Assessing the accuracy of a multisection robotic bronchoscope prototype in localization and targeting of small pulmonary lesions

Jorind Beqari,Jacob Hurd,Fumitaro Masaki,Bassel Tfayli,Hussein Kharroubi,Masahito Naito,Franklin King,Yolonda Colson
DOI: https://doi.org/10.1016/j.xjtc.2024.05.011
2024-05-24
Abstract:Objectives: Robotic bronchoscopy (RB) has emerged as a novel technique to address issues with the biopsy of small peripheral lung lesions. The objective of this study was to quantitatively assess the accuracy of a novel multisection robotic bronchoscope compared with current standards of care. Methods: This is a prospective, single-blind, comparative study where the accuracy of a multisection RB was compared against the accuracy of standard electromagnetic navigational bronchoscopy (EM-NB) during lesion localization and targeting. Five blinded subjects of varying bronchoscopy experience were recruited to use both RB and EM-NB in a swine lung model. Accuracy of localization and targeting success was measured as the distance from the center of pulmonary targets at each anatomic location. Subjects used both RB and EM-NB to navigate to 4 pulmonary targets assigned using 1:1 block randomization. Differences in accuracy and time between navigation systems were assessed using Wilcoxon rank-sum test. Results: Of the 40 total attempts per modality, successful targeting was achieved on 90% and 85% of attempts utilizing RB and EM-NB, respectively. Furthermore, RB demonstrated significantly lower median distance to the real-time EM target (1.1 mm; interquartile range [IQR], 0.6-2.0 mm) compared with EM-NB (2.6 mm; IQR, 1.6-3.8) (P < .001). Median target displacement resulting from lung and bronchus deformation during bronchoscopy was found to be significantly lower using RB (0.8 mm; IQR, 0.5-1.2 mm) compared with EM-NB (2.6 mm; IQR, 1.4-6.4 mm) (P < .001). Conclusions: The results of this study demonstrate that the multi-section RB prototype allows for improved localization and targeting of small peripheral lung nodules compared with current nonrobot bronchoscopy modalities.
What problem does this paper attempt to address?