Related Experiment Video
Updated: Aug 27, 2026
Robotic-assisted Bronchoscopy Combined with Multimodal Imaging for Targeted Lung Cryobiopsies
Published on: July 19, 2024
Vital study: Real-world comparison of robotic and imaging-guided bronchoscopy: Impact of real-time imaging and
Roshen Mathew1, Winnie Elma Roy1
1Department of Pulmonary and Critical Care Medicine, WVU Camden Clark Medical Center, Parkersburg, WV, USA.
Background:
Accurate bronchoscopic diagnosis of peripheral pulmonary lesions remains limited by navigation challenges and CT-to-body divergence. Integrated imaging techniques, including augmented fluoroscopy (AF) and C-arm-based tomography (CABT), have improved lesion localization across both robotic and non-robotic platforms. However, real-world comparative data in community settings remain limited.
Methods:
We conducted a retrospective, single-operator study comparing LungVision (LV) bronchoscopy (2022-2023, n = 75) and Galaxy robotic bronchoscopy (Galaxy RB) (2024-2025, n = 105) in a community hospital. Both platforms utilized integrated AF and CABT with a standard C-arm. Primary endpoints were strict diagnostic yield (DY) at index bronchoscopy (ATS/ACCP criteria) and diagnostic accuracy at 1.5 years. Secondary analyses evaluated determinants of diagnostic success and an exploratory randomized breath-hold subgroup.
Results:
Strict DY was higher with Galaxy RB compared to LV (87.6% vs 74.7%; OR 2.40, 95% CI 1.10-5.24; p = 0.03). Diagnostic accuracy at 1.5 years was also greater (97.1% vs 92%). Non-diagnostic rates were lower with Galaxy RB (12.4% vs 25.3%), with all cases resolved on follow-up. Within the Galaxy RB cohort, diagnostic success was strongly associated with AF localization (OR 10.76, p <0.001), visual tool adjustment (OR 9.57, p <0.001), and tool-lesion alignment (TIL/TTL vs NTL; OR 16.55, p = 0.040). Breath-hold was not associated with improved yield (81.4% vs 86.5%; OR 0.68, p = 0.57).
Conclusion:
In a real-world community setting using integrated imaging, robotic bronchoscopy demonstrated higher diagnostic yield and accuracy compared to a non-robotic platform. However, diagnostic success was primarily driven by effective tool-lesion alignment and real-time imaging rather than platform type or breath-hold utilization.
