动态模型指导的冠脉血管三维重建
|
摘要
X射线造影作为检测心血管疾病的“金标准”一直是医生最常采用的手段之一,然而随着科学的发展,二维的X射线血管造影图像已经不能满足医学检测的要求,医生们需要更为直观的三维甚至四维的血管图像,来定量地检测血管病变的有关参数,以及心脏和血管的运动情况,从而得到更加精确和全面的检查结果。而冠状动脉血管动态的三维重建就可以解决这个问题。本文将冠状动脉血管动态的三维重建这个大问题,分成了一系列的小问题进行研究,包括冠状动脉特性视图的建立,呼吸运动的提取,以及动态模型指导的动态的血管三维重建。
文中首先提出了一种冠状动脉动态特性视图的概念。将三维投影空间分成一些小的区间,根据X射线造影原理在每个投影空间对自建的冠状动脉三维动态模型进行投影,建立冠状动脉的动态特性视图。通过各个角度的动态特性视图一方面对冠状动脉的运动有了一个更全面的了解,另一方面可以用来对造影图中的血管进行运动补偿。
在X射线造影图中,冠状动脉会由于人体的呼吸和心脏的运动而发生运动。根据呼吸运动和心脏运动各自的特点,通过傅立叶级数变换结合频域滤波的方法进行呼吸和心脏运动的分离,提取呼吸运动曲线,并对心脏的运动进行分析。
在单臂造影的图像中,由于造影图时间上的不对应无法进行血管的三维重建。通过从造影图序列中提取的呼吸运动曲线将进行重建的两幅造影图补偿到同一个呼吸运动时相,并通过心脏和冠脉血管的三维动态模型将造影图补偿到同一个心脏运动时相,然后对补偿后的造影图进行重建。实验证明呼吸运动和心脏运动补偿都达到了很好的效果。
As the‘key criterion’, X-ray angiography is the most frequently used method to examine the cardiovascular disease. With the development of the science, two-dimensional angiographic images can not fulfill the requirement of the medical detection. The doctors need the more intuitionist three-dimensional or four-dimensional cardiovascular images, to measure the pathological changes of the vessels, and movement, so as to get a more precise and all-rounded detection result, which can be solved by the dynamic 3D reconstruction of the coronary artery. In the article, we divide this big problem into some small problems, including the establishment of the dynamic characteristic views of the coronary artery, the separation of the cardiac and respiratory movement from the angiographic images, and the dynamic 3D reconstruction of the coronary directed by the dynamic model of the cardiac and coronary.
Firstly, a kind of dynamic characteristic views of the coronary model is brought out. Divide the 3D projection space into some small zones, and then project the self-constructed dynamic model of the coronary in each zone, to construct the dynamic characteristic views. From the dynamic characteristic views we have got a more profound recognition of the movement of the coronary on the one side, on the other they can be utilized for the cardiac movement compensation of the coronary arteries.
In the X-ray angiographic images, the coronary will move as the result of the human respiratory movement and cardiac contraction. According to the characteristics of the respiratory movement and cardiac contraction, spread the cardiac contraction in the Fourier series, and then filter the components of the cardiac contraction from the frequency field of the coronary image series to get the respiratory component. De-transform the respiratory movement from the frequency filed to the time-zone, we can get the respiratory movement of the coronary separately.
In the single-plane angiographic images, the 3D reconstruction of the coronary can’t be done because of the un-correspondence of the time. First using the respiratory movement retrieved from the angiographic images, compensate the two images which are used to reconstruct the 3D coronary arteries to the same phase in one respiratory period. And then compensate the two images to the same phase of the cardiac periodic movement by the dynamic model of dynamic model of the cardiac and coronary model. Finally the coronary can be reconstructed from the compensated images. Very good results of the compensation have been shown in the experiments.
