Abstract
The advancement in the field of information and communication technology explores new paths for healthcare and E-health systems. Generally, the electrocardiographic (ECG) signals are contaminated with different impairments which impose difficulty to analysis and diagnosis the heart diseases. In contrast to the traditional artifacts removal techniques such as wavelet transform, and neural networks, this paper proposes a variable step size-least mean squared (VSS-LMS) algorithm for mitigating noise and cyber attacks in ECG signals. The proposed algorithm is obtained considering tuning parameters and errors, where the step size is adaptively updated based on the signal dynamics. The performances of the proposed approach is demonstrated by using MIT-BIH ECG signals database. It shows that the proposed method can properly remove noises from ECG signals.