Quality assessment of ECG compression techniques using a wavelet-based diagnostic measure.

Area of Science:

• Biomedical Engineering
• Signal Processing
• Medical Informatics

Background:

• High-quality Electrocardiograph (ECG) transmission faces challenges due to large data requirements and bandwidth limitations.
• Existing ECG compression techniques often rely on inaccurate or random error measures for performance evaluation.
• Advancements in hardware and software drive the development of novel ECG compression methods for higher compression rates.

Purpose of the Study:

• To introduce a novel wavelet-based quality measure for evaluating ECG compression techniques.
• To address the limitations of existing quality measures that are inaccurate or target random error behavior.
• To provide a more reliable method for assessing the quality of compressed ECG signals.

Main Methods:

• A new wavelet-based quality measure is proposed, decomposing ECG segments into frequency bands.
• A weighted scoring system is applied to each frequency band based on its dynamic range and diagnostic significance.
• Quantitative and qualitative performance evaluations were conducted to assess the proposed measure.

Main Results:

• The proposed wavelet-based quality measure demonstrates insensitivity to error variations.
• The new measure exhibits high accuracy in evaluating ECG compression performance.
• Results show a strong correlation between the proposed measure and subjective quality tests.

Conclusions:

• The developed wavelet-based quality measure offers a significant improvement over existing methods for ECG compression evaluation.
• This novel approach provides a reliable and accurate assessment of compressed ECG signal quality.
• The measure's robustness and correlation with subjective tests support its utility in clinical and research settings.