Development of adaptive filtering and artifact removal algorithms for cardiac signals in electrocardiograph and holter monitoring systems

Abstract

This study presents the development of adaptive filtering and artifact removal algorithms for cardiac signal processing in electrocardiograph and Holter monitoring systems. Electrocardiography is one of the most important non-invasive diagnostic methods for evaluating cardiac electrical activity, detecting rhythm abnormalities, identifying ischemic changes, and monitoring long-term heart function. However, ECG signals acquired in clinical and ambulatory environments are often affected by different types of noise and artifacts, including baseline wander, power-line interference, muscle activity noise, electrode motion artifacts, and motion-related disturbances. These interferences can distort the morphology of the P wave, QRS complex, ST segment, and T wave, thereby reducing diagnostic reliability. To address these challenges, this study proposes a multi-stage adaptive signal processing framework for ECG and Holter systems. The proposed methodology includes raw ECG acquisition, signal quality assessment, artifact classification, adaptive baseline wander suppression, power-line interference reduction, motion artifact cancellation, muscle noise attenuation, and final reconstruction of the clinically relevant ECG waveform. Adaptive algorithms such as LMS, NLMS, RLS, wavelet-based denoising, and hybrid deep learning-supported approaches are considered within the proposed framework. The results demonstrate that adaptive filtering methods can significantly improve ECG signal quality while preserving diagnostically important waveform components. In Holter monitoring systems, where signals are recorded continuously over long periods under variable motion conditions, adaptive and hybrid artifact removal approaches are especially important. The proposed algorithmic framework can improve diagnostic accuracy, support automated arrhythmia detection, and enhance the reliability of long-term cardiac monitoring systems.