Model-based segmentation of cardiac and vascular images

W.J. Niessen1, C.M. van Bemmel1, A.E. Frangi1, M.J.A. Siers1, O. Wink1
1Image Sciences Institute, University Medical Center Utrecht, Netherlands

Tóm tắt

Model-based approaches towards the segmentation of vascular and cardiac images are presented. For vessel segmentation, prior shape information is introduced based on the notion that vessels are elongated structures. For cardiac segmentation, shape information derived from a training set of segmented images is incorporated in an automatically constructed point distribution model of the heart.

Từ khóa

#Image segmentation #Biomedical imaging #Active shape model #Filters #Heart #Visualization #Robustness #Deformable models #Cardiology #Image processing

Tài liệu tham khảo

10.1109/42.811279 10.1002/1522-2594(200102)45:2<311::AID-MRM1040>3.0.CO;2-7 10.1109/42.848184 10.1002/mrm.10164 van bemmel, 2001, Blood pool agent CEMRA: improved arterial visualization of the aortoiliac vasculature in the steady-state using first-pass data, MICCAI, 699 van bemmel, 2002, Blood pool agent CE-MRA: improved arterial visualization of the aortoiliac vasculature in the steady-state using first-pass data, SPIE Medical Imaging frangi, 1998, Multiscale vessel enhancement filtering, Med Image Comput Comput Assisted Intervention, 130 10.1007/BF01386390 sethian, 1999, Level set methods and fast marching methods evolving interfaces in computational geome- try fluid mechanics computer vision and material sciences 10.1109/42.925294 10.1002/mrm.1910400509 10.1109/42.906421 10.1109/42.811277 10.1016/S1361-8415(98)80009-1 10.1016/S1361-8415(01)00040-8 10.1016/S1361-8415(01)00037-8 singh, 1998, Deformable Models in Medical Image Analysis 10.1109/42.974935 10.1016/S1361-8415(96)80007-7 10.1006/cviu.1995.1004 10.1016/S0262-8856(99)00077-3 10.1109/42.811260 10.1109/ICCV.2001.937567 10.1016/S1361-8415(99)80020-6 10.1109/42.796284 frangi, 2001, Automatic 3D pdm construction via atlas-based landmarking and volumetric elastic registration, Information Processing in Medical Imaging, 78, 10.1007/3-540-45729-1_7