Sammendrag
Virtual bronchoscopy is often used for planning a real bronchoscopy procedure. Software applications are developed for virtual bronchoscopy, involving usually segmentation of the tracheobronchial tree from the medical image scan, which is a difficult operation, both conceptually and from the computer implementation and running time point of view. That is why in this paper, a new method for bronchoscopy procedure planning that does not require such a segmentation is presented. The proposed procedure involves automatic path generation between the starting and ending points, skin removal, an algorithm for detection and resolution of collision with the airways walls and validation of the automatically created path. Results are presented for two datasets – one being the representation of a theoretical lungs model, with six levels of branches and the other one being the image scan of a real patient. Together with a system for tracking the bronchoscope during the real procedure, the proposed method can improve the diagnostic success rate of lung cancer using bronchoscopy and decrease the discomfort perceived by the patient.