An optimal technique for ECG noise reduction in real time applications

This paper presents a novel and efficient algorithm of ECG compression in real time monitoring systems, updated with each new input signal sample. This algorithm tries to improve the compression ratio of the captured signal by means of an optimal noise threshold in terms of hardware complexity and memory requirements. Threshold estimation is computed, using the instantaneous standard deviation, in order to decrease data sorting and storing resources, and allowing low-cost implementation in portable electronic systems. This method produces the highest number of null samples (more than 88.7%) using a low threshold and signal errors with very acceptable merit figures (99.876% of EPE, and 0.193% of MSE). The quality of the recovered signal is good for the clinical diagnosis, obtaining a superior compression rate in spite of using instantaneously captured ECG signals.