3D Robotic Catheter Shape Reconstruction and Localisation Using Appearance Priors and Adaptive C-Arm Positioning

Abstract

AbstractAccurate catheter navigation is necessary in endovascular interventions to avoid endothelial injury and subsequent complications. Although the use of robotic assistance has facilitated the navigation of catheters through complex anatomies, ambiguity in the catheter shape due to the 2D visualization provided by fluoroscopy can result in catheter and arterial wall collisions. The need for accurate shape reconstruction and localisation of the catheter has motivated the development of a range of 3D sensing techniques and augmented intraoperative imaging. The purpose of this paper is to present a 3D vision-based catheter shape reconstruction and localisation technique without the need for additional hardware. It is based on adaptive C-arm positioning under spatial constraints by incorporating appearance priors. On-line estimations of the 3D catheter shape can be achieved from the fluoroscopic images alone and are used to define the C-arm rotation that is optimal to reconstruct and localise the 3D catheter shape. The method is fully automatic and carried out without the burden of additional radiation and nephrotoxic risk to the patient. Detailed validation has been performed to demonstrate the potential clinical value of the technique.KeywordsEndovascular InterventionLeft Subclavian ArteryShape ReconstructionCatheter SegmentRobotic CatheterThese keywords were added by machine and not by the authors. This process is experimental and the keywords may be updated as the learning algorithm improves.