Real-Time Estimation of Lung Model Parameters and Breathing Effort During Assisted Ventilation

摘要

The estimation of lung mechanics’ parameters and the patient’s residual volitional breathing effort is a prerequisite to adjust the parameters of assisted ventilation in a patient-individual manner. A real-time capable approach is investigated that estimates the resistance and compliance of a first-order lung model in conjunction with the intrapleural pressure in real-time. Latter is a measure for the patient’s breathing effort. A signal generator model in the form of a Radial Basis Function (RBF) network is assumed for the intrapleural pressure. The Gaussian basis functions are periodic with the breathing cycle duration. This approach does not restrict the signal form of the patient-driven pressure curve. Recursive Least Squares (RLS) with selective forgetting is employed to consider the different dynamics of the estimated model parameters. A time-discrete version of the lung model is used for RLS. Computer simulations reveal that the approach is feasible and that selective forgetting is necessary to obtain satisfactory estimation results.