UVM Theses and Dissertations
Format:
Print
Author:
Schwartz, Benjamin Lincoln
Dept./Program:
Biomedical Engineering Program
Year:
2009
Degree:
M.S.
Abstract:
Bronchoconstriction is an inherently heterogeneous process. This has been modeled in a parallel fashion in mice. Modeling human bronchoconstriction in a serial arrangement has shown to be appropriate under some conditions. In this study we developed and fit three models each to experimental representations of three different lung pathologies. The first model, the Single Elastic Airway, holds the airways to be compliant, mechanically in parallel with the parenchymal tissue. They are modeled as a compartment with a finite elastance, in parallel with a viscoelastic tissue compartment. The second model, the Branching Elastic Airway, is an extension of the first model. The airways are assumed to be compliant and are modeled as branching compartments that have the same finite elastance. The branches terminate in identical viscoelastic tissue compartments.
The third model, the Dual Constant-Phase, holds the heterogeneity of lung mechanics to be parallel. The airways bifurcate and end in identical viscoelastic tissue compartments. Additionally, we fit the so called 'constantphase' model to the data of this study. The constant-phase model assumes the lungs behave homogeneously, represented by a rigid pipe with a viscoelastic balloon sealed over one end. The airways are meant to be represented by the former, whlle the tissue is meant to be represented by the latter. The goodness-of-fit of each model was determined by the Akaike Information Criterion (AIC). The lower is the score of a model when it is fit to data, the better that model is said to be for that data.
The three experimental procedures modeled asthma, allergic inflammation, and acute lung injury. We found that the constant-phase model was significantly better then the other three in modeling the acute lung injury with AIC scores of -45.2 and -35.5 for baseline and challenge states, respectively. The Single Elastic Airway model had the lowest AIC scores for the methacholine-challenged mice and allergically inflamed mice: -41 and 57.4, respectively. The Dual Constant-Phase model behaved like the constant-phase model, the common parameters being identical and the extra ones being physiologically nonsensical. Similarly, the Branching Elastic Airway model behaved like the Single Elastic Airway model. This study is important for understanding the fundamental mechanics of mouse models of lung pathologies.
The third model, the Dual Constant-Phase, holds the heterogeneity of lung mechanics to be parallel. The airways bifurcate and end in identical viscoelastic tissue compartments. Additionally, we fit the so called 'constantphase' model to the data of this study. The constant-phase model assumes the lungs behave homogeneously, represented by a rigid pipe with a viscoelastic balloon sealed over one end. The airways are meant to be represented by the former, whlle the tissue is meant to be represented by the latter. The goodness-of-fit of each model was determined by the Akaike Information Criterion (AIC). The lower is the score of a model when it is fit to data, the better that model is said to be for that data.
The three experimental procedures modeled asthma, allergic inflammation, and acute lung injury. We found that the constant-phase model was significantly better then the other three in modeling the acute lung injury with AIC scores of -45.2 and -35.5 for baseline and challenge states, respectively. The Single Elastic Airway model had the lowest AIC scores for the methacholine-challenged mice and allergically inflamed mice: -41 and 57.4, respectively. The Dual Constant-Phase model behaved like the constant-phase model, the common parameters being identical and the extra ones being physiologically nonsensical. Similarly, the Branching Elastic Airway model behaved like the Single Elastic Airway model. This study is important for understanding the fundamental mechanics of mouse models of lung pathologies.