ECG signal enhancement using adaptive Kalman filter and signal averaging

M.H. Moradi, M. Ashoori Rad, R. Baghbani Khezerloo
2014 International Journal of Cardiology  
The electrocardiogram (ECG) is widely used for diagnosis of heart diseases. Good quality ECG signals are utilized by physicians for interpretation and identification of physiological and pathological phenomena. Several techniques have been proposed to extract the ECG components contaminated with the background noise and allow the measurement of subtle features in the ECG signal. One of the common approaches is the adaptive filter architecture, which has been used for the noise cancelation of
more » ... s containing baseline wander, electromyogram (EMG) noise, and motion artifacts [1] . Statistical techniques such as principal component analysis [2] , independent component analysis [3] , and neural networks [4] have also been used to extract a noise free signal from the noisy ECG. Over the past several years, methods based on the wavelet transform (WT) have also received a great deal of attention for denoising of signals that possess multi-resolution characteristics such as the ECG [5] . Some of the noise and artifact problems that arise during these recordings can be suppressed by simple, frequency-selective filtering. However, due to the partial overlap of signal and noise bandwidths, this frequency-selective filtering can only help to some extent. We developed a filter that can do exactly this. This filter is derived using a Bayesian framework and constitutes a Kalman filter in which the dynamic variations in the ECG are modeled by a covariance matrix that is adaptively estimated every time new data arrive. Typically, ECG complexes that originate from consecutive heartbeats are very similar but not identical. Moreover, when recording the ECG, the signals are corrupted to some extent by noise and artifacts. In a simplified form, both the relation between consecutive ECG complexes and the corruption of the recorded ECG can be described by means of a state space model (Fig. 1 ) as follows: where, X k denotes the ECG complex for heartbeat k and Y k denotes the recorded signal. The evolution of the ECG complexes between heartbeats is modeled by the stochastic component V k (also referred to as the process noise). The measurement noise, i.e., corrupting signals, such as electromyogram signals, movement artifacts, and interferences from the power line grid, is represented by the vector W k . In the state-space description of Eq. (1), the problem of enhancing the SNR of the ECG is reduced to the problem of sequentially estimating the model parameter vector X k and the noise covariance Σ k and Λ k . Here, sequential estimation refers to the estimation of the relevant parameters based on the earlier estimate and all newly arriving data. With the adopted dipole model of the heart's electrical activity, it can be argued that dynamical variations in the ECG morphology are reflected in all recorded ECG signals Y. Analogously, measurement noise W that does not exhibit the same spatial correlation as the ECG is suppressed in the linear combination of ECG signals. As a result, the measurement noise vector W i for ith ECG signal can be approximated byŴ i using the estimateŶ i ¼ Y −i γ as follows: This also yields an estimation for measurement noise covariance Σ. The uncertainty in the state-space model of Eq. (1) and in the associated noise parameters suggests the use of a probabilistic approach for solving the parameter estimation problem [6] . In addition, the sequential nature of the estimation problem motivates the use of a Bayesian framework in which the prior probability distribution assigned to the unknown parameters is updated every time new data arrive. Here, again, sequential refers to the estimation of model parameters based on earlier parameter estimates and on newly arriving data.
doi:10.1016/j.ijcard.2014.03.128 pmid:24717324 fatcat:hludceuwnfctjo5hirkw3sa2dy