Our understanding of respiratory flow phenomena has been consolidated over decades with the exploration of in vitro and in silico canonical models that underscore the multiscale fluid mechanics spanning the vast airway complex. In recent years, there has been growing recognition of the significant intersubject variability characterizing the human lung morphometry that modulates underlying canonical flows across subjects. Despite outstanding challenges in modeling and validation approaches, exemplified foremost in capturing chronic respiratory diseases, the field is swiftly moving toward hybrid in silico whole-lung simulations that combine various model classes to resolve airflow and aerosol transport spanning the entire respiratory tract over cumulative breathing cycles. In the years to come, the prospect of accessible, community-curated datasets, in conjunction with the use of machine learning tools, could pave the way for in silico population-based studies to uncover unrecognized trends at the population level and deliver new respiratory diagnostic and pulmonary drug delivery endpoints.

中文翻译：

---

呼吸运输现象和主体间变异性的多尺度建模


几十年来，随着对体外和计算机规范模型的探索，我们对呼吸流动现象的理解得到了巩固，这些模型强调了跨越巨大气道复合体的多尺度流体力学。近年来，人们越来越认识到人类肺形态测量学的显著主体间变异性，该变异性调节了受试者间的潜在典型血流。尽管在建模和验证方法方面存在突出挑战，尤其是在捕获慢性呼吸系统疾病方面，但该领域正在迅速转向混合计算机全肺模拟，该模拟结合了各种模型类别，以解决累积呼吸周期内跨越整个呼吸道的气流和气溶胶运输问题。在未来几年，可访问的、社区策划的数据集的前景，结合机器学习工具的使用，可以为基于人群的计算机研究铺平道路，以揭示人群层面未被识别的趋势，并提供新的呼吸诊断和肺部药物递送终点。