基于决策树的胸痛辅助诊断研究

The Study of Auxiliary Diagnosis of Chest Pain Based on the Decision Tree

在院前救护的过程当中,胸痛患者病情较急,病发原因较多,危重程度差别较大,给胸痛种类的辨别造成了较大的困惑,为了帮助出诊医师提高胸痛患者抢救的效率,在胸痛中心信息化建设提供大量院前数据的基础之上,应用信息增益的方法对胸痛属性进行了简约,然后构建了胸痛诊断的决策树,通过测试模型分类的准确率超过82%,可以辅助胸痛中心医护人员有针对性的进行胸痛诊断,实现胸痛诊断智能化。

In the process ofpre-hospital rescue, chest pain patients have obvious characteristics, such as multiple quantities, numerous pathogeny and disparate critical level. It causeds great confusion to distinguish of chest pain. In order to help physicians to improve the efficiency of the rescue, the article come up with a model of the coronary heart disease diagnosis, which use information gain method to recognize the indispensable properties on the chest pain, and then build the chest pain diagnosis decision tree. Through test, the classification model has a good performance, accuracy is above 82%.So the method could be affirmed that can assist chest pain center staffto target the diagnosis of chest pain, to realize intelligent diagnosis.