اولین همایش بین المللی تحول خدمات سلامت: تلفیق مراقبت، درمان، بازتوانی و امیدبخشی

رفتن به صفحه اصلی

ارائه مجازی مقاله

کد مقاله 1121

عنوان Multi-Stage Hybrid Filtering Framework for Robust ECG Denoising Using an Unbiased Adaptive Model

نویسندگان Alireza Zarei - Ebrahim Daneshifar

چکیده Background: Electrocardiogram (ECG) signals are essential for cardiovascular diagnostics but are frequently corrupted by noise (e.g., powerline interference, baseline wander, and muscle artifacts), compromising clinical accuracy. This study proposes a novel hybrid denoising framework to enhance ECG quality for modern medical equipment. Methods: The proposed multi-stage algorithm integrates baseline wander correction, a two-stage moving average filter, an IIR comb filter, and the Steepest Descent adaptive algorithm. Performance was rigorously evaluated using standard records from the MIT-BIH Arrhythmia Database and compared with conventional methods. Findings: The hybrid method effectively suppressed various artifacts while strictly preserving critical morphological features, particularly the QRS complex. It achieved an outstanding Signal-to-Noise Ratio (SNR) of 30.03 dB and a Mean Squared Error (MSE) of 0.0062, significantly outperforming existing standard filters. Conclusion: By delivering high-fidelity noise reduction without distorting vital diagnostic markers, this adaptive hybrid algorithm offers a robust solution for smart healthcare systems, ambulatory Holter monitors, and ICU bedside equipment, thereby enhancing automated diagnostic reliability.

فایل ها

فایل پوستر مقاله

محتوای ارائه مجازی مقاله

نظرات