3D Registration of Ultrasound Images Based on Morphology Skeleton

In order to eliminate displacement and elastic deformation between images of adjacent frames in course of 3D ultrasonic image reconstruction, elastic registration based on morphology skeleton was adopted in this paper. Feature points of connected skeleton are extracted automatically by accounting topical curvature extreme points several times. Initial registration is processed according to barycenter of skeleton. Whereafter, elastic registration based on radial basis function are processed according to feature points of skeleton. Result of example demonstrate that according to traditional rigid registration, elastic registration based on skeleton feature retain natural difference in shape for organ's different part, and eliminate slight elastic deformation between frames caused by image obtained process simultaneously. This algorithm has a high practical value for image registration in course of 3D ultrasound image reconstruction

[1]  Benoit Godbout,et al.  3D Elastic Registration of Vessel Lumen from IVUS Data on Biplane Angiography , 2003, MICCAI.

[2]  Christian Barillot,et al.  Robust statistical registration of 3D ultrasound images using texture information , 2003, Proceedings 2003 International Conference on Image Processing (Cat. No.03CH37429).

[3]  Bostjan Likar Registration and restoration of medical images , 2000 .

[4]  Joan Serrat,et al.  Motion-induced error correction in ultrasound imaging , 2002, Proceedings. First International Symposium on 3D Data Processing Visualization and Transmission.

[5]  João M. Sanches,et al.  Alignment-by-reconstruction for 3D ultrasound imaging , 2000, Proceedings 15th International Conference on Pattern Recognition. ICPR-2000.

[6]  M. Viergever,et al.  Medical image matching-a review with classification , 1993, IEEE Engineering in Medicine and Biology Magazine.