Pre-processing ECG signals for smart home material application

Health monitoring plays a vital role with regards to early detection, prevention of any form of illness which will promote good health and overall, well-being of the people. Nowadays Smart home systems, IoT based monitoring systems, medical bracelets, Invasive/non-invasive medical sensors are widely used to monitor the physical health of the people. Electrocardiogram signals (ECG) which are used to access the electrical function of the heart. Proper monitoring of ECG signals will help us to prevent major heart illness. It also helps us to ensure the oxygen pumping ability of the heart which is very essential to maintain the required oxygen saturation level of the body. ECG recordings are in general prone to various type of noises. Turbulent ECG signals may also lead to wrong detection and evaluation. Thus, preprocessing of recorded signals plays a major role in health monitoring. This paper aims to denoise the ECG signals by using filters in an efficient manner. Performance analysis of the filters are evaluated by comparing the level of variation of the signal and the noise which is expressed in terms of SNR, Correlation coefficient (COR), Mean Absolute errors (MAE), Mean Square Error (MSE). Our preprocessing approach has been valuated using ECG signals from Physionet database.