Abstract Rationale Idiopathic pulmonary fibrosis (IPF) is characterized by spatially heterogeneous remodeling of collagen fibers in the lung parenchyma, resulting in the destruction of alveolar architecture. This pathological restructuring reflects tissue-level changes whose underlying mechanisms remain poorly understood. Building on an established agent-based model (ABM) that captures the increase in density and spatial heterogeneity of fibrotic tissue, we sought to incorporate quantitative morphological metrics. These metrics aim to refine the ability of the model to recapitulate the architectural alterations observed in diseased tissue. Methods Eight-week-old C57BL/6 mice were treated with either 3 units of bleomycin or vehicle and sacrificed three weeks post-instillation. Lungs were fixed at 5 cmH2O, contrasted with uranyl acetate and osmium tetroxide, embedded in glycol methacrylate, and imaged using a Bruker SkyScan 1173 micro-CT scanner. Regions of interest excluding major airway structures were isolated from the full scans. Control 3D structures served as initial conditions for an ABM in which stochastically moving pro- and anti-fibrotic agents modulate voxel densities over time to simulate parenchymal remodeling. Pore (alveolar compartment) size distributions were quantified within these parenchymal regions using a three-dimensional distance transform combined with watershed-based segmentation. Cumulative distribution functions were generated, and a permutational multivariate analysis of variance (PERMANOVA) was applied to assess group-level differences in pore size distributions. Results Although mean tissue volume fractions did not differ significantly between groups (control mean: 0.7347; bleomycin-treated mean: 0.6994), control lungs had a greater proportion of smaller alveolar compartments. In contrast, bleomycin-treated lungs exhibited a rightward shift and greater variance in the pore size distribution (Figure 1), reflecting an increased prevalence of larger pores. The model captured this rightward-tailed behavior consistent with diseased tissue. A PERMANOVA test confirmed a statistically significant difference in pore size distributions between control and bleomycin-treated lungs (pseudo-F=13.38, p = 0.0253), indicating a moderate-to-strong shift between groups. Conclusions We found that fibrotic lung tissue presents an increased frequency of large alveolar compartments despite preserving the total tissue volume fraction. The frequency distribution of alveolar volumes thus reflects structural remodeling in a way that total alveolar volume cannot and provides a clear target metric for our model outputs. These results suggest that fibrogenesis involves filling of the smallest airspaces with fibrotic tissue or coalescence of small adjacent spaces, together with expansion of larger ones via parenchymal tethering. Incorporating these mechanisms into our computational model may reveal how localized microarchitectural changes drive macroscopic disease progression. This abstract is funded by: NIH grant T32 HL-076122, U01 HL139466, R01HL151630, and NSF 2225554
Wilson et al. (Fri,) studied this question.