EUS with CT improves efficiency and structure identification over conventional EUS

Kirby G. Vosburgh, Nicholas Stylopoulos, R. San Jose Estepar, Randy E. Ellis, Eigil Samset, Christopher C. Thompson
Gastrointestinal Endoscopy
Volume 65, Number 6, Pages 866-870
2007

Download full paper

Abstract

Background: EUS is complicated because of the subtleties of US interpretation, small fields of observation, and uncertainty of probe position and orientation. Objective: Improved EUS performance is sought by providing contextual information to support US probe positioning and identification of features in US images. Our aims were to demonstrate the feasibility of the image registered gastroscopic US (IRGUS) system in a porcine model and to compare the effectiveness and the efficiency of IRGUS with traditional EUS. Design: Animal feasibility study. Interventions: The IRGUS system uses preprocedure CT and miniature US probe trackers to create real-time synthetic displays of the position of the probe tip and a matched slice of CT data for comparison with the US image. Participants used EUS and IRGUS systems in a porcine model to evaluate the speed and accuracy of structure identification. Main Outcome Measurements: The performance and utility of IRGUS were determined by the number of correctly identified structures in a timed trial, kinematic variables, and a structured survey. Results: IRGUS was twice as effective as EUS in localizing and identifying individual structures. In timed trials, IRGUS users identified over 25 accuracy of feature identification was statistically significant, and 90 for these tasks. Conclusions: IRGUS appears feasible and may be superior to conventional EUS in efficiency and accuracy of probe positioning and in image interpretation. IRGUS has the potential to shorten the EUS learning curve and to broaden the adoption of EUS techniques by gastroenterologists.

Display 1 shows the position of the tip of an endoscopic probe in the stomach relative to the ribs and the major vessels extracted from a preprocedure CT image of the subject. The oblique green slab shows the acquisition plane of the US transducer head. The red line in the plane is the nominal track of the biopsy needle. Display 2 shows the plane of the volumetric CT data that corresponds to the observed US image (ie, matches the green oblique plane in Display 1). Display 3 shows the unmodified image made by the US system. The operator uses Display 1 for overall orientation and identification of key landmarks, and then uses Displays 2 and 3 to identify features in the US image and to build confidence in the interpretation.

Reference

Vosburgh KG, Stylopoulos N, San Jose Estepar R, Ellis RE, Samset E, Thompson CC. EUS with CT improves efficiency and structure identification over conventional EUS. Gastrointestinal Endoscopy 2007;65(6):866-870.

Bibtex entry

@Article{vosburgh07,
  author         = {Kirby G. Vosburgh and Nicholas Stylopoulos and Raul {San   
                   Jose Estepar} and Randy E. Ellis and Eigil Samset and       
                   Christopher C. Thompson},                                   
  title          = {{EUS} with {CT} improves efficiency and structure          
                   identification over conventional {EUS}},                    
  journal        = {Gastrointestinal Endoscopy},                               
  year           = {2007},                                                     
  volume         = {65},                                                       
  number         = {6},                                                        
  pages          = {866--870}
}