Are Strategies Favoring Pattern Matching a Viable Way to Improve Complexity Estimation Based on Sample Entropy?

Area of Science:

• Physiology
• Complexity Science
• Biomedical Engineering

Background:

• Complexity estimation from physiological signals is crucial for understanding health and disease.
• Traditional methods like sample entropy (SampEn) can be limited by short data series.
• Improving pattern matching rate without increasing series length is a proposed strategy to enhance complexity estimation.

Purpose of the Study:

• To test the hypothesis that increasing pattern matching rate improves complexity estimation for short physiological time series.
• To compare novel pattern matching strategies with traditional SampEn in simulations and real-world physiological data.
• To evaluate the utility of these methods in distinguishing between Parkinson's disease patients and healthy controls.

Main Methods:

• Simulations of nonlinear deterministic and linear stochastic dynamics with varying noise levels.
• Application of signal transformations to increase pattern matching rate.
• Comparison of novel methods with standard SampEn for complexity estimation.
• Analysis of beat-to-beat heart period (HP) and systolic arterial pressure (SAP) variability in Parkinson's disease patients and healthy controls.

Main Results:

• Simulations showed strategies correctly estimated higher complexity in deterministic signals and greater regularity in stochastic or noisy signals.
• However, these novel techniques provided no practical advantage over traditional SampEn for short HP and SAP series.
• The ability to discriminate between groups and experimental conditions remained unchanged compared to SampEn.

Conclusions:

• Artificial increases in pattern matching rate do not offer methodological or practical value for complexity index assessment in short physiological series.
• The findings suggest that traditional SampEn remains a suitable method for analyzing complexity in short-term cardiac and vascular control variability.
• Further research may be needed to explore alternative or complementary methods for complexity analysis in limited data scenarios.