Multiscale Entropy Analysis of Short Signals: The Robustness of Fuzzy Entropy-Based Variants Compared to Full-Length...

Area of Science:

• Physiology
• Biomedical Engineering
• Data Science

Background:

• Multiscale entropy (MSE) analysis quantifies time series complexity across temporal scales.
• Traditional MSE may lack accuracy for short time series, necessitating algorithm adaptations.
• Existing MSE variations for short series lack systematic reliability comparisons.

Purpose of the Study:

• To systematically compare the reliability of different multiscale entropy (MSE) algorithm variations adapted for short time series.
• To evaluate the accuracy of fuzzy and non-fuzzy MSE algorithms using human and rat heart rate variability (HRV) data.
• To determine the optimal fuzzy exponents for improved MSE accuracy based on time series length.

Main Methods:

• Comparison of MSE algorithm variations: composite MSE (CMSE), refined composite MSE (RCMSE), modified MSE (MMSE), and their fuzzy counterparts.
• Analysis of MSE estimation errors across various fuzzy exponents.
• Validation using human and rat heart rate variability (HRV) time series, with long-term MSE as the reference.

Main Results:

• Fuzzy MSE versions demonstrated minimal estimation errors, particularly as a function of time series length.
• The traditional multiscale entropy with fuzzy counting (MFE) achieved accuracy comparable to other algorithms but with superior computational efficiency.
• Optimal fuzzy exponent selection is dependent on the specific time series length for best accuracy.

Conclusions:

• Fuzzy MSE algorithms offer improved accuracy and reliability for analyzing short time series, such as HRV.
• MFE presents a computationally efficient and accurate alternative for short-term complexity analysis.
• Tailoring fuzzy exponents to time series length is crucial for maximizing the precision of MSE estimations.