文中首先提出了一种冠状动脉动态特性视图的概念。将三维投影空间分成一些小的区间,根据X射线造影原理在每个投影空间对自建的冠状动脉三维动态模型进行投影,建立冠状动脉的动态特性视图。通过各个角度的动态特性视图一方面对冠状动脉的运动有了一个更全面的了解,另一方面可以用来对造影图中的血管进行运动补偿。
在X射线造影图中,冠状动脉会由于人体的呼吸和心脏的运动而发生运动。根据呼吸运动和心脏运动各自的特点,通过傅立叶级数变换结合频域滤波的方法进行呼吸和心脏运动的分离,提取呼吸运动曲线,并对心脏的运动进行分析。
在单臂造影的图像中,由于造影图时间上的不对应无法进行血管的三维重建。通过从造影图序列中提取的呼吸运动曲线将进行重建的两幅造影图补偿到同一个呼吸运动时相,并通过心脏和冠脉血管的三维动态模型将造影图补偿到同一个心脏运动时相,然后对补偿后的造影图进行重建。实验证明呼吸运动和心脏运动补偿都达到了很好的效果。
As the‘key criterion’, X-ray angiography is the most frequently used method to examine the cardiovascular disease. With the development of the science, two-dimensional angiographic images can not fulfill the requirement of the medical detection. The doctors need the more intuitionist three-dimensional or four-dimensional cardiovascular images, to measure the pathological changes of the vessels, and movement, so as to get a more precise and all-rounded detection result, which can be solved by the dynamic 3D reconstruction of the coronary artery. In the article, we divide this big problem into some small problems, including the establishment of the dynamic characteristic views of the coronary artery, the separation of the cardiac and respiratory movement from the angiographic images, and the dynamic 3D reconstruction of the coronary directed by the dynamic model of the cardiac and coronary.
Firstly, a kind of dynamic characteristic views of the coronary model is brought out. Divide the 3D projection space into some small zones, and then project the self-constructed dynamic model of the coronary in each zone, to construct the dynamic characteristic views. From the dynamic characteristic views we have got a more profound recognition of the movement of the coronary on the one side, on the other they can be utilized for the cardiac movement compensation of the coronary arteries.
In the X-ray angiographic images, the coronary will move as the result of the human respiratory movement and cardiac contraction. According to the characteristics of the respiratory movement and cardiac contraction, spread the cardiac contraction in the Fourier series, and then filter the components of the cardiac contraction from the frequency field of the coronary image series to get the respiratory component. De-transform the respiratory movement from the frequency filed to the time-zone, we can get the respiratory movement of the coronary separately.
In the single-plane angiographic images, the 3D reconstruction of the coronary can’t be done because of the un-correspondence of the time. First using the respiratory movement retrieved from the angiographic images, compensate the two images which are used to reconstruct the 3D coronary arteries to the same phase in one respiratory period. And then compensate the two images to the same phase of the cardiac periodic movement by the dynamic model of dynamic model of the cardiac and coronary model. Finally the coronary can be reconstructed from the compensated images. Very good results of the compensation have been shown in the experiments.
