Assessing Spline-Based Multi-resolution 2D-3D Image Registration for Practical Use in Surgical Guidance
A spline-based multi-resolution 2D-3D image registration algorithm has recently been introduced [1-3]. However, its accuracy, robustness, and efficiency have not been fully investigated. In this paper, we focus on assessing and improving this newly introduced 2D-3D registration algorithm. A phantom and a cadaver test, together with their respective ground truths, were specially designed for this purpose. A novel least-squares normalized pattern intensity (LSNPI) similarity measure was proposed to improve the accuracy and robustness. Several parameters that may also affect its robustness, accuracy, and efficiency are experimentally determined, including the final resolution level, the initial guess of the patient pose, the number of 2D projection images, and the angle between 2D projection images. Our experiments show that it is feasible for the assessed 2D-3D registration algorithm to achieve sub-millimeter accuracy in a realistic setup in less than two minutes, when it is used together with the newly proposed similarity measure.
KeywordsGround Truth Projection Image Fiducial Marker Registration Algorithm Target Registration Error
Unable to display preview. Download preview PDF.
- 1.Jonić, S., Thévenaz, P., Unser, M.: Multiresolution spline-based 3D/2D registration of CT volume and C-arm images for computer-assisted surgery. In: Sonka, M., Hanson, K.M. (eds.) Proc. SPIE. Medical Imaging 2001: Image Processing, vol. 4322, pp. 1101–1109 (2001)Google Scholar
- 2.Jonić, S., Thévenaz, P., Unser, M.: Multiresolution-based registration of a volume to a set of its projections. In: Sonka, M., Fitzpatrick, J.M. (eds.) Proc. SPIE. Medical Imaging 2003: Image Processing, vol. 5032, pp. 1049–1052 (2003)Google Scholar
- 3.Jonić, S., Thévenaz, P., Zheng, G., Nolte, L.-P., Unser, M.: Spline-based registration of a 3D CT to a set of C-arm images. IEEE T Med Imaging (submitted)Google Scholar
- 6.Weese, J., Buzug, T.M., Lorenz, C., Fassnacht, C.: An approach to 2D/3D registration of a vertebra in 2D x-ray fluoroscopies with 3D CT images. In: Troccaz, J., Grimson, E., Mijsges, R. (eds.) Proc. CVRMED-MRCAS 1997, pp. 119–128 (1997)Google Scholar
- 8.LaRose, D., Hayouth, J., Kanade, T.: Transgraph: Interactive intensity-based 2D/3D registra-tion of X-ray and CT data. In: Kanson, K.M. (ed.) Proc. SPIE Medical Imaging 2000: Im-age Processing, vol. 3979, pp. 385–396 (2000)Google Scholar
- 9.Rohlfing, T., Russakoff, D.B., Murphy, M.J., Maurer, C.R.: An intensity-based registration algorithm for probabilistic images and its application for 2-D to 3-D image registration. In: Sonka, M., Fitzpatrick, J.M. (eds.) Proc. SPIE Medical Imaging 2002: Image Processing, vol. 4684, pp. 581–591 (2002)Google Scholar
- 12.Yao, J., Taylor, R.: Assessing accuracy factors in deformable 2D/3D medical image regis-tration using a statistical pelvis model. In: Proc. of the Ninth IEEE International Conference on Computer Vision (ICCV 2003), pp. 1329–1334 (2003)Google Scholar