Performance Evaluation and Implementation of FPGA Based SGSF in Smart Diagnostic Applications

Area of Science:

• Biomedical Engineering
• Signal Processing
• Digital Electronics

Background:

• Electroencephalogram (EEG) and Electrocardiogram (ECG) signals contain critical information in their peaks.
• Noise and artifacts in EEG/ECG signals can obscure vital diagnostic information.
• Effective signal pre-processing is crucial for accurate real-time smart medical diagnostic systems.

Purpose of the Study:

• To implement the Savitzky Golay Smoothing Filter (SGSF) for pre-processing real-time EEG and ECG signals.
• To evaluate the noise rejection efficiency and shape-preserving capabilities of SGSF.
• To compare SGSF performance against the Moving Average Filter (MAF) and implement it on reconfigurable architectures.

Main Methods:

• Implemented the Savitzky Golay Smoothing Filter (SGSF).
• Tested SGSF on synthetic EEG and ECG signals with varying noise levels.
• Compared SGSF with Moving Average Filter (MAF) using parameters like SNR, SSNR, SNRI, MSE, COR, and signal distortion.
• Realized SGSF on a Field-Programmable Gate Array (FPGA) platform for real-time applications.

Main Results:

• SGSF demonstrated efficient noise rejection with minimal signal distortion.
• SGSF exhibited superior smoothing performance compared to MAF.
• The FPGA-SGSF implementation significantly reduced processing time.
• The proposed methodology preserved essential signal features in real-time EEG and ECG data.

Conclusions:

• SGSF is a highly effective filter for pre-processing EEG and ECG signals in smart medical diagnostics.
• SGSF offers advantages over MAF in terms of efficiency and signal preservation.
• FPGA-based SGSF implementation is suitable for high-speed, low-power, real-time medical systems.