引文
[1]张泽宝.医学影像物理学.第一版.北京:人民卫生出版社, 2000
[2]王国栋.血管造影图像处理与血管三维重建.华中科技大学博士毕业论文, 2008
[3] I. Liu and Y. Sun. Fully automated reconstruction of 3-D vascular tree structures from two orthogonal views using computational algorithms and production rules. Opt. Eng., 1992, 31(10): 2197-2207
[4]黄家祥,郁道银,陈晓东等.冠状动脉树结构的三维重建.中国生物医学工程学报, 2005, 23(2)
[5] Claire Chalopin, Ge′rard Finet, Isabelle E. Magnin, 2001. Modeling the 3D coronary tree for labeling purposes. Medical Image Analysis, 2001(5): 301-315
[6] Renaudin C. P., Barbier B., Roriz R., et al. Coronary arteries: new design for three-dimensional arterial phantoms. Radiology, 1994, 190: 579-582
[7] G. Coppini, M. Demi, R. Mennini et al. 3-D knowledge driven reconstruction of coronary trees. Med. Biol. Eng. Comput., 1991: 535-542
[8]肖晶.静态模型指导的血管三维重建.华中科技大学硕士毕业论文, 2009
[9] V. Rasche, A. Buecker, M. Grass, et al. ECG-gated 3-D rotational coronary angiography (3-DRCA), in Proc. Computer Assisted Radiology and Surgery(CAR 02), 2002: 827-831
[10] M. E. Olszewski, R. M. Long, S. C. Mitchell, et al. A Quantitative analysis of 3-D coronary Vasculature in Four Dimensions, Proceedings of the 22nd Annusl EMBS International Conference, Chicago IL, 2000: 2621-2624
[11] B. Movassaghi, V. Rasche, M. A. Viergever, et al. Quantitative analysis of 3-D coronary modeling in 3-D rotational X-ray imaging, presented at the IEEE Nuclear Science Symp. (NSS)Medical Imaging Conf. (MIC), 2002: 878-880
[12] T. Saito, M. Misaki, K. Sherato, et al. Three-dimensional quantitative coronary angiography. IEEE Trans. Biomed. Eng, 1990: 768-777
[13] A. Wahle, E. Wellnhofer, I. Mugaragu, et al. Assessment of diffuse coronary artery disease by quantitative analysis of coronary morphology based upon 3-D reconstruction from biplane angiograms. IEEE Trans. Med. Imag, 1995, 14: 230-241
[14] B. Movassaghi, S. Young, V. Rasche. An accurate coronary modeling procedure using 2-D calibrated projections based on extracted 3-D vessel centerlines, in Proc. Computer Assisted Radiology and Surgery(CARS 2003), 2003: 1397-1401
[15] S. Young, B. Movassaghi, J. Weese, et al. 3-D vessel axis extraction using 2-D calibrated X-ray projections for coronary modeling, in Proc. SPIE Medical Imaging: Image Processing, San Diego, CA, 2003, 5032: 1491-1498
[16] A. Navab, A. Bani-Hashemi, M. Mitshke. Dynamic geometrical calibration for 3-D cerebral angiography, in Proc. SPIE Medical Imaging; Image Processing, Newport Beach, CA, 1996, 2708: 361-370
[17] A. Rougee, C. Picard, C. Ponchut, et al. Geometrical calibration of X-ray imaging chains for three-dimensional reconstrution, Comput. Med. Imag. Graphics, 1993, 17: 295-300
[18] K. R. Hoffmann, J. Esthappan, S. Li, et al. A simple technique for calibraing imaging geometries, in Proc. SPIE Medical Imaging: Image Processing, Newport Beach, CA, 1996, 2708: 371-375
[19] D. G. W. Onnasch, G. P. M. Prause. Geometric image correction and iso-center calibration at oblique biplane angiographic views, in Proc. IEEE Computer in Cardiology, 1992: 647-650
[20] M. Mitschke, N. Navab. Optimal configuration for dynamic calibration of projection geometry of X-ray C-arm systems, in Proc. MMBIA, June 2000: 204-209
[21] D. A. Reimann, M. J. Flynn. Automatied distortion correction of X-ray image intensifier images, in IEEE Nuclear Science Symp. Medical Imaging Conf. Rec, Orlando, FL, Oct. 1993: 1339-1341
