Modélisation et segmentation d’images échographiques endovasculaires - Modelization and segmentation of intravascular ultrasound images

Modélisation et segmentation d’images échographiques endovasculaires

Modelization and segmentation of intravascular ultrasound images

Frédéric Guérault Philippe Delachartre  Gérard Finet  Isabelle E. Magnin 

SIMAG Développement, 13397 Marseille Cedex 20, France

CREATIS, UMR CNRS 5515, affilié à l’INSERM, 69621 Villeurbanne Cedex, France

Hôpital Cardiologique de Lyon, Laboratoire d'hémodynamique, BP Lyon Monchat, 69394 lyon cedex 03

Corresponding Author Email: 
guerault@simag.fr
Page: 
517-529
|
Received: 
26 April 2000
| |
Accepted: 
N/A
| | Citation

OPEN ACCESS

Abstract: 

We present a new algorithm for the segmentation of intravascular ultrasound images. The model used to describe the endoluminal section of the artery takes into account the granular aspect of the texture and all the geometric deformations inherent in intravascular ultrasound images. The segmentation algorithm is developped in the framework of the statistical theory of estimation. It is adapted to the image model and allows to determine all the concentric contours of the artery section. The use of a specific image model limits the field of possible solutions, and allows to obtain a performant and reliable algorithm. We illustrate the performances of this algorithm on synthetic images and on real intravascular ultrasound images.

Résumé

Nous présentons un nouvel algorithme pour la segmentation d’images échographiques endovasculaires. Le modèle introduit pour décrire la section endoluminale de l’artère prend en compte l’aspect granulaire des textures et l’ensemble des déformations géométriques inhérentes aux images échographiques endovasculaires. L’algorithme de segmentation est développé dans le cadre général de la théorie statistique de l’estimation. Il est adapté au modèle d’image et permet de déterminer sur les images échographiques endovasculaires les différents contours concentriques de la section artérielle. Le modèle utilisé permet de restreindre fortement le domaine des solutions admissibles, ce qui permet d'obtenir un algorithme performant et fiable. Nous illustrons les performances de cet algorithme sur des images synthétiques et sur des images ultrasonores endovasculaires réelles acquises chez l’homme.

Keywords: 

Deformable modele, maximum likelihood, segmentation, ultrasound image, artifact

Mots clés

Modèle déformable, maximum de vraisemblance, segmentation, image ultrasonore, artefact

1. Introduction
2. Nature De l’image Échographique
3. Modélisation De I’image
4. Estimation Statistique De La Forme
5. Résultats Expérimentaux
6. Conclusion Et Perspectives
  References

[1] G. Finet, Imagerie échographique endovasculaire : Validations et traitements numériques, Thèse de Doctorat, Univ. Claude Bernard - Lyon I, 1994, p. 284.

[2] M. Kass, A. Witkin, and D. Terzopoulos, Snakes: Active Contour Models, International Journal of Computer Vision, Vol. 1, pp. 321-331, 1988.

[3] R. Ronfard, Region-Based Strategies for Active Contour Models, International Journal of Computer Vision, Vol. 13(2), pp. 229-251, 1994.

[4] C. Kervrann and F. Heitz, Statistical deformable model-based segmentation of image motion, IEEE Trans. on Image Processing, Vol. 8(4), pp. 593-588, 1999.

[5] G. Storvik, A bayesian approach to dynamic contours through stochastic sampling and simulated annealing, IEEE Trans PAMI, Vol. 16, pp. 976-986, 1994.

[6] A.K. Jain, Y. Zhong, and S. Lakshmanan, Object Matching Using Deformable Templates, IEEE PAMI, Vol. 18(3), pp. 267-278, 1996.

[7] R.O. Duda and P.E. Hart, Pattern classification and scene analyse. John Willey and sons ed. 1973, New York.

[8] O. Germain and Ph. Réfrégier, Optimal snake-based segmentation of a random luminance target on a spatially disjoint background, Opt. Lett., Vol. 21 (22), pp. 1845-1847, 1996.

[9] Ph. Réfrégier, O. Germain, and T. Gaidon, Optimal snake segmentation of target and background with independent Gamma density probabilities, application to speckled and preprocessed images, Opt. Com., Vol. 137, pp. 382-388, 1997.

[10] C. Chesnaud and Ph. Réfrégier. Optimal snake region based segmentation for different physical noise models and fast algorithm implementation. in Conference PSIP'99 (Physics in Signal and Image Processing). 1999. Paris.

[11] F. Lizzi and e. al, Relationship of ultrasonic spectral parameters to features of tissue microstructure, IEEE Trans. Ultrason. Ferroelect., Freq. Contr., Vol. 33, pp. 319-329, 1986.

