An Enhanced Mechanistic Model For Capnography, With Application To CHF-COPD Discrimination

Area of Science:

• Pulmonary physiology and respiratory monitoring.
• Medical diagnostics and signal processing.
• Cardiopulmonary disease characterization.

Background:

• Capnography, a noninvasive monitoring tool, provides insights into lung ventilation-perfusion dynamics.
• Existing models effectively differentiate normal respiration from obstructive lung disease (COPD).
• Congestive heart failure (CHF) presents unique capnogram characteristics not fully captured by current models.

Purpose of the Study:

• To develop an enhanced mechanistic model of CO2 exhalation to capture CHF-specific capnogram features.
• To estimate respiratory parameters from capnograms using the enhanced model.
• To create a classification system distinguishing between CHF and COPD patients based on these parameters.

Main Methods:

• An enhanced mechanistic model incorporating lung fluid inertance was developed.
• Four model parameters were estimated breath-by-breath by fitting to measured capnograms.
• A quadratic discriminator was trained using parameters from CHF and COPD patients to classify conditions.

Main Results:

• The enhanced model successfully captured CHF-specific capnogram characteristics.
• The trained quadratic discriminator achieved an area under the ROC curve of 0.94 on the training set.
• In a test set, the discriminator correctly identified 87.5% of CHF and COPD patients.

Conclusions:

• The proposed enhanced mechanistic model and capnogram analysis can effectively distinguish between CHF and COPD.
• This approach offers a noninvasive, accurate method for differentiating these cardiopulmonary conditions.
• Capnography-derived parameters hold significant potential for clinical diagnosis of respiratory and cardiac diseases.