人工神经网络在类风湿关节炎诊断中的应用

The Application of the Artificial Neural Network in the Diagnosis of Rheumatoid Arthritis Disease

目的:应用BP人工神经网络原理,设计一种类风湿关节炎疾病诊断的方法。方法:选用对类风湿关节炎敏感的8个指标,作为BP人工神经网络的输入数据,对样本进行训练和预测。结果:BP人工神经网络经通过对150例样本的运算,训练集的113例样本,训练正确率为97.4%;预测集的37例样本,预测正确率为91.9%。结论:BP人工神经网络能为类风湿关节炎作出较准确的诊断,能提高诊断的客观性。

Objective. To set up a diagnostic method of Rheumatoid ar-thritis disease by the BP Artificial Neural Network theory. Methods： Selectingeight sensitive indexes as the inputting data for the BP Artificial Neural Netw-ork to forecast and train. Results. After the neural network computing about the 150 samples, the correct rate of the training set including 113 samples is 97.4% and the forecasting set including 37 samples is 91.9%. Conclusion： TheBP Artificial Neural Network can he made very correct diagnosis of Rheumatoid arthritis and can improve the objectivity about it.