论文部分内容阅读
目的研究一种基于多心电(ECG)周期融合和优先权分类的心室早期收缩(premature ventricular contraction,PVC)高精度检测方法。方法利用再定义ECG样本和2种不同ECG分割方法得到4个以非线性Hermite系数为特征的向量集。文中的数据取自MIT-BIH数据库,包括正常窦性心律(normal sinus rhythm,NSR)和PVC。进行一种基于类优先条件约束的改建二次判别函数(improved quadratic discriminant function,IQDF)的分类,其中以贝叶斯分类阈值为基准寻找在优先限定PVC错误率条件下使NSR错误率为最小的拉格朗日分类阈值。结果 PVC和NSR分别取得了99.29%和96.73%的分类精度。结论文中方法不仅能使PVC高分类精度得到优先保证,而且能使NSR分类精度保持在理想的高水平上。
Objective To study a method for the high accuracy detection of premature ventricular contraction (PVC) based on multi-cardiac cycle (ECG) fusion and prior classification. Methods Four recursive ECG samples and two different ECG segmentation methods were used to obtain four vector sets characterized by nonlinear Hermite coefficients. Data from the MIT-BIH database are included, including normal sinus rhythm (NSR) and PVC. An improved quadratic discriminant function (IQDF) classification based on class-first conditional constraint is proposed, in which the Bayesian classification threshold is used as a benchmark to find the NSR error rate minimized under the preferential PVC error rate Lagrangian classification threshold. Results The classification accuracy of PVC and NSR were 99.29% and 96.73% respectively. Conclusion The method in this paper can not only give high priority to the accuracy of PVC classification, but also keep NSR classification accuracy at a high level.