“A Holistic Approach to Designing Nearly Linear-Phase IIR Filters"

Electrocardiogram (ECG) signals are widely used in the medical field to assess the heart's electrical activity. However, these signals often contain artefacts—undesired noise or disturbances—that can obscure accurate interpretation. These artefacts typically arise from sources unrelated to cardiac activity and must be mitigated or eliminated for reliable diagnosis. This study investigates various methods for designing Infinite Impulse Response (IIR) filters with near-linear phase characteristics to remove artefacts from ECG signals. Initially, conventional methods such as Butterworth, Chebyshev, Inverse Chebyshev, and Elliptic filter designs are employed to develop IIR filters. The ECG data for this research is sourced from the MIT-BIH Physionet Database. Subsequently, innovative techniques, including Minimum Phase Transformation, Frequency-Dependent Phase Compensation, Kautz Filtering, and Z-Transform, are applied to achieve nearly linear phase characteristics. The proposed approach is assessed based on Signal-to-Noise Ratio (SNR) and group delay performance. The results highlight the effectiveness of the proposed techniques in enhancing filter performance, reducing artefacts, and preserving the integrity of the ECG signals.