Premium
SU‐GG‐J‐104: Image‐Guided Intervention Based On Registration Between C‐Arm Projection and Cone‐Beam Reprojection of Reconstructed Volume
Author(s) -
Chen Z
Publication year - 2008
Publication title -
medical physics
Language(s) - English
Resource type - Journals
SCImago Journal Rank - 1.473
H-Index - 180
eISSN - 2473-4209
pISSN - 0094-2405
DOI - 10.1118/1.2961654
Subject(s) - volume rendering , computer vision , voxel , artificial intelligence , cone beam computed tomography , volume (thermodynamics) , projection (relational algebra) , image registration , iterative reconstruction , computer science , segmentation , reprojection error , image segmentation , scanner , nuclear medicine , rendering (computer graphics) , medicine , radiology , physics , image (mathematics) , computed tomography , algorithm , quantum mechanics
Purpose: To provide an image‐guided manipulation of the catheter introduction during intervention procedure by registering the C‐arm projection image with the cone‐beam reprojection image of a reconstructed volume that was generated previously by a volume CT (64‐slice CT or flat‐panel‐ based cone‐beam CT). Method and Materials First, reconstruct a volume for the interventional region by volume CT scanning and reconstruction, i.e., generate an angiographic volume; Second, perform volume segmentation to segment 3D blood vessels; Third, capture the C‐arm projection image during catheter introduction; Fourth, calculate the cone‐beam reprojection image on the angiographic volume, and adjust the reprojection parameters based on the matching or coregistration with the C‐arm image; Finally, perform 3D display of the segmented vessels where vessel segments are colored as indications of the catheter position. Results: The angiographic volume for the interventional region is reconstructed from a 64‐slice CT scanner or from a cone‐beam CT scanner. The volume segmentation is based on voxel value and vessel tree structure knowledge. The 2D image registration between the C‐arm image and the cone‐beam reprojection image of the angiographic volume requires a trial‐and‐error process by adjusting the reprojection geometric settings. The 3D vessels are displayed by fast isosurface or volume rendering. Conclusion: By reprojecting a reconstructed angiographic volume and co‐registering the projection image with the C‐arm projection image, the catheter in the 3D blood vessels can be visualized by 3D rendering, which provides a 3D intuitive image‐based guidance for the intervention procedure.