Reconstruction of noisy Electrocardiograms by approximation using Lagrange form of Hermite interpolating polynomial with Chebyshev nodes

• Published in: NeuroQuantology Vol. 20; no. 10; p. 7830
• Main Authors: Chouhan, Vandana, Ray, Shashwati
• Format: Journal Article
• Language: English
• Published: Bornova Izmir NeuroQuantology 01.01.2022
• Subjects: Algorithms; Amplitudes; Chebyshev approximation; Correlation coefficients; Diagnosis; Electrocardiography; Hermite polynomials; Morphology; Nodes; Noise reduction
• ISSN: 1303-5150
• DOI: 10.14704/nq.2022.20.10.NQ55771

Elecrocardiograms (ECG) are recording of heart’s electrical activity and most common test for ambulatory and intensive care unit. Medical conditions such as diabetes, arrhythmias, ectopy, electrolyte imbalance, conduction block, and fibrillation also have impact on ECG. Real time ECG is acquired using surface electrodes placed on patients skin, during acquisition they are often contaminated by various types of noise originating from external source or from physiological process of human body. This noise will distort time domain morphological features such as position, interval, amplitude and slope of complex and segment which are important for diagnosis. Efficacy of computer assisted cardiac diagnosis greatly relies on accurate estimation of amplitude and duration of morphologies; hence it is important that ECG should be denoised. In this paper Lagrange form of Hermite interpolating polynomial with chebyshev nodes is proposed for denoising ECG signal. Proposed algorithm is applied on ECG signal from MIT/BIH arrhythmia database contaminated with different level of noise. The proposed algorithm is compared in terms of RMS, PRDN, and correlation coefficient. On the basis of objective assessment of reconstructed ECG signal the proposed technique outperforms other existing denoising algorithm and together with noise reduction it retains the details of morphological features for clinical use