家用血液检测服务系统中诊断与预测算法的研究

Research of Diagnosis Algorithm and Prediction Algorithm in Domestic Blood Test Services System

为了准确地获得C反应蛋白(C-reactive protein,CRP)的浓度,研制了一种家用血液检测服务系统,并设计了诊断与预测算法。诊断算法采用以决策树模型为基础的病情诊断算法,预测算法采用以灰色预测模型为基础的病情预测算法。两种算法的实验结果表明:病情诊断算法的测试集准确率为88.45%,符合诊断要求;病情预测算法的小误差概率P>0.95,拟合效果较好。该研究可为辅助心脑血管疾病的医学治疗提供参考。

In order to get the concentration of C-reactive protein （CRP） precisely, test services system was developed. Meanwhile, a diagnosis algorithm and a prediction algorithm weredesigned. Decision tree model was u e d in diagnosis algorithm and grey forecasting prediction algorithm. Experiments of two algorithms show that, diagnosis algorithm, s precision is 88.4 5 % and prediction algorithm , s minor error probability P is bigger than 0.95 , which meets thesystem requirements. These algorithms offer a reference for medical treatment of cardiocerebrovasculardisease.