Optimized ECG Processing System using FPGA with Polyphase Filter for Arrhythmia Classification

This study introduces a high efficiency ECG signal processing system that leverages a specialized filtering technique integrated with feature extraction and classification mechanisms for enhanced cardiac monitoring. Utilizing a polyphase Finite Impulse Response (FIR) filter structure optimized via Distributed Arithmetic (DA), the design significantly reduces noise and computational load. ECG data sourced from the MIT-BIH Arrhythmia Database is processed to identify key features, particularly the QRS complex. A Support Vector Machine (SVM) classifier is employed to detect arrhythmias using these features. The system implementation uses 1093 Look-Up Tables (LUTs) and consumes 25.89 W of power, demonstrating notable advantages in low power consumption and reduced hardware usage. Experimental results validate the system's efficiency, establishing it as a viable candidate for advanced ECG analysis in modern healthcare systems.