Segmentation automatique application aux angioscanners 3D du foie
Automatical Segmentation : Application to 3D Angiograms of the Liver
OPEN ACCESS
As part of a hepatic surgery simulator, we have developed a new method for the extraction of the portal vein's vascular tree in 3D liver angioscanners. In practice, this tree is used to localize the different anatomical segments that correspond to the unit of surgical ablation of the liver. Our method thus facilitates the surgeon's task by automatically giving the 3D model of the portal vein in a three-step segmentation . The first step reduces the image to the ROI defined by the liver contours and increases its quality by an anisotropic filtering . The second step performs the segmentation of vascular networks by a global thresholding followed by a local analysis. The third step translates a priori knowledge in topological and geometrical constraints . This last step allows to remove mistakes due to the anisotropy of the images by disconnecting the different vascular trees in order to extract the portal vein. Results on 12 patients, validated by a radiologist, showed that the algorithm automatically extracts the principal branches of the portal vein, allowing to delimit the anatomical segments defined in the conventional liver anatomy.
Résumé
Dans le cadre de la réalisation d'un simulateur de chirurgie Iaparoscopique t du foie, nous avons développé une nouvelle méthode permettant d'extraire dans les angioscanners 3D du foie, le réseau vasculaire de la veine porte . Ce réseau est utilisé en pratique pour repérer les différents segments anatomiques, qui représentent l'unité d'intervention dans les exérèses' du foie. Notre méthode facilite ainsi la tâche des chirurgiens, en leur fournissant automatiquement le modèle 3D de la veine porte par une segmentation décomposée en trois étapes. La première étape réduit l'image à la région d'intérêt correspondant au contour du foie et améliore sa qualité en réalisant un filtrage anisotrope . La seconde segmente les réseaux vasculaires et appliquant un seuillage global, suivi d'une analyse locale. La troisième étape traduit les connaissances a priori que nous avons des réseaux vasculaires, en contraintes topologiques et géométriques . Cette dernière étape permet de corriger les problèmes résultant de l'anisotropie des images, en déconnectant les différentes arborescences du foie pour en extraire la veine porte . Les résultats obtenus sur douze patients, et vérifiés par un radiologue, montrent que l'algorithme extrait automatiquement les principales branches de la veine porte, permettant de délimiter les segments anatomiques définis dans l'anatomie conventionnelle du foie.
Segmentation, angioscanner 3D, liver, portal vein, vessels, thresholding, digital topology
Mots clés
Segmentation, angioscanner 3D, foie, veine porte, vascularisation, seuillage, topologie discrète
[1] M .S . Atkins and B .T. Mackiewich . Automatic Segmentation of the Brain in MRI. In K .H . Höhne, and R . Kikinis, editors, Visualization in Biomedical Computing, pages 241-246 . Springer Verlag, 1996 . Lecture Notes in Computer Science, Vol . 1131.
[2] K .T.Bae and M .L . Giger. C .-T. Chen and C .E . Kahn . Automatic segmentation of liver structure in CT images. Medical Physics, 20 (1) : 71-78, January 1993 .
[3] E . Bellon, M . Feron, F. Maes, L. Van Hoe, D . Delaere, F. Haven, S . Sunaert , A.L. Baert, G . Marchai, P. Suetens, Evaluation of manual vs semi-automated delineation of liver lesions on CT images . European Radiology, 7 : 432-438 , 1997 .
[4] G. Bertrand, G . Malandain, A new characterization of three-dimensional simple points, Pattern Recognition Letters, 15(2) : 169-175, February 1994 .
[5] M .E . Brummer, R .M . Mersereau, R.L. Eisner, R .J. Lewine. Automatic Detection of Brain Contours in MRI Data Sets,IEEE Trans. on Medical Imaging, 12(2) : 153-166, June 1993 .
