Abstract
Background: The gold standard for diagnosing excessive tracheal collapse is still evaluation during bronchoscopy. Today, multidetector computed tomography (MDCT) is used to confirm a suspicion of abnormal tracheal collapse. There is no gold standard for computed tomography (CT) image analysis of tracheal collapse. Purpose: To evaluate four different methods for the diagnosis of tracheal collapse using the images obtained through MDCT to help clinicians evaluate the images in daily practice. Objectives: 374 consecutive high-resolution CT scans with full inspiratory and end-expiratory CT scans were retrospectively analyzed. Methods: The images were analyzed in four different ways. The degree of collapse was based on cross-sectional areas of individual locations or volumes of entire regions: (1) 1 cm above the carina, (2) the level of maximal collapse of the trachea, (3) the entire region from the carina to the thoracic inlet, and (4) the trachea and bronchial region as defined by the software. Results: We compared three existing and one new method for image analysis of tracheal collapse by MDCT. The prevalence of tracheal collapse varied from 10.7% to 19.5% in this cohort of patients suffering from mixed lung diseases when using an expiratory collapse of ≥50% as a threshold. The four methods were comparable with highly significant Pearsons correlation coefficients (0.764-0.856). However, the four methods identified different patients with collapse of ≥50%. There was no correlation between symptoms and the degree of collapse. Conclusion: The different methods identify tracheal collapse in different patients. Hence, the diagnosis of excessive tracheal collapse can not rely solely on MDCT images. Generally, there is a poor correlation between symptoms and the degree of collapse in the different methods. However, when using the maximal collapse, there is some correlation with symptoms. When in doubt regarding the diagnosis, further investigations, such as bronchoscopy, should be carried out.