Functional Residual Capacity Predictions through Three Personalized Basis Functions in a Virtual Patient Model for VCV

The current approach to invasively mechanically ventilating a patient is generalized, and determining a patient-specific positive-end-expiratory-pressure (PEEP) is not standardized. This raises issues not only around the efficiency of ventilation but the safety of such. The inclusion of recruitment maneuvers with subsequent PEEP in mechanical ventilation has proven highly effective in recruiting lung volume and preventing alveolar collapse. The introduction of patient-specific, personalized monitoring enables a more appropriate delivery of ventilation that evolves as the patients condition does. In this study, function residual capacity has been analysed using hysteresis loop analysis (HLA) and three separate potential basis functions, Exponential (EXP), Parabolic (PARA) and Cumulative (CUMU). These basis function sets were compared based on their performance in predicting functional residual capacity (VFRC). Additional components of lung mechanics have been previously analysed and compared, however this particular study prioritized the accuracy of VFRC predictions. Data was provided from the McREM trial which spanned across 19 patients and 7 different baseline PEEP levels ranging from 6 cmH2O through 12 cmH2O. Up to 6 prediction steps were analysed from each baseline PEEP to determine the accuracy across a range of information yielding 623 cases. The results showed that all three basis function sets displayed the highest R2 values for cumulative prediction steps 1-6. All three had a final R2 of 0.84, however the PARA set showed higher R2 over each prediction step, yielding it the most efficient as the higher PEEP levels are the more clinically relevant ones for invasive mechanical ventilation (IMV).