2D-3D registration based on shape matching

Abstract

This paper describes a method for determining an object's pose given its 3D model and a 2D view. This 2D-3D registration problem arises in a number of medical applications, e.g. image guided spine procedures. Previous approaches often rely on a good initial estimate of the pose parameters and an optimization procedure t o r e ne this initial pose estimate, e.g. the iterative closest point (ICP). However, such algorithms can identify local minima as global minima, leading to registration errors, if the initial pose is not carefully chosen. The speci cation of the appropriate initial conditions, however requires user interaction and is time consuming. We pr opose an approach where sample 2D views are generated f r om the 3D model and matched against the given view (2D-3D registration). Additional views are then generated in the vicinity of the best view and the procedure i s r epeated until convergence. Results of estimating the coordinates of a vertebrae spine bone from its 3D model, obtained f r om volumetric (CT or MR) data, and a 2D view, as might be obtained f r om uoroscopic data, demonstrates that the pose can be reliably obtained without requiring extensive user interface.

3D reconstruction of the human rib cage from 2D projection images using a statistical shape model

“…In recent years, several methods for the pose reconstruction of 3D objects from 2D projections were introduced [14,[19][20][21]. The work of Lavallée et al [19] deals with determining position and orientation of arbitrary surfaces based on contour points in 2D image data.…”

Co-registration of intra-operative brain surface photographs and pre-operative MR images

“…Burschka et al [6] present a method to register CT scans to endoscopic images by a 3D reconstruction of the 2D images followed by a registration of the reconstruction to the CT data. The use of 2D views for pose estimation of a pre-surgical 3D image has been investigated in [9]. A variational approach is proposed in [12] for the matching of 3D CT data to 2D ultrasound slices that, unlike other volume-to-slice approaches, only uses the given data and relies on a higher order regularization.…”

An Adaptive Technique for Accurate Feature Extraction from Regular and Irregular Image Data

“…Typically approaches for feature extraction in intensity or range images can only be applied principally to that specific image type; none of the typical approaches can be readily applied to both image types without image preprocessing. As much research has been undertaken in trying to integrate 2D and 3D images [2], specifically in application areas such as medical imaging [6] and face recognition [5,21], we believe that a single approach that can be used on both image types would greatly aid the process of image fusion.…”