[6] M .J . Carreira, D . Cabello, M .G . Penedo, J .M . Pardo, Computer-aided Lung Nodule Detection in Chest Radiography . In R . Chin, H Ip, A. Naiman, and T.-C . Pong, editors, Image Analysis Applications and Computer Graphics , pages 331-338 . Springer Verlag, 1995 . Lecture Notes in Computer Science , Vol . 1024.
[7] J .-S . Chou, S .-Y. j . Chen, G.S . Sudakoff, K .R . Hoffmann, C .-T. Chen, A .H . Dachman . Image fusion for visualization of hepatic vasculature and tumors . In M .H . Loew, editor, Medical Imaging 1995 : Image Processing, pages 157 -163 . SPIE, May 1995 . Proceedings Vol. 2434 .
[8] J.L . Coatrieux, M . Garreau, R . Collorec, C . Roux, Computer Vision Approaches for the Three-Dimensional Reconstruction of Coronary Arteries : Review and Prospects, Critical Reviews in Biomedical Engineering, 22(1) : 1-38, 1994 .
[9] Couinaud . Le foie, études anatomiques et chirurgicales, Masson, 1957 .
[10] H. Delingette, Modélisation, déformation et reconnaissance d'objets tridimensionnels à l'aide de maillages simplexes . PhD thesis, École Centrale de Paris, July 1994 .
[11] T.E . Dufresne, A . Sarwal, and A .P. Dhawan, A Grey-Level Thinning Method for Delineation and Representation of Arteries, Computerized Medical Imaging and Graphics, 18(5) : 343-355, 1994 .
[12] J .J . Everhart, T.M . Cannon, J . Newell,and D . Linch, Image Segmentation applied to CT examination of lymphangioleiomyomatosis, M .H. Loew, editor , Medical Imaging 1994 : Image Processing, pages 87-95, SPIE, may 1994 . Prooceeding Vol . 2167 .
[13] J .H.D Fasel, D . Selle, C .J .G . Evertsz, F. Terrier, H.O . Peitgen, and P. Gailloud , Segmental Anatomy of the Liver : Poor correlation with CT, Radilogy, 206 : 151-156, 1998 .
[14] E.K . Fishman, B .S . Kuszyk, D .G . Heath, and L. Gao, Surgical planning for liver resection, IEEE Computer, Innovative Technology for Computer Professionals, 29(1) : 64-72, January 1996 .
[15] L . Gao, D .G . Heath, B .S . Kuszyk and E .K . Fishman, Automatic Liver Segmentation Techniques for Three-Dimensional Visualization of CT Data , Radiology, 201(2) : 359-364, November 1996 .
[16] G . Gerig, O . Kühler, R . Kikinis, and F.A. Jolesz, Nonlinear Anisotropie Filtering of MRI Data,IEEE Trans . on Medical Imaging . 11(2) : 221-232 , June 1992 .
[17] G . Gerig, J . Martin, R . Kikinis, O . Kühler, M . Shenton and F.A . Jolesz, Unsupervised tissue type segmentation of 3D dual-echo MR head data, Image and Vision Computing, 10(6) : 349-360, August 1992.
[18] A. Goshtasby, and W.D. O'Neill, Curve Fitting by a Sum of Gaussians , CVGIP : Graphical Models and Image Processing, 56(4) : 281-288, July 1994.
[19] D .G . Heath,P.A . Soyer,B .S . Kuszyk, D .F. Bliss, P.S . Calboun, D .A. Bluemke, M.A . Choti and E.K. Fishman, Three-dimensional Spiral CT during Arterial Portography : Comparison of three rendering techniques, Radiographics, 15(4) : 1001-1011, July 1995
[20] G.T. Herman G .T. J . Zheng and C.A. Bucholtz, Shape-Based Interpolation IEEE Computer Graphics and Applications, 12 : 69-79, May 1992.
[21] K .R. Hoffmann, S .-Y j . Chen and Kormano, Segmentation and display of hepatic vessels and metastases . In M .H. Loew, editor, Medical Imaging 1993 : Image Processing, pages 263-270 . SPIE, September 1993 . Proceedings Vol. 1898 .
