Purpose: To evaluate the performance of discrete compactness (DC) and tortuosity (TR) metrics in the structures of the trachea and pulmonary vascular tree, by processing computed tomography (CT) images to quantify significant differences between populations of control subjects (HC) and patients diagnosed with COVID-19.
Methods: A total of 156 subjects were analyzed, 78 of which are HC and 78 patients diagnosed with COVID-19. The structures of the trachea and vascular tree were segmented using image processing, and volume, DC, and TR metrics were calculated. Subsequently, a statistical analysis was carried out to quantify whether there are significant differences between the groups analyzed in this research.
Results: The statistical analysis presented significant differences for all the metrics evaluated with a p value < 0.05 between the HC populations and patients diagnosed with COVID-19. In particular, DC and TR measurements were very marked in the structure of the trachea with a value of p < 0.01.
Conclusions: All the evaluated metrics presented statistically significant changes. In the case of CD and TR, the statistical variation was more significant in the region of the trachea and airways in patients diagnosed with COVID-19; This suggests that the proposed metrics are sensitive to the morphological changes that occur in these structures and could be considered as potentially useful imaging markers to quantify the damage caused by this pathology.