Real ECG signal acquisition with shimmer platform and using of compressed sensing techniques in the offline signal reconstruction

In this paper, the best ECG signal recovery quality of the realistic ECG signal and on which compressed sensing techniques have been applied, is investigated. The contribution of this paper is two folds: Firstly, a real-time ECG signal acquisition is performed using a Shimmer platform. Secondly, the collected are processed by the mean of compressed sensing approach to reduce the signal size and cope with the sensors requirements of high energy efficiency, thus a longer battery life. The recovery of the original signal is afterwards ensured by resorting to Orthogonal Matching Pursuit (OMP) and Matching Pursuit (MP) algorithms and PSNR is used as a metric for the reliability of the signal reconstruction. Several tests are carried out each time by either modifying the type and the size of the dictionaries or the number of iterations. It is shown that the best signal reconstruction is achieved by a value of PSNR that reaches 57 dB.