Monitoring of total positive end-expiratory pressure during mechanical ventilation by artificial neural networks.

Ventilation treatment of acute lung injury (ALI) requires the application of positive airway pressure at the end of expiration (PEEPapp) to avoid lung collapse. However, the total pressure exerted on the alveolar walls (PEEPtot) is the sum of PEEPapp and intrinsic PEEP (PEEPi), a hidden component. To measure PEEPtot, ventilation must be discontinued with an end-expiratory hold maneuver (EEHM). We hypothesized that artificial neural networks (ANN) could estimate the PEEPtot from flow and pressure tracings during ongoing mechanical ventilation. Ten pigs were mechanically ventilated, and the time constant of their respiratory system (τRS) was measured. We shortened their expiratory time (TE) according to multiples of τRS, obtaining different respiratory patterns (Rpat). Pressure (PAW) and flow (V'AW) at the airway opening during ongoing mechanical ventilation were simultaneously recorded, with and without the addition of external resistance. The last breath of each Rpat included an EEHM, which was used to compute the reference PEEPtot. The entire protocol was repeated after the induction of ALI with i.v. injection of oleic acid, and 382 tracings were obtained. The ANN had to extract the PEEPtot, from the tracings without an EEHM. ANN agreement with reference PEEPtot was assessed with the Bland-Altman method. Bland Altman analysis of estimation error by ANN showed -0.40 ± 2.84 (expressed as bias ± precision) and ±5.58 as limits of agreement (data expressed as cmH2O). The ANNs estimated the PEEPtot well at different levels of PEEPapp under dynamic conditions, opening up new possibilities in monitoring PEEPi in critically ill patients who require ventilator treatment.