Stereoscopic particle image velocimetry analysis of healthy and emphysemic acinus models

Inhaled particles reaching the alveolar walls have the potential to cross the blood gas barrier and enter the blood stream. Pulmonary dosimetry, however, is not well understood. Numerical and experimental studies shed some light on the mechanisms of particle transport, but realistic geometries have not been investigated. In order to accurately predict particle deposition, the characteristics that affect deposition need to be understood. This includes, but is not limited to, fluid flow, lung morphology, breathing conditions, and particle concentration. Various geometries have been used for research, but very few are close representations to in vivo geometry. Most studies have used simplified or idealized geometries based on published dimensions, but none replicate actual in vivo geometry; even fewer examine the differences that exist between healthy and diseased lung geometries. The following work analyzes and compares the flow fields that exist in replica healthy and emphysemic lungs by using realistic geometries and breathing conditions. Actual human lung casts for in vivo healthy and emphysemic geometries were obtained, scanned, and used to reconstruct three dimensional replica models. From these geometries, hollow compliant models were created and used to simulate breathing under healthy and diseased conditions. It was shown that major geometric differences exist between the healthy and emphysemic models. Specifically, the emphysemic model alveoli appeared to merge into a single large alveolus with no potential regions of recirculation, as compared to the healthy model, which contained smaller, more distinct alveoli, and an overall model volume 11x smaller than the emphysemic model. Each experimental geometry was examined using stereoscopic particle image velocimetry (stereoPIV) techniques. Realistic flow conditions were derived by the application of scaling theory to convert from the in vivo size to the large scale experimental set up. Experimental techniques were validated by comparing to computational fluid dynamic (CFD) results when using a simplified three-dimensional (3D) bulb geometry. Following validation, experimental flow fields were examined, on the large experimental scale, using velocity and streamline plots, for healthy and emphysemic geometries. It was shown that reversible flow was present in both models; even in locations representing a high probability for recirculating or irreversible flow. Each of the experimental flow fields was then scaled to represent in vivo velocity predictions. The emphysemic model (run under normal emphysemic breathing conditions) had a flow rate 8x that of the healthy model with normal breathing. The inlet velocity for the healthy normal breathing model, however, was 1.6x larger than the emphysemic. These flow results are a function of both the model geometry and the applied realistic breathing conditions. The distribution of in vivo velocity magnitude over the flow field was also different between the healthy and emphysemic models. Specifically, the healthy model enters at higher velocities than the emphysemic, and then uniformly slows as the fluid moves towards the walls. The emphysemic model yields a large region of fast flow near the inlet and slows at random locations as it approaches the walls. It was reasoned that even though inhaled particles would likely travel further into the emphysemic model compared to the healthy, the distance to reach the wall would be much greater in emphysema as compared to healthy. This would result in higher deposition efficiencies for healthy models as compared to emphysema, which is consistent with results found in the literature