基于人工神经网络的肺癌诊断研究

Research of Lung Cancer Diagnosis Based on Artificial Neural Network

目的利用人工神经网络模型ＢＰ算法的多层前馈网络模型原理，建立用于肺癌诊断的神经网络模型。方法利用人工神经网络的函数逼近功能模型，建立输入到输出的函数映射关系。结果该模型可较好地反映系统的动态性和数据的时序关联性。对肺癌诊断数据的应用结果显示肺癌病人的正确检出率为９６．２％，误诊率为３．８％；非肺癌病人的正确检出率为８８％，误诊率为１２％。结论基于人工神经网络的肺癌诊断方法具有较高的准确性。

Objective To set up neural network model to be applied in lung cancer diagnosis,utilizing the principle of artificial neural network model BP algorithm's multilayer proceeding feedback network model. Methods This paper set up the functional projective relationship from the imput to the output ,utilizing the function approximating model of artificial neural network. Results The model can reflect the dynamic characteristic of system and the timeconnected characteristic of data well.The result applied in data of lung cancer diagnosis shows that sensitivity is 962%,false negative rate is 38%;specificity is 88%,false positive rate is 12%. Conclusion The method of lung cancer diagnosis based on artificial neural network has a higher accuracy.