Signal conditioning with almost no multiplications via filter sharpening

Area of Science:

• Biomedical Engineering
• Digital Signal Processing

Background:

• Signal conditioning often requires computationally intensive filters.
• Existing methods may struggle to meet stringent performance constraints, especially with fixed-point arithmetic.

Purpose of the Study:

• To introduce and evaluate "Filter Sharpening" for efficient, high-quality signal conditioning.
• To demonstrate the application of Filter Sharpening in removing electrocardiogram baseline wander.

Main Methods:

• Developing a filter network using computationally simple, repeatable filter structures.
• Implementing the filter network with fixed-point arithmetic and minimal add/subtract operations.
• Applying a highpass filter with a 0.25 Hz corner frequency for baseline wander removal.

Main Results:

• The Filter Sharpening method enables high-quality signal conditioning with minimal computational cost.
• Achieved over 40 dB attenuation for highpass filtering.
• The proposed filter meets the stringent performance requirements of the EC-13 medical standard.

Conclusions:

• Filter Sharpening offers an efficient and effective approach for signal conditioning, particularly in resource-constrained environments.
• This technique is well-suited for medical signal processing applications, such as removing baseline wander from ECG data.