论文部分内容阅读
目的:探讨肺癌CT诊断中应用模糊神经网络辅助诊断的效果。方法:选取2012年1月-2013年12月远安县人民医院收治的60例肺癌患者,同时选取60例肺良性疾病患者作为研究对象。随机选择其中75例作为训练集,通过隶属度函数对训练集样本特征实施模糊处理,输入BP神经网络,进行网络训练,对剩余45例患者进行预测,观察FNN和3层前向BP神经网络(BPNN)预测结果。结果:模糊神经网络(FNN)诊断肺癌正确率、灵敏度和特异度均高于BPNN。结论:模糊神经网络应用于肺癌诊断的预测结果与病理检查结果相近,是一种有效的肺癌CT诊断计算机辅助诊断方法。
Objective: To explore the application of fuzzy neural network in the diagnosis of CT in diagnosis of lung cancer. Methods: Sixty patients with lung cancer admitted from Yuan’an County People’s Hospital from January 2012 to December 2013 were selected, and 60 patients with benign lung diseases were selected as the study subjects. Among them, 75 cases were chosen randomly as the training set, the training set sample characteristics were obfuscated by the membership function, then BP neural network was input into the network for training. The remaining 45 cases were predicted. FNN and 3-layer forward BP neural network BPNN) forecast results. Results: The accuracy, sensitivity and specificity of FNN in diagnosis of lung cancer were higher than that of BPNN. Conclusion: The result of fuzzy neural network applied to the diagnosis of lung cancer is similar to the result of pathological examination, which is an effective computer-aided diagnosis method for CT diagnosis of lung cancer.