[22] N . Inaoka, H . Suzuki and M . Fukuda, Hepatic Blood Vessels Recognition using Anatomical Knowledge . In M .H . Loew, editor, Medical Imaging VI : Image Processing, pages 509-513 . SPIE, June 1992 . Proceedings Vol. 1652 .
[23] R .C . Jansen, K. Reinink, and G .W.A .M. van der Heijden . Analysis of Grey Level histograms by using statistical methods for mixtures of distribution Pattern Recognition Letters, 14 : 585-590, 1993 .
[24] V.A . Kovalev, Rule-Based Method for tumor Recognition in Liver Ultrasonic Images . In C. Braccini, L. DeFloriani, and G . Vernazza, editors Image Analysis and Processing, pages 217-22 . Springer Verlag, September 1995 . Lecture Notes in Computer Science, Vol . 974 .
[25] K . Krissian, G. and Malandain, and N . Ayache, Directional Anisotropic Diffusion Applied to Segmentation of Vessels in 3D Images, Tecnical Report RR-3064, Inria, December 1996 .
[26] K . Krissian, G . Malandain, and N. Ayache, Directional Anisotropic Diffusion Applied to Segmentation of Vessels in 3D Images . In B . Haar Romeny, L . Florack, J. Koenderink, and M . Viergever, editors, Scale-Space Theory in Computer Vision, pages 345-348 . Springer Verlag, july 1997 . Lecture Notes in Computer Science, Vol . 1252 .
[27] T. Langenberger, R.A . Meuli, P. Schroeter, P. Maeder, J .M . Vesin, and P. Schnyder, Automatic Segmentation and Volume measuring of brain with MRI . In H .U. Lemke and al ., editors, Computer Assisted Radiology, pages 219-224 . Elsevier Science B .V.,1996 .
[28] G . Malandain, G . Bertrand, and N. Ayache, Topological Segmentation of Discrete Surfaces . International Journal of Computer Vision, 10(2) : 183-197 , February 1993 .
[29] Y. Masutani, Y. Yamauchi, M . Suzuki, Y. Ohta, T. Dohi, M. Tsuzuki, M . and D . Hashimoto, Development of interactive vessel modelling system for hepatic vasculature from MR images Medical and Biomedical Engineering and Computing, 33 : 97-101, january 1995 .
[30] S . Matsushita, H . Oyamada, M . Kusakabe,and N . Suzuki, Attempt to extract 3-D image of liver automatically out of Abdominal MRI . In M .H. Loew, editor, Medical Imaging 1993 : Image Processing, pages 803-808 . SPIE, September 1993 . Proceedings Vol. 1898 ,
[31] F. Meyer, Cytologie quantitative et morphologie mathématique, PhD thesis . École des mines, may 1979 .
[32] J . Montagnat, and H . Delingette, Volumetric Medical Images Segmentation using Shape Constrained Deformable Models . In J. Troccaz, E. Grimson, and R . Mösges, editors,CVRMed-MRCAS, pages 13-22 . Springer verlag, march 1997 . Lecture Notes in Computer Science, Vol . 1205 .
[33] J .F. O'Brien, N .F. Ezquerra, Automated Segmentation of Coronary Vessels in Angiographie Image Sequences Utilizing Temporal, Spatial and Structural Constraints . In R.A . Robb, editor, Visualization in Biomedical Computing , pages 25-37. SPIE, September 1994 . Proceeding Vol . 2359 .
[34] P. Perona, and J . Malik, Scale-space and edge detection using anisotropic diffusion, IEEE Trans . on Pattern Analysis and Machine Intelligence, 12(7) : 629-639, july 1990 .
[35] P. Perona, T. Shiota, and J. Malik, Anisotropic Diffusion. In B .M . ter Haar Romeny, editor, Geometry-Driven Diffusion in Computer Vision, pages 73-92 . Kluwer academic, 1994 .
