Effect of nasal airway nonlinearities on oscillometric resistance measurements in infants
Résumé
Oscillometric measurements of respiratory system resistance (R-rs) in infants are usually made via the nasal pathways, which not only significantly contribute to overall R-rs but also introduce marked flow (V')-dependent changes. We employed intrabreath oscillometry in casts of the upper airways constructed from head CT images of 46 infants. We examined oscillometric nasal resistance (R-n) in upper airway casts with no respiratory flow (R-0) and the effect of varying V' on R-n by simulating tidal breathing. A characteristic nonlinear relationship was found between R-n and V', exhibiting segmental linearity and a prominent breakpoint (V'(bp)) after log-log transformation. V'(bp) was linearly related to the preceding value of end-expiratory volume acceleration (V `'(eE); on average r(2) = 0.96, P < 0.001). R-n depended on V', and R at end-expiration (R-eE) showed a strong dependence on V `'(eE) in every cast (r(2) = 0.994, P < 001) with considerable interindividual variability. The intercept of the linear regression of R-eE versus V `'(eE) was found to be a close estimate of R-0. These findings were utilized in reanalyzed R-rs data acquired in vivo in a small group of infants (n = 15). Using a graphical method to estimate R-0 from R-eE, we found a relative contribution of V'-dependent nonlinearity to total resistance of up to 33%. In conclusion, we propose a method for correcting the acceleration-dependent nonlinearity error in R-eE. This correction can be adapted to estimate R-0 from a single intrabreath oscillometric measurement, which would reduce the masking effects of the upper airways on the changes in the intrathoracic resistance. NEW & NOTEWORTHY Oscillometric measurements of respiratory system resistance (R-rs) in infants are usually made via the nasal pathways, which not only significantly contribute to overall R-rs but also introduce marked flow acceleration-dependent distortions. Here, we propose a method for correcting flow acceleration-dependent nonlinearity error based on in vitro measurements in 3D-printed upper airway casts of infants as well as in vivo measurements. This correction can be adapted to estimate R-rs from a single intrabreath oscillometric measurement.