Enhancement and Denoising of ECG Signal Using Extended Kalman Filter and Extended Kalman Smoother

Considering the medical fields, there are many instruments and electronic gadgets where the resulting output is Bio-signals does play vital role as per the need viz. EEG (Electro-encephalogram), ECG (Electrocardiogram), EMG (Electro-myogram) and MEG (Magneto cardiogram). The output signals helps in diagnosing different diseases and medical disorders into the patients. Concerning with the output signals from ECG machines/gadgets find irrelevant linkage with different noise frequencies resulting in faulty or false output. These noises are basically from electrode artifacts, muscles and lines. This paper focuses on a nonlinear Bayesian filtering framework for the filtering of noisy electro-cardiogram (ECG) recordings. While analyzing the same there had been made the practical usage of filtering framework through an Extended Kalman Filter (EKF) and an Extended Kalman Smoother (EKS). The respective methods applied helps in filtering of noisy ECG signals and enhancement of ECG signals. This method mentioned has practically been conducted on various ECGs (normal in their operations), there had been an artificial addition of white Gaussian noises outputting and suggesting for the visual inspection and cleaning of the ECG recordings. Usual ECG de-noising approaches when compared to all mentioned, likewise band pass filtering, wavelet denoising, low pass filtering and adaptive filtering, over a broad range of ECG SNRs the study gives finer results. The EKF and EKS output is liable enough tracking the original ECG signal's irrespective of the noisiest period of the ECG signal is the result.