Objective: To assess three-dimensional computed tomography 'bronchoscopic' (3-DCTB) reconstruction of routine CT data as a non-invasive method of airway visualization, and compare it with fibreoptic bronchoscopy (FOB).
Methods: Fourteen datasets were acquired from 13 patients undergoing both FOB and CT examination of the chest. Standard continuous volume CT using 6 mm collimation and clinical FOB techniques were employed. Images were obtained from 3-DCTB reconstructions by segmentation and surface recognition algorithms generating surface rendered 'bronchoscopic views'. These were scored for technical quality and anatomical detail. The most distal bronchi seen in left upper and right lower lobes were recorded for FOB and 3-DCTB.
Results: On FOB, the subsegmental bronchi were seen in right lower and in left upper lobe in 10/14 cases and 4/14 cases, respectively. Visualization of the subsegmental airways was not achieved with 3-DCTB, as they could not be identified with confidence. 3-DCTB never achieved a more distal view than obtained by FOB. Using 3-DCT, the right, lower lobe segmental bronchi were seen in 10/14 cases, and lobar bronchus in 14/14 cases (two occluded). In the left upper lobe, 3-DCT showed segmental bronchi in 6/14 cases, lobar bronchus in 11/14 cases (one occluded) and the left main bronchus appeared occluded in 3/14 cases. Overall, technical quality and anatomical detail scores of the carina and proximal bronchi ranked significantly higher than views of segmental bronchi.
Conclusions: 3-DCTB cannot routinely replace FOB for inspection of major and segmental bronchi. Subsegmental bronchi cannot be adequately demonstrated by 3-DCTB using 6 mm collimation datasets.