TY - GEN
T1 - Symbolic description of 3-D structures applied to cerebral vessel tree obtained from MR angiography volume data
AU - Gerig, G.
AU - Koller, Th
AU - Székely, G.
AU - Brechbühler, Ch
AU - Kübler, O.
N1 - Funding Information:
Part of this research work was funded by COVIRA (Computer Vision in Radiology), project A2003 of the AIM (Advanced Informatics in Medicine) programme of the European Commission. Participants in the COVIRA consortium are: Philips Medical Systems, Best (NL) and Madrid (E), Corporate Research, Hamburg (D) (prime contractor); Siemens AG, Erlangen (D) and Munich (D); IBM US Scientific Centre, Winchester (UK); Gregorio Maranon General Hospital, Madrid (E); University of Tfibingen, Neuroradiology and Theoretical Astrophysics (D); German Cancer Research Centre, Heidelberg (D); University of Leuven, Neurosurgery, Radiology and Electrical Engineering (B); University of Utrecht, Neurosurgery and Computer Vision (NL); Royal Marsden Hospital/Institute of Cancer Research, Sutton (UK); National Hospital for Neurology and Neurosurgery, London (UK); Foundation of Research and
Publisher Copyright:
© Springer-Verlag Berlin Heidelberg 1993.
Copyright:
Copyright 2020 Elsevier B.V., All rights reserved.
PY - 1993
Y1 - 1993
N2 - The present paper focuses on the conversion of multidimensional image structures to art object-centered, abstract description encoding shape features and structure relationships. We describe a prototype system that extracts three-dimensional (3-D) curvilinear structures from volume image data and transforms them into a symbofic description which represents topological and geometrical features of tree-fike, filamentous objects. The initial segmentation is performed by 3-D hysteresis thresholding. A skeletal structure is derived by 3-D binary thinning, approximating the center-fines while fully preserving the 3-D topology. The local width of the fine structures is characterized by a separate 3-D Eucfidean distance transform. Compilation, or raster-to-vector transformation, converts the maximally thinned voxel fists into a vector description. The final graph data-structure encodes the spatial course of fine sections, the estimate of the local diameter, and the topology at important key locations fike branchings and end-points. The analysis system is appfied to the characterization of the cerebral vascular system segmented from magnetic resonance angiography (MRA).
AB - The present paper focuses on the conversion of multidimensional image structures to art object-centered, abstract description encoding shape features and structure relationships. We describe a prototype system that extracts three-dimensional (3-D) curvilinear structures from volume image data and transforms them into a symbofic description which represents topological and geometrical features of tree-fike, filamentous objects. The initial segmentation is performed by 3-D hysteresis thresholding. A skeletal structure is derived by 3-D binary thinning, approximating the center-fines while fully preserving the 3-D topology. The local width of the fine structures is characterized by a separate 3-D Eucfidean distance transform. Compilation, or raster-to-vector transformation, converts the maximally thinned voxel fists into a vector description. The final graph data-structure encodes the spatial course of fine sections, the estimate of the local diameter, and the topology at important key locations fike branchings and end-points. The analysis system is appfied to the characterization of the cerebral vascular system segmented from magnetic resonance angiography (MRA).
KW - 3-D Binary thinning
KW - 3-D Raster-tovector transform
KW - Cerebrospinal vasculature
KW - Magnetic resonance angiography
KW - Multidimensional image analysis
KW - Symbofic representation
UR - http://www.scopus.com/inward/record.url?scp=0002854447&partnerID=8YFLogxK
UR - http://www.scopus.com/inward/citedby.url?scp=0002854447&partnerID=8YFLogxK
U2 - 10.1007/bfb0013783
DO - 10.1007/bfb0013783
M3 - Conference contribution
AN - SCOPUS:0002854447
SN - 9783540568001
T3 - Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)
SP - 94
EP - 111
BT - Information Processing in Medical Imaging - 13th International Conference, IPMI 1993, Proceedings
A2 - Barrett, Harrison H.
A2 - Gmitro, Arthur F.
PB - Springer Verlag
T2 - 13th International Conference on Information Processing in Medical Imaging, IPMI 1993
Y2 - 14 June 1993 through 18 June 1993
ER -