Nonlinear identification algorithm for online and offline study of pulmonary mechanical ventilation

TL;DR

This work tackles the challenge of accurately characterizing a patient’s respiratory mechanics under assisted ventilation by estimating nonlinear per-cycle parameters. It develops a grey-box identification framework that first fits a linear model to initialise a nonlinear quadratic description $P_c(V)=a_1V+a_2V^2$, and then refines the nonlinear parameters per respiratory cycle using mouth pressure $P_v$ and flow $F$ data. Through simulations with sigmoidal and hysteretic P–V curves and real data from sedated patients (including COVID-19 cases), the nonlinear approach yields closer fits than traditional linear methods, particularly outside the linear region, while providing clinically useful indicators such as P–V curvature and region classification (Atelectasis, Linear, Overdistension). This enhanced information supports safer ventilation, e.g., by steering operation toward the linear, best-compliance region and enabling online monitoring of regional lung status. The method demonstrates potential for real-time bedside use and offline validation, with future work aiming to validate on larger patient cohorts and correlate estimated parameters with clinical outcomes.

Abstract

This work presents an algorithm for determining the parameters of a nonlinear dynamic model of the respiratory system in patients undergoing assisted ventilation. Using the pressure and flow signals measured at the mouth, the model's quadratic pressure-volume characteristic is fit to this data in each respiratory cycle by appropriate estimates of the model parameters. Parameter changes during ventilation can thus also be detected. The algorithm is first refined and assessed using data derived from simulated patients represented through a sigmoidal pressure-volume characteristic with hysteresis. As satisfactory results are achieved with the simulated data, the algorithm is evaluated with real data obtained from actual patients undergoing assisted ventilation. The proposed nonlinear dynamic model and associated parameter estimation algorithm yield closer fits than the static linear models computed by respiratory machines, with only a minor increase in computation. They also provide more information to the physician, such as the pressure-volume (P-V) curvature and the condition of the lung (whether normal, under-inflated, or over-inflated). This information can be used to provide safer ventilation for patients, for instance by ventilating them in the linear region of the respiratory system.

Figures (13)

• Figure 1: Characterisation of volumes in the system.
• Figure 2: In dashed lines, the Q-S P-V curve measured from FRC to TLC is shown. In solid lines, illustrative respiratory cycles curves starting from different equilibrium points $Eq_i$, $i=1,2,3$, showing the various curve shapes that can be observed depending on the region where the patient is ventilated. $P_A$ is the alveolar pressure, $V_{L}$ is the total volume and $V_{PEEP_i}$ is the volume due to the ventilator $PEEP_i$.
• Figure 3: Hysteresis in the lungs-chest wall system in a typical respiratory cycle of a patient. The arrow pointing upwards corresponds to the inspiration path, and the arrow pointing downwards to the expiration path.
• Figure 4: Electrical equivalent model of the respiratory system.
• Figure 5: Proposed algorithm flow chart.