Low-Order Mechanistic Models for Volumetric and Temporal Capnography: Development, Validation, and Application.

Area of Science:

• Respiratory Physiology
• Computational Modeling
• Medical Diagnostics

Background:

• Capnography, measuring exhaled CO2 over time (Tcap) or volume (Vcap), provides insights into cardiorespiratory function.
• Existing methods for analyzing capnograms often lack detailed mechanistic underpinnings.
• Accurate, quantitative models are needed to extract maximal information from capnograms.

Purpose of the Study:

• To develop low-order mechanistic models that quantitatively analyze capnograms.
• To enable robust estimation of physiological parameters from capnogram data.
• To explore the potential of these models for diagnosing respiratory and cardiac conditions.

Main Methods:

• Developed mechanistic models incorporating alveolar mechanics and airway gas mixing.
• Fitted models breath-by-breath to capnogram data (Tcap and Vcap) from ventilated patients.
• Estimated model parameters characterizing the capnogram and airflow.

Main Results:

• Models demonstrated close fits to measured capnograms (mean RMSE < 2% of end-tidal CO2).
• Robust estimation of subject-specific physiological parameters was achieved.
• Tcap-based analysis showed potential for discriminating COPD from CHF with high accuracy (80.6%) and AUROC (0.84).

Conclusions:

• Mechanistic capnogram models provide accurate and quantitative analysis of exhaled CO2.
• Subject-specific parameter estimation enhances diagnostic capabilities.
• These models offer a novel, non-invasive approach for cardiorespiratory assessment and disease differentiation.