[22] L. Lannay. Quantitative evaluation of an algorithm for correcting geometrical distorsions in DSA images: Application to stereotaxy, in SPIE, San Diego, CA, Feb. 1995, 2434: 520-529
[23] E. Gronenschild. The accuracy and reproducibility of a global to correct for geometric image distortion in the X-ray imaming chain, Med. Pys, 1997, 24(2): 1875-1888
[24] Z. Zhang, R. Deriche, O. Faugeras, et al. A robust technique for matching two uncalibrated images through the recovery of the unkown epipolar geometry. INRIA, Sophia-Antipolis, France, Tech. Rep. RR-227, 1994
[25] Z. Zhang. A flexible new technique for camera calibration, IEEE Trans. Pattern Anal. Machine Intell, 2000, 22(10): 1330-1334
[26] R. Curwen, A. Amini, J. Duncan, et al. Tracking vascular motion in X-ray image sequences with Kalman snakes, Coput. Cardiol., 1994: 109-112
[27] M. Pierre, D. Jolly, C. Liang, A. Gupta. Optimal polyline tracking for artery motion compensation in coronary angiography, in Proc. ICCV, 1998: 414-419
[28] J. H. Rong, J. L. Coatrieux, R. Collorec. Motion estimation in digital subtraction angiography, in Proc. IEEE Eng. Med. Biol. Soc. Conf., 1989: 567-568
[29] S. Ruan, A. Bruno, R. Collorec, et al. 3-D motion and reconstruction of coronary networks, Proc. IEEE EMBS, 1992, 5: 2048-2049
[30] S. Ruan, A. Bruno, J. L. Coatrieux. 3-D motion and reconstruction of coronary arteries from biplane cineangiography, Image Vision Comput., 1994, 12(10): 683-689
[31] R. Toledo, P. Radeva, C. Von Land, et al. 3-D dynamic model of the coronary tree, Comput. Cardiol., 1998, 25: 777-780
[32] Guy Shechter*, Jon R. Resar, Elliot R. McVeigh. Displacement and Velocity of the Coronary Arteries: Cardiac and Respiratory Motion, IEEE Transactions on Medical Imaging, 2006, 25: 369-375
[33] M. Sermesant, H. Delingette, N. Ayache. An Electromechanical Model of the Heart for Image Analysis and Simulation, IEEE Transactions on Medical Imaging, 2006, 25: 612-625
[34] Guy Shechter*, Barak Shechter, Jon R. Resar, et al. Prospective Motion Correction of X-Ray Images for Coronary Interventions. 2005, 24: 441-450
[35] S. Zhang, G. D. Sullivan, K. D. Baker. The automatic construction of a view-independent relational model for 3D object recognition. IEEE Trans. PAMI, 1993, 15(6): 531-544
[36] S. Chen, H. Freeman. The dominant views of solid objects. 11th Int. Conf. Pattern Recogn., IEEE, New York, 1992, 1: 332-336
[37] O. Munkelt. Aspect-trees: generation and interpretation. CVIU, 1995, 61(3): 365-386
[38] Z. Gigus, J. Malik. Computing the aspect graph for line drawings of polyhedral objects. IEEE Trans. PAMI, 1990, 12(2): 113-122
[39] D. W. Eggert. The scale space aspect graph. IEEE Trans. PAMI, 1993, 15(11): 1114-1129
[40] A. M. Robey, G. A. W. West, S. Venkatesh. An Investigation into the use of physical modeling for the prediction of various feature types visible from different view points. in second CAD-Based Vision Workshop, IEEE Computer Society Press, Los Alamitos, CA, 1994: 282-290
[41]刘芳.冠状动脉模型的建立和分析.华中科技大学硕士毕业论文. 2007
[42] M. Sonka, M. D. Winniford, S. M. Collins. Robust simultaneous detection of coronary borders in complex images. IEEE Trans. Med. Imaging, 1995, 14(1): 151-161
[43] Y. Wang, S. Riederer, R. Ehman. Respiratory motion of the heart: kinematics and the implications for the spatial resolution in coronary imaging, Magnetic Resonance in Medicine, 1995, 33(5): 713- 719
[44]李昭.心脏三维及四维数学建模.华中科技大学硕士毕业论文, 2007
[45]郑君里,应启珩,杨为理.信号与系统.第二版.北京:高等教育出版社, 2000
