Modified multiscale fuzzy entropy: A robust method for short-term physiologic signals

Area of Science:

• Physiological signal analysis
• Complex systems science
• Biomedical engineering

Background:

• Multiscale entropy (MSE) is a key method for analyzing complex physiological signals.
• Existing variants like modified multiscale entropy (MMSE) have limitations with short time series due to hard similarity criteria.
• There is a need for improved entropy analysis methods for short physiological data.

Purpose of the Study:

• To introduce Modified Multiscale Fuzzy Entropy (MMFE) as an advancement over MMSE.
• To evaluate the robustness of MMFE and MMSE for short physiological time series.
• To compare MMFE and MMSE performance against classical MSE using varying signal lengths.

Main Methods:

• Development of the Modified Multiscale Fuzzy Entropy (MMFE) method, integrating fuzzy entropy concepts with MMSE.
• Evaluation of MMSE and MMFE robustness using segmented stochastic noise and heart rate variability (HRV) series.
• Comparison of MMFE and MMSE results with classical MSE on full signals.

Main Results:

• MMFE demonstrates significantly higher robustness than MMSE for short physiological time series.
• MMFE performance closely approximates classical MSE results for time series as short as 400 samples.
• An exponential relationship between the MMFE fuzzy parameter and signal size was identified.

Conclusions:

• MMFE is a more robust and reliable method for analyzing short physiological time series compared to MMSE.
• The identified exponential relationship provides a guideline for selecting optimal MMFE parameters based on signal length.
• MMFE offers a promising tool for complex physiological signal analysis when data is limited.