[12] R. Wagner, M. Insana, and D. Brown, Statistical properties of radio-frequency and envelope detected signals with applications to medical ultrasound, J. Opt. Soc. Am., Vol. 4, pp. 910-922, 1987.

[13] N. Friedland and D. Adam, Automatic ventricular boundary detection from sequential ultrasound images using simulated annealing, IEEE Trans. Med. Imag., Vol. 4, pp. 344-353, 1989.

[14] J.M.B. Dias and J.M.N. Leitào, Wall Position and Thickness Estimation from Sequences of Echocardiographic Images, IEEE Trans. on Medical Imaging, Vol. 15(1), pp. 25-38, 1996.

[15] Ph. Delachartre, C. Cachard, G. Finet, L. Gerfault, and D. Vray, Modeling geometric artefacts in intravascular ultrasound imaging, Ultrasound in Med. Biol., Vol. 25(4), pp. 567-575, 1999.

[16] G. Finet, C. Cachard, Ph. Delachartre, E. Maurincomme, and J. Beaune, Artifacts in intravascular ultrasound imaging during coronary artery stent implantation, Ultrasound in Med. Biol., Vol. 24(6), pp. 793-802, 1998.

[17] P. Wells, Developments in medical ultrasonics, World Med Electron, Vol. 66, pp. 272-277, 1966.

[18] N. Bom, C. Lancée, and F.V. Egmond, An ultrasonic intracardiac scanner, Ultrasonics, Vol. 10, pp. 72-76, 1972.

[19] G. Finet, E. Maurincomme, A. Tabib, R. J. Crowley, I. Magnin, R. Roriz, J. Beaune and M. Amiel, Artifacts in intravascular ultrasound imaging: analyses and implications, Ultrasound in Med. Biol., Vol. 19(7), pp. 533-547, 1993.

[20] Y. Yakimovsky, Boundary and Object Detection in Real World Images, Journal of the Association for Computing Machinery, Vol. 23(4), pp. 599-618, 1976.

[21] F. Guérault and Ph. Réfrégier. Optimal location of an intensity fluctuating target in nonhomogeneous and nonoverlapping background. in International Symposium on Optical Science, Engineering, and Instrumentation. 1998. San Diego, California USA: SPIE.

[22] F. Guérault, Techniques statistiques pour l'estimation de la position d'un objet dans des images bruitées, Thèse de Doctorat, Univ. Aix-Marseille, 1999, p. 179.

[23] L. D. Cohen, On active contour models and balloons, CVGIP : Image Understanding, Vol. 53, pp 211-218, 1991.

[24] B. Fua and C. Brechb_hler, Imposing hard constraints on deformable models through optimization in orthogonal subspaces, Computer Vision and Image Understanding, Vol. 65(2), pp 148-162, 1997.

[25] R. Malladi, J. Sethian, and B. Venuri, Shape modelling with front propagation, IEEE Trans. Pattern Anal. And Machine Intell., Vol. 17, pp 158-175, 1995.

[26] V. Caselles, J. Sethian, and G. Sapiro, Geodesic Active Contours, Int. J. Computer Vision, Vol. 22(1), pp 61-79, 1997.

[27] L. H. Staib and J. S. Ducan, Boudary finding with parametrically deformable contour models, IEEE Trans Pattern Anal. And Machine Intell., Vol. 14, pp 1061-1075, 1992.

[28] S. Menet, Ph. Saint-Marc and G. Medioni, B-sankes : implementation and application to stereo, Artificial Intelligence and Computer Vision, pp 223-236, 1991.

[29] Y. Amit, U. Grenander, and M. Piccioni, Structural image restauration through deformable templates, Journal of the Americain Statistical Association, Vol 86, pp 376-397, 1991.

[30] T. Cootes, A. Hill, C. J. Taylor, and J. Haslam, Use of active models for locating structure in medical images, Image and Vision Computing, Vol 12, pp 355-365, 1994.

[31] C. Kervrann and F. Heitz, A hierarchical statistical framework for the segmentation of deformable objects in image sequences, Proc. IEEE Conf. Comp. Vision Pattern Recognition, Seattle, pp 724-728, 1994.

[32] T. Cootes, C. J. Taylor, D. H. Cooper, and J. Graham, Active shape models – their training and application, CVGIP : Image Understanding, Vol. 61, pp 38-59, 1995.

[33] C. Chesnaud, Ph. Réfrégier, and V. Boulet, Statistical region snake-based segmentation adapted to different physical noise models, IEEE Trans. Pattern Anal. And Machine Intell., Vol. 21(11), 1999.