grade related; RBF-NN as a predictive model or other prognostic models of glioma pathological grading prediction of significance.