支持向量机和人工神经网络在冠状动脉旁路移植术后晚期静脉移植血管病患病风险预测中的应用

Application of Support Vector Machines and Artificial Neural Networks in the Risk Prediction of Advanced Saphenous Vein Graft Disease after Coronary Artery Bypass Grafting

目的探讨支持向量机和人工神经网络在预测个体冠状动脉旁路移植术后晚期静脉移植血管病患病风险中的应用。方法选取2015年3月-2017年12月天津市胸科医院CABG术后超过一年的冠状动脉粥样硬化性心脏病患者,分别应用径向基SVM、多项式SVM和BP神经网络建立晚期SVGD预测模型。通过受试者工作特征曲线下面积、精确率、召回率及F1指标评价模型的预测性能。结果 BP神经网络在测试集中反映模型精确率和召回率的F1值为0.84,而ROC曲线下面积均值为0.773,大于其他两种SVM预测模型。结论 BP神经网络对晚期SVGD的预测表现更佳,有助于临床的辅助诊断。

Objective To investigate the application of support vector machines and artificial neural networks in the prediction of the risk of advanced saphenous vein graft disease after coronary artery bypass grafting.Methods CHD patients who had been more than one year after CABG surgery in Tianjin Chest Hospital from March 2015 to December 2017 were selected to build advanced SVGD prediction models using radial basis SVM,polynomial SVM and BP neural network respectively.The predictive efficacy of models was evaluated by area under the receiver operator characteristic curve,precision rate,recall rate and F1 index.Results The F1 index of BP neural network reflecting model precision rate and recall rate in the test set was 0.84,while the average area under the ROC curve was 0.773,higher than the other two SVM prediction models.Conclusion BP neural network performs better in predicting advanced SVGD,which is helpful for clinical auxiliary diagnosis.