Software for CT-image Analysis to Assist the Choice of Mechanical-Ventilation Settings in Acute Respiratory Distress Syndrome

摘要

Acute respiratory distress syndrome (ARDS) is a critical impairment of the lung function, which occurs - among others - in severe cases of patients with Covid-19. Its therapeutic management is based on mechanical ventilation, but this may aggravate the patient's condition if the settings are not adapted to the actual lung state. Computed tomography images allow for assessing the lung ventilation with fine spatial resolution, but their quantitative analysis is hampered by the contrast loss due to the disease. This article describes software developed to assist the clinicians in this analysis by implementing semi-automatic algorithms as well as interactive tools. The focus is the assessment of the cyclic hyperinflation, which may lead to ventilator-induced lung injury. For this purpose aerated parts of the lungs were segmented in twenty ARDS patients, half with Covid-19. The results were in very good agreement with manual segmentation performed by experts: 5.3% (5.1 ml) mean difference in measured cyclic hyperinflation.