Synthesizing imaging and computational models to link heterogeneous mechanical and ventilation dysfunction in asthma

The primary dysfunction in asthma is that the airways are hyper-reactive, which causes heterogeneous constriction throughout the airway tree. Mechanically, this causes an increase in the mean level and frequency dependence of lung resistance (RL) and elastance (EL). With regard to ventilation, imaging data reveal a heterogeneous and patchy distribution of ventilation. This dissertation synthesized ventilation distribution data derived from Positron Emission Tomographic (PET) images and dynamic RL and EL into a 3D computational airway tree model. The goals were (1) to identify the relative importance of small and large airways in the deterioration of both mechanical function and ventilation; (2) to examine whether the defects require narrowing of only airways leading to ventilation defects; and (3) to establish the link between ventilation heterogeneities and indices of mechanical heterogeneity.; Measures of PET, RL and EL were acquired in healthy and mild-to-moderate asthmatics before and after bronchoprovocation. We identified the largest size airways of the 3D model that, if constricted, could match the size and anatomical location of ventilation defects imaged by PET. We found that these airways had diameters mostly less than 0.44 mm. We then imposed constriction to the remaining airways and showed that matching mechanical impairment and PET ventilation defects requires either constriction of small airways alone, or simultaneous constriction of small and large airways, but not just large airways alone. We further found that airway narrowing throughout the lung must reach a critical level before ventilation defects are developed. Additional constriction adds to mechanical obstruction by increasing the extent of the ventilation defects. Finally, we also showed that measures of mechanical heterogeneity derived from RL reflect the heterogeneity in only the ventilating acini and are fairly insensitive to increases in the number of non-ventilating acini. Meanwhile, ventilation heterogeneity reflects the distribution of both the poorly and well ventilated acini.; In summary, we have advanced a new paradigm for synthesizing structural models of the lung with imaging and dynamic lung mechanics data. We generated new insights about crucial roles played by all airways during bronchoconstriction. Finally, these data and models explain better the relationship between mechanical and ventilation heterogeneity.