Discovering shared cardiovascular dynamics within a patient cohort.

Area of Science:

• Physiology
• Biomedical Engineering
• Systems Biology

Background:

• Cardiovascular variables like heart rate (HR) and blood pressure (BP) are tightly regulated by complex control systems.
• HR and BP time series display unique interaction dynamics influenced by physiological perturbations and pathological states.
• Identifying common dynamical patterns across heterogeneous patient populations remains a challenge.

Purpose of the Study:

• To develop and validate a technique for identifying shared dynamical patterns in cardiovascular time series.
• To capture nonlinear dynamics and non-Gaussian perturbations within cardiovascular regulation.
• To assess the applicability of the method across a cohort of individuals.

Main Methods:

• Utilized switching linear dynamical systems (SLDS) to model cardiovascular time series.
• Employed a mixture of linear dynamical systems with shared components across all subjects.
• Validated the technique using a simulated cardiovascular system and real-world data from a tilt-table test.

Main Results:

• The proposed SLDS technique successfully identified consistent dynamical patterns in HR and BP time series.
• The method demonstrated robustness in capturing complex cardiovascular system behaviors.
• Exploratory results confirmed the ability to detect shared dynamics across multiple recordings.

Conclusions:

• Switching linear dynamical systems offer a powerful approach for uncovering common cardiovascular regulatory patterns.
• This technique facilitates the analysis of physiological time series from diverse patient cohorts.
• The findings contribute to a deeper understanding of cardiovascular system dynamics and inter-individual variability.