Outlier-resilient complexity analysis of heartbeat dynamics

Area of Science:

• Physiological complexity
• Nonlinear dynamics
• Biomedical signal processing

Background:

• Physiological complexity is a hallmark of healthy control.
• Quantifying complexity in physiological signals with outliers is challenging for clinical application.
• Nonlinear dynamic theory offers novel insights into physiological control.

Purpose of the Study:

• To develop a robust method for estimating signal complexity resilient to outliers.
• To apply the method to human heartbeat recordings for clinical relevance.
• To assess the method's ability to detect cardiac dysfunction and predict mortality.

Main Methods:

• Analyzing the irregularity of sign time series from coarse-grained time series at various scales.
• Utilizing surrogate data to validate robustness against noise and outliers.
• Applying the method to heartbeat recordings from healthy individuals, heart failure patients, and critically ill ECMO patients.

Main Results:

• The proposed method reliably assesses complexity in noisy data and is resilient to outliers.
• Cardiac control degradation was detected in heart failure and ECMO patients without outlier removal.
• Complexity measures predicted mortality in ECMO patients.

Conclusions:

• The new method provides a reliable way to quantify physiological signal complexity, even with outliers.
• This approach can aid in monitoring cardiac function and predicting outcomes in clinical settings.
• The findings support the clinical utility of nonlinear dynamic theory in patient care.