FPGA-core defibrillator using wavelet-fuzzy ECG arrhythmia classification

Area of Science:

• Biomedical Engineering
• Digital Signal Processing
• Computer Engineering

Background:

• Electrocardiogram (ECG) analysis is crucial for diagnosing cardiac conditions.
• Automated systems are needed for efficient and real-time arrhythmia detection.
• Wavelet transform and fuzzy logic offer robust methods for ECG feature extraction and classification.

Purpose of the Study:

• To develop and evaluate an ECG feature extraction and classification system using Field Programmable Gate Array (FPGA).
• To detect three types of cardiac arrhythmias and abnormalities in real-time.
• To compare the performance of an FPGA implementation with a microcontroller-based system.

Main Methods:

• ECG beats were analyzed in the wavelet domain for QRS complex detection and wave peak localization.
• A fuzzy classifier was employed to classify ECG beats based on extracted features.
• The algorithm was implemented on an FPGA (Altera) and an AVR microcontroller, with signal input via a custom software interface.

Main Results:

• The developed FPGA prototype successfully performed real-time ECG monitoring for arrhythmia detection.
• The FPGA achieved significantly higher computational speed compared to the microcontroller due to parallel processing capabilities.
• Resource usage was higher on the FPGA, but it provided superior processing performance.

Conclusions:

• FPGA implementation enables high-speed, real-time ECG signal processing for arrhythmia detection.
• The proposed system demonstrates the feasibility of using FPGAs for advanced cardiac monitoring.
• FPGA offers a speed advantage over microcontrollers for complex signal processing tasks at the cost of increased resource utilization.