[2]王国栋.血管造影图像处理与血管三维重建.华中科技大学博士毕业论文, 2008
[3] I. Liu and Y. Sun. Fully automated reconstruction of 3-D vascular tree structures from two orthogonal views using computational algorithms and production rules. Opt. Eng., 1992, 31(10): 2197-2207
[4]黄家祥,郁道银,陈晓东等.冠状动脉树结构的三维重建.中国生物医学工程学报, 2005, 23(2)
[5] Claire Chalopin, Ge′rard Finet, Isabelle E. Magnin, 2001. Modeling the 3D coronary tree for labeling purposes. Medical Image Analysis, 2001(5): 301-315
[6] Renaudin C. P., Barbier B., Roriz R., et al. Coronary arteries: new design for three-dimensional arterial phantoms. Radiology, 1994, 190: 579-582
[7] G. Coppini, M. Demi, R. Mennini et al. 3-D knowledge driven reconstruction of coronary trees. Med. Biol. Eng. Comput., 1991: 535-542
[8]肖晶.静态模型指导的血管三维重建.华中科技大学硕士毕业论文, 2009
[9] V. Rasche, A. Buecker, M. Grass, et al. ECG-gated 3-D rotational coronary angiography (3-DRCA), in Proc. Computer Assisted Radiology and Surgery(CAR 02), 2002: 827-831
[10] M. E. Olszewski, R. M. Long, S. C. Mitchell, et al. A Quantitative analysis of 3-D coronary Vasculature in Four Dimensions, Proceedings of the 22nd Annusl EMBS International Conference, Chicago IL, 2000: 2621-2624
[11] B. Movassaghi, V. Rasche, M. A. Viergever, et al. Quantitative analysis of 3-D coronary modeling in 3-D rotational X-ray imaging, presented at the IEEE Nuclear Science Symp. (NSS)Medical Imaging Conf. (MIC), 2002: 878-880
[12] T. Saito, M. Misaki, K. Sherato, et al. Three-dimensional quantitative coronary angiography. IEEE Trans. Biomed. Eng, 1990: 768-777
[13] A. Wahle, E. Wellnhofer, I. Mugaragu, et al. Assessment of diffuse coronary artery disease by quantitative analysis of coronary morphology based upon 3-D reconstruction from biplane angiograms. IEEE Trans. Med. Imag, 1995, 14: 230-241
[14] B. Movassaghi, S. Young, V. Rasche. An accurate coronary modeling procedure using 2-D calibrated projections based on extracted 3-D vessel centerlines, in Proc. Computer Assisted Radiology and Surgery(CARS 2003), 2003: 1397-1401
[15] S. Young, B. Movassaghi, J. Weese, et al. 3-D vessel axis extraction using 2-D calibrated X-ray projections for coronary modeling, in Proc. SPIE Medical Imaging: Image Processing, San Diego, CA, 2003, 5032: 1491-1498
[16] A. Navab, A. Bani-Hashemi, M. Mitshke. Dynamic geometrical calibration for 3-D cerebral angiography, in Proc. SPIE Medical Imaging; Image Processing, Newport Beach, CA, 1996, 2708: 361-370
[17] A. Rougee, C. Picard, C. Ponchut, et al. Geometrical calibration of X-ray imaging chains for three-dimensional reconstrution, Comput. Med. Imag. Graphics, 1993, 17: 295-300
[18] K. R. Hoffmann, J. Esthappan, S. Li, et al. A simple technique for calibraing imaging geometries, in Proc. SPIE Medical Imaging: Image Processing, Newport Beach, CA, 1996, 2708: 371-375
[19] D. G. W. Onnasch, G. P. M. Prause. Geometric image correction and iso-center calibration at oblique biplane angiographic views, in Proc. IEEE Computer in Cardiology, 1992: 647-650
[20] M. Mitschke, N. Navab. Optimal configuration for dynamic calibration of projection geometry of X-ray C-arm systems, in Proc. MMBIA, June 2000: 204-209
[21] D. A. Reimann, M. J. Flynn. Automatied distortion correction of X-ray image intensifier images, in IEEE Nuclear Science Symp. Medical Imaging Conf. Rec, Orlando, FL, Oct. 1993: 1339-1341
