基于蚁群聚类算法的动脉硬化无创检测

发布时间：2018-10-16 22:46

【摘要】：动脉硬化无创检测对于预防心血管事件具有重要意义。然而,基于心电信号或脉搏波信号的单一特征源的无创动脉硬化检测无法全面反映心血管动脉硬化事件。为了提高动脉硬化无创检测识别精度,提出了基于心电信号、脉搏波信号的多源数据无创动脉硬化识别方法,构建了具有变异特性的蚁群聚类算法,对提取的40组临床心电、脉搏波信号的特征值向量进行监督分类。通过对系统测试结果与专家分类结果对比分析,表明该方法提高了单一特征源的动脉硬化识别率,是一种有效的动脉硬化无创识别方法。
[Abstract]:Noninvasive detection of arteriosclerosis is of great significance in preventing cardiovascular events. However, noninvasive arteriosclerosis detection based on single characteristic source of ECG signal or pulse wave signal can not fully reflect cardiovascular arteriosclerosis events. In order to improve the accuracy of noninvasive detection and recognition of arteriosclerosis, a multi-source data noninvasive arteriosclerosis recognition method based on ECG and pulse wave signals was proposed, and an ant colony clustering algorithm with variation characteristics was constructed. 40 groups of clinical ECG were extracted. The eigenvalue vectors of pulse wave signal are supervised and classified. By comparing the system test results with the expert classification results, it is shown that the method improves the recognition rate of arteriosclerosis of a single characteristic source and is an effective noninvasive method for the identification of arteriosclerosis.
【作者单位】：内蒙古师范大学物理与电子信息学院;内蒙古农业大学机电工程学院;内蒙古大学电子信息工程学院;
【基金】：内蒙古自然科学基金(No.2013MS0924) 国家自然科学基金(No.61461042) 内蒙古师范大学科研基金(No.2012ZRYB001)
【分类号】：R54;TP18

【相似文献】