Fast 2d-3d point-based registration using GPU-Based preprocessing for image-guided surgery

  • Authors:

  • Helen Hong;Kyehyun Kim;Seongjin Park

  • Affiliations:

  • Division of Multimedia Engineering, College of Information and Media, Seoul Women's University, Seoul, Korea;School of Computer Science and Engineering, Seoul National University, Seoul, Korea;School of Computer Science and Engineering, Seoul National University, Seoul, Korea

  • Venue:
  • CIARP'06 Proceedings of the 11th Iberoamerican conference on Progress in Pattern Recognition, Image Analysis and Applications
  • Year:
  • 2006

Quantified Score

Hi-index 0.00

Visualization

Abstract

We propose a 2D-3D point-based registration method that provides fast and efficient alignment of X-ray fluoroscopy and CT images. Our method is divided into two procedures: pre-operative and intra-operative procedures. For pre-operative procedures, we generate digitally reconstructed radiographs (DRRs) from 3D volume using graphics hardware. In intra-operative procedures, we perform a hierarchical registration that includes in-plane registration using principal axes method and out-plane registration using minimal error searching method in spherical coordinates. This method reduces a degree of freedom from 6-DOF to 2-DOF. Experimental results using 2 cardiac phantoms show that our DRRs generation method is more than 150 times faster than software-based ray casting methods, and our hierarchical registration technique effectively matches DRRs and 2D images.