Bufferless Compression of Asynchronously Sampled ECG Signals in Cubic Hermitian Vector Space

Area of Science:

• Biomedical Engineering
• Signal Processing
• Data Compression

Background:

• Asynchronous level crossing sampling analog-to-digital converters (ADCs) offer energy efficiency but face challenges with increasing data rates at lower thresholds.
• Existing compression methods may negate the energy benefits of asynchronous ADCs due to data buffering requirements.

Purpose of the Study:

• To introduce a computationally efficient, online data compression technique for asynchronously sampled electrocardiogram (ECG) signals.
• To maintain the energy efficiency and timing accuracy advantages of asynchronous ADCs during signal compression.

Main Methods:

• A cubic Hermitian vector-based technique for online compression of asynchronously sampled ECG signals.
• The algorithm operates with O(n) complexity and requires no data buffering.
• Signal boundary parameters are extracted a priori to ensure computational efficiency.

Main Results:

• Achieved compression ratios of up to 90% for ECG signals.
• Obtained Percentage Root Mean Square Difference (PRD) ratios as low as 0.97, indicating high signal fidelity.
• Preserved the superior feature-to-feature timing accuracy inherent in asynchronously sampled data.

Conclusions:

• The proposed cubic Hermitian vector-based compression method is well-suited for asynchronous ADCs.
• The technique effectively reduces data rates and energy consumption while maintaining signal integrity and timing accuracy.
• This approach enables the practical application of energy-efficient asynchronous ECG acquisition systems.