Sepsis insult generates different vascular response phenotypes: an unsupervised time-series clustering preclinical...

Area of Science:

• Cardiovascular Physiology
• Sepsis Pathophysiology
• Biomedical Engineering

Background:

• Sepsis is associated with significant vascular dysfunction.
• Early detection and severity assessment of sepsis are critical for effective treatment.
• Conventional hemodynamic parameters may not fully capture the complex cardiovascular changes in sepsis.

Purpose of the Study:

• To investigate the temporal evolution of cardiovascular response to sepsis.
• To evaluate novel arterial blood pressure (ABP) waveform-derived indices for sepsis detection and severity assessment.
• To combine conventional hemodynamic parameters with advanced cardiovascular markers.

Main Methods:

• Continuous ABP recording at aortic and femoral sites in pigs undergoing induced septic shock.
• Computation of standard beat-to-beat indices (mean ABP, heart rate, pulse pressure) and advanced markers (baroreflex sensitivity, characteristic time constant, PP amplification, harmonic distortion).
• Time series clustering using K-mean with soft-dynamic time warping on selected features (PP amplification, femoral characteristic time constant, aortic systolic ABP, femoral harmonic distortion).

Main Results:

• All animals exhibited severe cardiovascular decompensation during septic shock.
• Time series cluster analysis identified two distinct temporal patterns in cardiovascular variables.
• Shapelet analysis confirmed consistent, variable-specific trends within the identified clusters.

Conclusions:

• Novel ABP waveform-derived indices show promise for early sepsis identification and severity monitoring.
• These indices may support timely, personalized therapeutic strategies in sepsis management.
• Further research is needed to correlate these indices with treatment response and clinical outcomes.