M.Sc. Lu Guo

B.Sc. Tridib Ghosh

Bronchoscopy plays a crucial role in pulmonary diagnostics and interventions. While conventional computer vision techniques have been integrated into several vision-based bronchoscopy systems, deep learning offers promising opportunities to further enhance their capabilities. However, a key challenge in applying deep learning methods lies in the need for diverse and high-quality training data to ensure generalization. Among these, bronchoscopic video data is particularly valuable but remains scarce in publicly available datasets. In contrast, virtual bronchoscopic videos can be easily synthesized from CT scans. This project focuses on bridging the gap between virtual and real bronchoscopic videos, with the goal of unlocking potential of deep learning methods for simulation and navigation assistance in bronchoscopy.