Heterogeneidade pulmonar e lesão induzida pela ventilação mecânica

• Ano de defesa: 2018
• Autor(a) principal: Ribeiro, Gabriel Casulari da Motta
• Tipo de documento: Tese
• Tipo de acesso: Acesso aberto
• Idioma: por
• Instituição de defesa: Universidade Federal do Rio de Janeiro; Brasil; Instituto Alberto Luiz Coimbra de Pós-Graduação e Pesquisa de Engenharia; Programa de Pós-Graduação em Engenharia Biomédica; UFRJ
• Palavras-chave em Português: Sistema respiratório; Imagens médicas; Modelagem matemática; CNPQ::ENGENHARIAS::ENGENHARIA BIOMEDICA::ENGENHARIA MEDICA
• Link de acesso: http://hdl.handle.net/11422/20284

Lung expansion during a respiratory cycle is spatially heterogeneous with possible consequences for the outcome of mechanical ventilation. This thesis studied this heterogeneity in a large animal model of the first 24 h of mechanical ventilation of initially normal lungs, and in a theoretical model of gas mixing during a multiplebreath N2 washout maneuver. The animal experiments compared two initial states of lung expansion, prone position (homogeneous) and supine position (heterogeneous). The ventilator settings were in accordance to current clinical practice, increasing the translational relevance. Computed and positron emission tomography were used to asses voxel-level distribution of static (aeration) and dynamic (strain) lung expansion, and regional metabolic activity (marker of inflammation). Supine animals showed progressive deterioration of regional lung mechanics resulting in spatially distinct mild tissue injury and inflammation, as well as gene expression. ln contrast, the prone position had mild inflammation and loss in aeration without increase in heterogeneity. A new model to explain the N2 washout maneuver was proposed. This considers the lungs as parallel compartments with a series dead space. Compared to the classical model, the new model imposes fewer restrictions on the ventilatory cycles during the maneuver, potentially allowing its application in mechanically ventilated patients. Computational simulations and bench experiments demonstrated that the proposed model outperforms the classical in the reconstruction of distributions of specific ventilation (linked to dynamic expansion). The current findings will help to define mechanical ventilation strategies for patients without initial lung injury and advance bedside monitoring, with potential impact on clinical practice.