Evaluating the impact of artificial obstructions on the aerodynamics of a personalized computational fluid dynamic model of human tracheobronchial tree

Long-term exposure to pollutants and irritating agents may trigger the development of respiratory system infections and diseases, representing the second leading cause of death in the Western Word. Obstructive lung diseases, such as asthma, and chronic obstructive pulmonary disease are characterized by a reduction in bronchi lumen area, with consequent alteration of bronchial tree functionality. The main causes of obstructive diseases (ODs), often asymptomatic until significant lung damage has occurred, are still not completely understood. Studying the aerodynamics of the respiratory system might help the comprehension and characterization of the factors involved in ODs onset and development, supporting diagnosis, prediction and treatment evaluation. In this context, computational fluid dynamics (CFD) represents an attractive tool to obtain quantitative and local information about tracheobronchial airflow dynamics, overcoming the limitations of current clinical approaches (i.e., spirometry and cardiopulmonary exercise tests), which provide an average at-the-mouth estimation of residual functionality. The aim of this thesis work was to evaluate the impact of artificial pulmonary obstructions on the airflow dynamics of a personalized CFD model of human tracheobronchial tree. In detail, starting from computed tomography images of a healthy subject, a personalized model of human tracheobronchial tree was reconstructed, extending from trachea to third-generation bronchi. Three additional models were created by generating a virtual stenosis at three different locations along the healthy model: at mid trachea (G0), at first- (G1) and at second-generation bronchi (G2). Unsteady-state simulations were performed by imposing a generalized pressure waveform at the inlet section, and reference pressure at the outlet boundaries. k-ω approach was used to model turbulence. The entire breathing cycle was simulated, including inspiration and expiration phasis. Healthy and obstructed models were compared in terms of pressure and velocity distribution. Additionally, near-wall and intrabronchial biomechanical quantities were computed based on wall shear stress (WSS) and helicity, respectively. Overall, a decrease in tracheal flow rate (higher for proximal obstruction) was observed for constricted models with respect to healthy one (-64%, -36%, and -27%, for G0, G1, and G2 respectively). Local changes in pressures, velocity and WSS-based quantities clearly emerged at the obstruction. Time-average WSS (TAWSS) and topological shear variation index (TSVI) increased locally at the constricted bronchial region. The latter did not reflect on global TAWSS median values, decreasing for diseased model. Interestingly, higher values of TAWSS (+100%) emerged at the inspiration vs. expiration phase. As for bulk descriptors, higher helicity intensity (h2) characterized the (1) healthy vs. diseased models, and the (2) inspiration vs. expiration phase. The presence of OD did not affect the balance of counter-rotating helical flow structures, clearly evolving from trachea to bronchi in all the models. In conclusion, tracheobronchial obstructions affect bronchial airflow dynamics, affecting pressure and velocity distributions. Higher local values of TAWSS and TSVI emerged at the constricted region in inspiration/expiration phases, suggesting their potential as obstruction site markers. This study is a first investigation of ODs aerodynamics impact; further investigations on patient-specific diseased models are needed.