AIRWAY GEOMETRY AND WALL MECHANICAL-PROPERTIES ESTIMATED FROM SUBGLOTTAL INPUT IMPEDANCE IN HUMANS

We measured input impedance between 16 and 2,048 Hz in intubated subjects at functional residual capacity. The corresponding subglottal impedances (Z(SG)) were then computed using a model where the endotracheal tube was represented by a distributed-parameter two-port network. Z(SG) was well described by a model based on Horsfield's asymmetric airway geometry at total lung capacity (TLC) with nonrigid walls. The walls of the cartilaginous airways included separate cartilage and soft tissue compartments, whereas the noncartilaginous airway walls had only a soft tissue compartment. Both compartments consisted of a series resistance, inertance, and compliance, the values of which were computed from airway dimensions and wall material properties (viscosity, density, and Young's modulus). Airway wall thickness was determined by scaling an airway wall area-diameter relationship. Airway lengths and diameters were scaled from the Horsfield TLC values by a single factor and by an order-dependent sigmoidal curve, respectively. The estimated soft tissue viscosity and Young's modulus were 1.04 +/- 0.21 cmH(2)O.s and 593 +/- 319 cmH(2)O, respectively. Airway lengths and tracheal diameters were not statistically different from the Horsfield values. The estimated diameters of the more peripheral airways were significantly reduced compared with the Horsfield TLC values (e.g., similar to 40% at the terminal airway), which is consistent with the reduction in airway caliber when the lung deflates from TLC to functional residual capacity. These results indicate that high-frequency Z(SG) is sensitive to subglottal airway geometry and wall properties and that by use of appropriate structural models one can estimate airway geometry and airway wall parameters.