期刊名称：International Journal of Computer Trends and Technology
电子版ISSN：2231-2803
出版年度：2017
卷号：45
期号：1
页码：4-9
DOI：10.14445/22312803/IJCTT-V45P102
出版社：Seventh Sense Research Group
摘要：Electrocardiogram (ECG) is the bioelectric signals generated during the cardiac cycle. ECG shows the actual functioning of the heart and various heart diseases can be diagnosed after analyzing ECG signals. In addition to that ECG signals are required to be compressed for costeffective storage and fast transmission over low bandwidth channels to the remote locations. Principal Component Analysis identifies the most important features of ECG signals and ECG signal can be compressed using these Principal Components. Principal Components with high variance can be retained and low variance can be discarded. Thus the ECG signal can be compressed. Multiple leads of ECG signal can depict the clear picture of the heart therefore in this paper singular value decomposition (SVD) is used to obtained combined Principal Components of twelve lead PhysikalischTechnische Bundesanstalt (PTB) Data Base ECG signal. Along with this ECG signal is also compressed and reconstructed with the help of first few Eigen Vectors and quantity of compression is measured by Compression Ratio (CR) and reconstructed signal quality is measured by Percentage Root Mean Square (PRD) difference.
关键词：Compression Ratio (CR); Percentage Root Mean Square Difference (PRD) difference; Eigen Value; Eigen Vector; Principal Components analysis
