论文部分内容阅读
瞬时胎心率是监测胎儿健康状态的一种重要方式。当前,监控胎儿心率是重要而复杂的任务,正确的自动化分类和规则提取是非常必要的。医疗诊断自动化系统,不仅加强医疗保健,同时也可以降低成本。设计了一个有效挖掘规则,并根据给定的参数来预测胎儿的风险水平。采用C4.5、Classification and Regression Tree(CART)、随机森林分类器来进行系统比较。该系统的性能评价由分类精度、产生规则数量构成。实验结果表明,基于随机森林分类器的系统具有高精度(99.4%)的预测胎儿健康状态的潜力,同时,产生的规则数量精简且可供于医生决策。
Instantaneous fetal heart rate is an important way to monitor fetal health status. Currently, monitoring fetal heart rate is an important and complex task, the correct automated classification and rule extraction is necessary. Medical diagnostic automation system, not only to strengthen health care, but also can reduce costs. An effective mining rule is designed and the fetal risk level is predicted based on the given parameters. C4.5, Classification and Regression Tree (CART), random forest classifier for system comparison. The system performance evaluation by the classification accuracy, the number of rules to form. The experimental results show that the system based on stochastic forest classifier has the potential of predicting fetal health status with high precision (99.4%), while the rules generated are streamlined and available for physician decision-making.