University of Isfahan Computational Intelligence in Electrical Engineering 2821-0689 8 2 2017 08 23 Multi-lead ECG Compression Based on Compressed Sensing Theory Multi-lead ECG Compression Based on Compressed Sensing Theory 13 24 21794 10.22108/isee.2017.21794 FA Siavash Eftekharifar Faculty of Electrical and Computer Engineering, University of Tabriz, Tabriz, Iran Tohid Yousefi Rezaii Faculty of Electrical and Computer Engineering, University of Tabriz, Tabriz, Iran Sabalan Daneshvar Faculty of Electrical and Computer Engineering, University of Tabriz, Tabriz, Iran Amir Rastegarnia Department of Electrical Engineering, University of Malayer, Malayer, Iran Azam Khalili Department of Electrical Engineering, University of Malayer, Malayer, Iran Journal Article 2016 08 19 The purpose of this paper is to exploit the compressed sensing theory in order to compress multi-lead ECG channels with a high compression ratio and minimum reconstruction error. In order to obtain the sparse representation of the ECG signals a basis matrix with Gaussian kernels which have the maximum resemblance with ECG signals, is constructed. Then using Orthogonal matching pursuit, algorithm which is a greedy/iterative optimization technique, the sparse representation is acquired. Finally, utilizing the compressed sensing theory is possible. In order to prove the accuracy of the algorithm the same optimization technique is used to reconstruct the compressed signal. Using a wavelet basis is also common to obtain the sparse representation. The compressed sensing theory is also applied to the ECG signals for which their sparse representations have been obtained using a wavelet basis. The results show the superiority of the proposed method over the wavelet basis. The purpose of this paper is to exploit the compressed sensing theory in order to compress multi-lead ECG channels with a high compression ratio and minimum reconstruction error. In order to obtain the sparse representation of the ECG signals a basis matrix with Gaussian kernels which have the maximum resemblance with ECG signals, is constructed. Then using Orthogonal matching pursuit, algorithm which is a greedy/iterative optimization technique, the sparse representation is acquired. Finally, utilizing the compressed sensing theory is possible. In order to prove the accuracy of the algorithm the same optimization technique is used to reconstruct the compressed signal. Using a wavelet basis is also common to obtain the sparse representation. The compressed sensing theory is also applied to the ECG signals for which their sparse representations have been obtained using a wavelet basis. The results show the superiority of the proposed method over the wavelet basis. https://isee.ui.ac.ir/article_21794_6899dc92a6b06337fd7e76d16de55459.pdf