Benefits of Zero-Phase or Linear Phase Filters to Design Multiscale Entropy: Theory and Application

Area of Science:

• Signal Processing
• Complexity Science
• Biomedical Engineering

Background:

• Multiscale entropy (MSE) is crucial for signal complexity analysis and classification.
• Traditional MSE uses coarse-graining via filtering and decimation, which can introduce distortions.

Purpose of the Study:

• To develop a new variant of MSE that minimizes distortions introduced during the coarse-graining process.
• To improve the accuracy and reliability of complexity measures derived from signals.

Main Methods:

• A novel coarse-graining method is proposed, employing linear-phase or null-phase low-pass filters before decimation.
• The modified coarse-grained sequences are used to calculate a new MSE variant.
• The method's behavior is illustrated using simulated signals (sinusoids, noises) and applied to attentional tunneling detection.

Main Results:

• The new MSE variant effectively reduces distortions associated with decimation in traditional MSE.
• Illustrations show distinct behaviors of the new MSE across different signal types.
• Application to attentional tunneling detection demonstrated significant improvement in differentiating states (lower p-values).

Conclusions:

• The proposed MSE variant offers a more accurate characterization of signal complexity.
• This method enhances the sensitivity for detecting subtle changes in signal dynamics, such as in cognitive states.
• The improved coarse-graining technique is applicable to MSE and other entropy variants.