[36] J . Piper, and E . Granum, Computing distance transformations in convex and non-convex domains, Pattern Recognition, 20(6) : 599-615, 1987 .
[37] W.H . Press, B .P. Flannery, S .A . Teukolsky, and W.T. Vetterling, Numerical Recipes in C. Cambridge University Press,1988 .
[38] A .R . Sanderson, E. Cohen, T.C . Henderson, and D .L . Parker, Vascular Models : From Raw Data to geometric Models . In AAAI'94, Applications of Computer Vision in Medical Image Processing, pages 141-144, Stanford, CA, march 1994 . Spring Symposium series .
[39] A .R . Sanderson, D .L. Parker, and T.C. Henderson, Simultaneous Segmentation of MR and X-ray Angiograms for Visualization of Cerebral Vascular Anatomy. In M .A . Viergever, editor, International Conference on Volume Image Processing, pages 11-14, Utrecht, The Netherlands, June 1993.
[40] M . Schmitt, and J . Mattioli, Morphologie mathématique, Masson, 1994 .
[41] P. Schroeter . Unsupervised 2-D and 3-D Image Segmentation . PhD Thesis , École Polytechnique Fédérale de Lausanne, 1996 .
[42] J. Serra . Image analysis and mathematical morphology, Academic Press , 1982.
[43] M . Sonka, G . Sundaramoorthy, E .A . and Hoffman . Knowledge-Based Segmentation of Intrathoracic Airways from Multidimensional High Resolution CT Images . In E .A. Hoffman, and R .S . Acharya, editors, Medical Imaging 1994 : Physiology and Function from Multidimensional Images, pages 73-85 . SPIE, May 1994 . Proceedings vol . 2168 .
[44] P. Soyer, D . Heath, D .A. Bluemke, M .A . Choti, J.E . Kuhlman, J .E . R. Reichte, and E .K. Fishman . Three-Dimensionnal Helical CT of intrahepatic Venous structures : Comparison of three rendering techniques, Journal of Computer Assisted Radiography, 20(1) : 122-127, 1996 .
[45] Ying Sun, Richaard Lucariello, and Salvatore A . Chiaramida. Directional Low-Pass Filtering for Improved Accuracy and Reproducibility of Stenosis Quantification in Coronary Arteriograms,IEEE Trans. on Medical Imaging, 14(2) : 242-248, June 1995 .
[46] C . Toumoulin, R. Collorec, J .L. Coatrieux, Vascular network segmentation in subtraction angiograms : a comparative study, Medical Information, 15(4) : 333-341, 1990 .
[47] J.M. Tubiana, J.P. Deutsch, J. Taboury, and B . Martin . Imagerie des métastases hépatiques des cancers colorectaux . Diagnostic et résécabilité . In B . Nordlinger and D . Jaek, editors, Traitement des métastases hépatiques des cancers colorectaux, pages 61-75 . Springer-Verlag, September 1992 .
[48] W.M. Wells, W.E .L . Grimson, R. Kikinis, and F.A. Jolesz . Adaptative Segmentation of MRI Data. In N. Ayache, editor, Computer Vision, Virtual Reality and Robotics in Medecine, pages 59-69. Springer Verlag, 1995 . lecteureNotes in Computer Science, Vol . 905 .
[49] C . Zahlten, H . Jürgens, C .J .G . Evertsz, R. Leppek, H .-O . Peithen, and K .J . Klose . Portal Vein Reconstruction Based on Topology, European Journal of Radiology, 19 : 96-100, 1995 .
[50] C. Zahlten, H . Jürgens, H .-O . Peitgen, Reconstruction of branching blood vessels from CT-data, In R. Scateni, J .J . van Wijk, and P. Zanarini, editors Workshop of Visualization in Scientific Computing, pages 41-523 .Springer verlag, 1995 . Eurographics .