Computational models for respiratory gas exchange and their applications

The primary function of the lungs is the transport and exchange of oxygen and removal of carbon dioxide that is critical in supporting normal function of vital body organs. Various modelling studies have attempted to investigate and capture aspects of the gas exchange process and its regulation with different levels of complexity and detail. The aim of this thesis is to assess the trade-off between gas exchange model complexity and feasibility and within a respiratory system modelling framework, and its applications to facilitate understanding of lung physiology during normal function and pathology. An integrated comprehensive modelling framework that allows gas exchange prediction within anatomically based lung geometry is presented. Structural and functional simplifications are assessed to result in a class of models with increasing complexity in their description of gas exchange in the human lungs, which span from simple steady state prediction, to fitting empirical equations that capture characteristic behaviour, to complex equations derived from underlying physiological principles; and the model can be scaled from a few compartments to distributions of approximately 32,000 compartments in the whole lung. The classes of gas exchange models are assessed for their applicability in modelling key pulmonary functions and appropriate models are used to investigate three questions relating to physiology, experimentation/imaging, and gas exchange during clinical therapy. First, the simplest steady state model is used to study structure-function relationships in the normal lung and reconcile differing experimental observations on the relative importance of passive ventilation-perfusion matching mechanisms. Simulation results show that during quiet supine breathing, the effects of gravity introduce significant heterogeneity in ventilation and perfusion but also provide spatial correlation, while the effects of ‘matched’ airway and arterial structure play a relatively minor role. Second, the model at its highest resolution is validated through simulations of two well-established experimental protocols: the gold standard multiple inert gas elimination technique, and high resolution specific ventilation imaging (SVI). The model is able to mimic the two experimental protocols and can give accurate O2 predictions of global and regional function in normal and abnormal lung states. Furthermore, an in-silico examination of the assumptions of the specific ventilation imaging technique are performed, which pointed to the confounding influence of venous blood flow and image misalignment on experimental obtained SVI maps. Third, the feasibility of model application to a clinical setting is examined by applying the simplified model to systematically investigate several mechanisms of efficacy for nasal high flow therapy in a cohort of 20 post-cardiac surgery patients. Results showed that this generic model can be parameterised to represent individualised patient respiratory response. Moreover, model predictions show that flow induced nasopharyngeal washout largely reduce respiratory efforts without improving oxygenation, when arterial carbon dioxide is within the normal range. The highly debated mechanism of pressure induced alveolar recruitment is required to produce model predictions that are consistent with clinical measurements for some individuals in this patient cohort. The models and studies presented here provide a basis for extension and application to future research in studying interaction of underlying physiology mechanisms, biomedical imaging of pulmonary function, and clinical problems of gas exchange.

Name thesis:degree_name

PhD

Level thesis:degree_level

Doctoral

Discipline thesis:degree_discipline

Bioengineering

Grantor dc:publisher

ResearchSpace@Auckland

Year dc:date.issued

2017

Author dc:creator

• Kang, Wenying

Advisors dc:contributor.advisor

• Tawhai, M
• Clark, A

Statement dc:rights

• Items in ResearchSpace are protected by copyright, with all rights reserved, unless otherwise indicated. Previously published items are made available in accordance with the copyright policy of the publisher.

Licence dc:rights.uri

• https://researchspace.auckland.ac.nz/docs/uoa-docs/rights.htm

Handle dc:identifier.uri

https://hdl.handle.net/2292/36832

OAI identifier oai:identifier

oai:researchspace.auckland.ac.nz:2292/36832