Respiratory infections, such as the novel coronavirus (SARS-COV-2) and other lung injuries, damage the pulmonary epithelium. In the most severe cases this leads to acute respiratory distress syndrome (ARDS). Due to respiratory failure associated with ARDS, the clinical intervention is the use of mechanical ventilation. Despite the benefits of mechanical ventilators, prolonged or misuse of these ventilators may lead to ventilation- associated/ventilation-induced lung injury (VILI). Damage caused to epithelial cells within the alveoli can lead to various types of complications and in- creased mortality rates. A key component of the immune response is recruitment of macrophages, immune cells that differentiate into phenotypes with unique pro- and/or anti-inflammatory roles based on the surrounding environment. An imbalance in pro- and anti-inflammatory responses can have deleterious effects on the individual's health. To gain a greater understanding of the mechanisms of the immune response to VILI and post-ventilation outcomes, we develop a mathematical model of interactions between the immune system and site of damage while accounting for macrophage polarization. Through Latin hypercube sampling we generate a virtual cohort of patients with biologically feasible dynamics. We use a variety of methods to analyze the results, including a random forest decision tree algorithm and parameter sensitivity with eFAST. Analysis shows that parameters and properties of transients related to epithelial repair and M1 activation and de-activation best predicted outcome. Using this new information, we hypothesize interventions and use these treatment strategies to modulate damage in select virtual cases.

中文翻译：

---

呼吸机诱发的肺部炎症的数学模型

呼吸道感染，例如新型冠状病毒（SARS-COV-2）和其他肺损伤，会损害肺上皮。在最严重的情况下，这会导致急性呼吸窘迫综合征（ARDS）。由于与ARDS相关的呼吸衰竭，临床干预是使用机械通气。尽管使用机械呼吸机有好处，但长期使用或误用这些呼吸机仍可能导致与呼吸有关的/通气引起的肺损伤（VILI）。对肺泡内的上皮细胞造成的损害可导致各种类型的并发症并增加死亡率。免疫反应的关键组成部分是募集巨噬细胞，巨噬细胞是根据周围环境分化为具有独特的促炎和/或抗炎作用的表型的免疫细胞。促炎和抗炎反应的不平衡会对个体的健康产生有害影响。为了更好地了解对VILI的免疫反应和换气后结果的机制，我们在考虑巨噬细胞极化的同时建立了免疫系统与损伤部位之间相互作用的数学模型。通过拉丁超立方采样，我们生成了具有生物学可行性动态的患者虚拟队列。我们使用多种方法来分析结果，包括随机森林决策树算法和eFAST的参数敏感性。分析表明，与上皮修复，M1激活和失活相关的瞬态参数和特性最能预测结果。利用这些新信息，