[22] L. Lannay. Quantitative evaluation of an algorithm for correcting geometrical distorsions in DSA images: Application to stereotaxy, in SPIE, San Diego, CA, Feb. 1995, 2434: 520-529
[23] E. Gronenschild. The accuracy and reproducibility of a global to correct for geometric image distortion in the X-ray imaming chain, Med. Pys, 1997, 24(2): 1875-1888
[24] Z. Zhang, R. Deriche, O. Faugeras, et al. A robust technique for matching two uncalibrated images through the recovery of the unkown epipolar geometry. INRIA, Sophia-Antipolis, France, Tech. Rep. RR-227, 1994
[25] Z. Zhang. A flexible new technique for camera calibration, IEEE Trans. Pattern Anal. Machine Intell, 2000, 22(10): 1330-1334
[26] R. Curwen, A. Amini, J. Duncan, et al. Tracking vascular motion in X-ray image sequences with Kalman snakes, Coput. Cardiol., 1994: 109-112
[27] M. Pierre, D. Jolly, C. Liang, A. Gupta. Optimal polyline tracking for artery motion compensation in coronary angiography, in Proc. ICCV, 1998: 414-419
[28] J. H. Rong, J. L. Coatrieux, R. Collorec. Motion estimation in digital subtraction angiography, in Proc. IEEE Eng. Med. Biol. Soc. Conf., 1989: 567-568
[29] S. Ruan, A. Bruno, R. Collorec, et al. 3-D motion and reconstruction of coronary networks, Proc. IEEE EMBS, 1992, 5: 2048-2049
[30] S. Ruan, A. Bruno, J. L. Coatrieux. 3-D motion and reconstruction of coronary arteries from biplane cineangiography, Image Vision Comput., 1994, 12(10): 683-689
[31] R. Toledo, P. Radeva, C. Von Land, et al. 3-D dynamic model of the coronary tree, Comput. Cardiol., 1998, 25: 777-780
[32] Guy Shechter*, Jon R. Resar, Elliot R. McVeigh. Displacement and Velocity of the Coronary Arteries: Cardiac and Respiratory Motion, IEEE Transactions on Medical Imaging, 2006, 25: 369-375
[33] M. Sermesant, H. Delingette, N. Ayache. An Electromechanical Model of the Heart for Image Analysis and Simulation, IEEE Transactions on Medical Imaging, 2006, 25: 612-625
[34] Guy Shechter*, Barak Shechter, Jon R. Resar, et al. Prospective Motion Correction of X-Ray Images for Coronary Interventions. 2005, 24: 441-450
[35] S. Zhang, G. D. Sullivan, K. D. Baker. The automatic construction of a view-independent relational model for 3D object recognition. IEEE Trans. PAMI, 1993, 15(6): 531-544
[36] S. Chen, H. Freeman. The dominant views of solid objects. 11th Int. Conf. Pattern Recogn., IEEE, New York, 1992, 1: 332-336
[37] O. Munkelt. Aspect-trees: generation and interpretation. CVIU, 1995, 61(3): 365-386
[38] Z. Gigus, J. Malik. Computing the aspect graph for line drawings of polyhedral objects. IEEE Trans. PAMI, 1990, 12(2): 113-122
[39] D. W. Eggert. The scale space aspect graph. IEEE Trans. PAMI, 1993, 15(11): 1114-1129
[40] A. M. Robey, G. A. W. West, S. Venkatesh. An Investigation into the use of physical modeling for the prediction of various feature types visible from different view points. in second CAD-Based Vision Workshop, IEEE Computer Society Press, Los Alamitos, CA, 1994: 282-290
[41]刘芳.冠状动脉模型的建立和分析.华中科技大学硕士毕业论文. 2007
[42] M. Sonka, M. D. Winniford, S. M. Collins. Robust simultaneous detection of coronary borders in complex images. IEEE Trans. Med. Imaging, 1995, 14(1): 151-161
[43] Y. Wang, S. Riederer, R. Ehman. Respiratory motion of the heart: kinematics and the implications for the spatial resolution in coronary imaging, Magnetic Resonance in Medicine, 1995, 33(5): 713- 719
[44]李昭.心脏三维及四维数学建模.华中科技大学硕士毕业论文, 2007
[45]郑君里,应启珩,杨为理.信号与系统.第二版.北京:高等教育出版社, 2000