Correntropy-based nonlinearity test applied to patients with chronic heart failure

Area of Science:

• Biomedical Engineering
• Nonlinear Dynamics
• Respiratory Physiology

Background:

• Chronic heart failure (CHF) is associated with altered respiratory patterns.
• Detecting nonlinearities in breathing can indicate disease severity or risk.
• Current methods may not fully capture complex respiratory dynamics in CHF.

Purpose of the Study:

• To introduce the correntropy function as a novel measure for detecting nonlinearities in respiratory patterns of CHF patients.
• To differentiate between periodic breathing (PB) and nonperiodic breathing (nPB) in CHF using correntropy.
• To assess the relationship between respiratory complexity, CHF risk, and breathing patterns.

Main Methods:

• Utilized the correntropy function to analyze respiratory data from CHF patients.
• Employed surrogate data methods, specifically iterative refined amplitude adjusted Fourier transform (IAAFT), for statistical testing.
• Tested the null hypothesis that respiratory data follows a Gaussian linear stochastic process.
• Calculated correntropy spectral density (CSD) and derived parameters, including the ratio R (modulation/respiratory frequency power bands).

Main Results:

• Correntropy successfully identified nonlinearities in respiratory patterns.
• Reduced complexity was observed in CHF patients with higher risk levels.
• The ratio R was significantly different in nPB patients compared to PB patients, indicating greater nonlinearity in nPB.
• No significant differences in correntropy-derived parameters were found using IAAFT surrogate data.

Conclusions:

• Correntropy is a suitable measure for preserving nonlinear information in respiratory signals.
• Nonperiodic breathing in CHF patients exhibits significantly higher nonlinearities than periodic breathing.
• Correntropy analysis offers a promising approach for characterizing respiratory dynamics and potentially assessing risk in CHF.