3D-2D image registration by nonlinear regression

We propose a 3D-2D image registration method that relates image features of 2D projection images to the transformation parameters of the 3D image by nonlinear regression. The method is compared with a conventional registration method based on iterative optimization. For evaluation, simulated X-ray images (DRRs) were generated from coronary artery tree models derived from 3D CTA scans. Registration of nine vessel trees was performed, and the alignment quality was measured by the mean target registration error (mTRE). The regression approach was shown to be slightly less accurate, but much more robust than the method based on an iterative optimization approach.

[1]  Sim Heng Ong,et al.  Rapid surface registration of 3D volumes using a neural network approach , 2008, Image Vis. Comput..

[2]  Cheng Wang,et al.  Image analysis by modified Legendre moments , 2007, Pattern Recognit..

[3]  J. Zhang,et al.  A neural network approach for 3D surface modeling and registration , 2004, IEEE International Workshop on Biomedical Circuits and Systems, 2004..

[4]  Ronald Chung,et al.  Viewpoint determination of image by interpolation over sparse samples , 2008, Image Vis. Comput..

[5]  L. Freire,et al.  FMRI 3D registration based on Fourier space subsets using neural networks , 2010, 2010 Annual International Conference of the IEEE Engineering in Medicine and Biology.

[6]  Simon Haykin,et al.  Neural Networks: A Comprehensive Foundation , 1998 .

[7]  Paul Suetens,et al.  Robust initialization for 2D/3D registration of knee implant models to single-plane fluoroscopy , 2007, SPIE Medical Imaging.

[8]  Stefan Klein,et al.  Alignment of 4D Coronary CTA with Monoplane X-ray Angiography , 2011, AE-CAI.

[9]  Ashutosh Kumar Singh,et al.  The Elements of Statistical Learning: Data Mining, Inference, and Prediction , 2010 .

[10]  Arthur Flexer,et al.  Statistical evaluation of neural networks experiments: Minimum requirements and current practice , 1994 .

[11]  Bostjan Likar,et al.  A review of 3D/2D registration methods for image-guided interventions , 2012, Medical Image Anal..

[12]  Aldo Maalouf,et al.  Image Registration based on Neural Network and Fourier Transform , 2006, 2006 International Conference of the IEEE Engineering in Medicine and Biology Society.

[13]  Stefan Klein,et al.  GPU accelerated alignment of 3-D CTA with 2-D X-ray data for improved guidance in coronary interventions , 2009, 2009 IEEE International Symposium on Biomedical Imaging: From Nano to Macro.

[14]  Benjamin B. Kimia,et al.  2D-3D registration based on shape matching , 2000, Proceedings IEEE Workshop on Mathematical Methods in Biomedical Image Analysis. MMBIA-2000 (Cat. No.PR00737).

[15]  Graeme P. Penney,et al.  Standardized evaluation methodology for 2-D-3-D registration , 2005, IEEE Transactions on Medical Imaging.

[16]  S.A. Banks,et al.  Accurate measurement of three-dimensional knee replacement kinematics using single-plane fluoroscopy , 1996, IEEE Transactions on Biomedical Engineering.