摘要
研究诊断心血管病,结合形象化,利用图像检测,提高诊断水平。心音分段是心音图分析的一个重要环节。分段效果将直接影响心脏疾病的诊断水平。为了提高心音的分段正确率,根据动力学复杂性原理,提出一种简单度和归一化平均香农能的心音分段算法。对于复杂多样的心音信号,算法不仅分段正确率高,而且还能准确计算出第一心音和第二心音的时限。通过对144例正常心音和异常心音信号进行分段仿真实验,结果表明,算法是一种有效的和鲁棒性强的心音分段技术,为后期心音信号的识别奠定了良好的基础。
Heart sounds segmentation is an important part of the phonocardiogram （PCG） analysis. The quality of heart sounds segmentation will directly affect the level of heart disease diagnosis. In this paper, we have proposed a segmentation algorithm based on simplicity and normalized average Shannon energy （NASE） in order to improve the accuracy of heart sounds segmentation, by utilizing the dynamical complexity theory. Whereas the heart sound signals were intricate, the algorithm not only has high accuracy, but also accurately calculates the first and second heart sound time limit. Tested with 144 cases of normal and abnormal heart sounds, the results show that the proposed al- gorithm is an efficient and robust technique for heart sounds segmentation.
出处
《计算机仿真》
CSCD
北大核心
2011年第1期340-343,共4页
Computer Simulation
基金
光纤声传感阵列语音处理技术研究(桂科自(0832007))
关键词
心音分段
简单度
归一化平均香农能
峰逐层算法
Heart sounds segmentation
Simplicity
Normalized average Shannon energy
Peak peeling algorithm
