An Automatic Registration Method for Radiation-Free Catheter Navigation Guidance

Abstract

Catheter navigation is typically based on fluoroscopy. This implies exposure to harmful radiation, lack of depth perception and limited soft-tissue contrast. Catheter navigation would benefit from guidance that makes better use of detailed pre-operatively acquired MR/CT images, while reducing radiation exposure and improving spatial awareness of the catheter pose and shape. A prerequisite for such guidance is an accurate registration between the catheter tracking system and the MR/CT scans. Existing registration methods are lengthy and cumbersome as they require a lot of user interaction. This forms a major obstacle for their adoption into clinical practice. This paper proposes a radiation-free registration method that minimizes the impact on the surgical workflow and avoids most user interaction. The method relies on catheters with embedded sensors that provide intra-operative data that can either belong to the vessel wall or to the lumen of the vessel. Based on the acquired surface and lumen points an accurate registration is computed automatically, with minimal user interaction. Validation of the proposed method is performed on a synthetic yet realistic aorta phantom. Input from electromagnetic tracking, force sensing, and intra-vascular ultrasound are used as intra-operative sensory data.