A novel adaptive control system for noisy pressure-controlled ventilation: a numerical simulation and bench test study.

Abstract

There is growing interest in the use of both variable and pressure-controlled ventilation (PCV). The combination of these approaches as "noisy PCV" requires adaptation of the mechanical ventilator to the respiratory system mechanics. Thus, we developed and evaluated a new control system based on the least-mean-squares adaptive approach, which automatically and continuously adjusts the driving pressure during PCV to achieve the desired variability pattern of tidal volume (V (T)). The controller was tested during numerical simulations and with a physical model reproducing the mechanical properties of the respiratory system. We applied step changes in respiratory system mechanics and mechanical ventilation settings. The time needed to converge to the desired V (T) variability pattern after each change (t (c)) and the difference in minute ventilation between the measured and target pattern of V (T) (DeltaMV) were determined. During numerical simulations, the control system for noisy PCV achieved the desired variable V (T) pattern in less than 30 respiratory cycles, with limited influence of the dynamic elastance (E*) on t (c), except when E* was underestimated by >25%. We also found that, during tests in the physical model, the control system converged in <60 respiratory cycles and was not influenced by airways resistance. In all measurements, the absolute value of DeltaMV was <25%. The new control system for noisy PCV can prove useful for controlled mechanical ventilation in the intensive care unit.