Segmentation compétitive de l’hippocampe et de l’amygdale à partir de volumes IRM

Segmentation compétitive de l’hippocampe et de l’amygdale à partir de volumes IRM

Competitive segmentation of the hippocampus and the amygdala from MRI scans

M. Chupin D. Hasboun  É. Bardinet  S. Baillet  L. Lemieux  L. Garnero 

Department of Clinical and Experimental Epilepsy, Institute of Neurology, University College of London, Queen Square, WC1N 3BG, London, UK

Laboratoire de Neurosciences Cognitives et Imagerie Cérébrale, CNRS UPR640, Hôpital de la Salpêtrière, 47 bvd de l’hôpital, 75651 Paris Cedex 13

Unité de Neuroradiologie, Hôpital de la Salpêtrière, 47 bvd de l’hôpital, 75651 PARIS Cedex 13

Corresponding Author Email: 
marie.chupin@chups.jussieu.fr
Page: 
503-516
|
Received: 
14 October 2005
|
Accepted: 
N/A
|
Published: 
31 December 2006
| Citation

OPEN ACCESS

Abstract: 

The hippocampus and the amygdala are two brain structures which play a central role in several fundamental cognitive processes. Their segmentation from Magnetic Resonance Imaging (MRI) scans is a unique way to measure their atrophy in some neurological diseases, but it is made difficult by their complex geometry. Their simultaneous segmentation is considered here through a competitive homotopic region growing method. It is driven by relational anatomical knowledge, which enables to consider the segmentation of atrophic structures in a straightforward way. For both structures, this fast algorithm gives results which are comparable to manual segmentation with a better reproducibility. Its performances regarding segmentation quality, automation and computation time, are amongst the best published data.

Résumé

L’hippocampe et l’amygdale sont deux structures cérébrales intervenant dans plusieurs fonctions cognitives fondamentales. Leur segmentation, à partir de volumes d’imagerie par résonance magnétique (IRM), est un outil essentiel pour mesurer leur atteinte dans certaines pathologies neurologiques, mais elle est rendue difficile par leur géométrie complexe. Nous considérons leur segmentation simultanée par une méthode de déformation homotopique compétitive de régions. Celle-ci est guidée par des connaissances anatomiques relationnelles; ceci permet de considérer directement des structures atrophiées. Rapide, l’algorithme donne, pour les deux structures, des résultats comparables à la segmentation manuelle avec une meilleure reproductibilité. Ses performances, concernant la qualité de la segmentation, le degré d’automatisation et le temps de calcul, sont parmi les meilleures de la littérature.

Keywords: 

Medical imaging, segmentation, MRI, Brain, Anatomical priors

Mots clés

Imagerie médicale, Segmentation, IRM, Cerveau, a priori anatomiques

1. Introduction
2. Méthode
3. Résultats
4. Discussion
5. Conclusion
  References

[1] D. HASBOUN, M. CHANTÔME, A. ZOUAOUI, M. SAHEL, M. DELA-DOEUILLE, N. SOUROUR, M.DUYMES, M. BAULAC, C. MARSAULT et D. DORMONT, « MR determination of hippocampal volume: comparison of three methods», American Journal of NeuroRadiology, vol. 17, p. 1091-1098, 1996.

[2] J. PRUESSNER, L. LI, W. SERLES, M. PRUESSNER, D. COLLINS, N. KA-BANI, S. LUPIEN et A. EVANS, «Volumetry of hippocampus and amygdala with high-resolution MRI and tree-dimensional analysis software: Minimizing the discrepancies between laboratories », Cerebral Cortex, vol. 10, p. 433-442, 2000.

[3] E. ASHTON, K. PARKER, M. BERG et C. CHEN, « A novel volumetric feature extraction technique with applications to MR images », IEEE Transactions on Medical Imaging, vol. 16 #4, p. 365-371, 1997.

[4] G. BUENO, O. MUSSE, F. HEITZ et J. ARMSPACH, «Threedimensional segmentation of anatomical structures in MR images on large data bases », Magnetic Resonance Imaging, vol. 19, p. 73-88, 2001.

[5] R. PÉREZ de ALEJO, J. RUIZ-CABELLO, M. CORTIJO, I. RODRIGUEZ, I. ECHAVE, J. REGADERA, J. ARRAZOLA, P. AVILÉS, P. BARREIRO, D. GARGALLO et M. GRANA, « Computer-assisted enhanced volumetric segmentation Magnetic Resonance Imaging data using a mixture of artificial neural networks », Magnetic Resonance Imaging, vol. 21, p. 901-912, 2003.

[6] S. HIRANO, N. KAMIURA, N. MATSUI, Y. HATA, R. SIROMONEY et K. SUBRAMANIAN, «Hyppocampus extraction based on parallel multiscale structure matching. parallel image processing and analysis », International Journal of Pattern Recognition and Artificial Intelligence, vol. 14 #4, p. 427-439, 2000.

[7] J. KLEMENCIC, V. VALENCIC et N. PECARIC, « Deformable contour based algorithm for segmentation of the hippocampus from MRI », CAIP 2001, vol. LNCS 2124, p. 298-308, 2001.

[8] A. GHANEI, H. SOLTANIAN-ZADEH et J. WINDHAMS, « A 3D deformable surface model for segmentation of objects from volumetric data in medical images », Computers in Biology and Medicine, vol. 28, p. 239-253, 1998.

[9] A. GHANEI, H. SOLTANIAN-ZADEH, K. ELISEVICH et J. FESSLER, « A knowledge-based deformable surface model with application to segmentation of brain strucures in MRI », SPIE Medical Imaging 2001, vol. 4322, p. 356-365, 2001.

[10] S. PIZER, S. JOSHI, P. FLETCHER, M. STYNER, G. TRACTON et J. CHEN, « Segmentation of single-figure object by deformable Mreps », MICCAI 2001, vol. 2208, p. 862-871, 2001.

[11] D. SHEN, S. MOFFAT, S. RESNICK et C. DAVATZIKOS, «Measuring size and shape of the hippocampus in MR images using a deformable shape model », NeuroImage, vol. 15 #2, p. 422-434, 2002.

[12] A. KELEMEN, G. SZEKELY et G. GERIG, « Elastic model-based segmentation of 3-D neuroradiological data sets », IEEE Transactions on Medical Imaging, vol. 18#10, p. 828-839, 1999.

[13] J. YANG, L. STAIB et J. DUNCAN, « Neighbor-constrained segmentation with 3D deformable models », IEEE Transactions on Medical Imaging, vol. 23 #8, p. 940-948, 2004.

[14] J. YANG et J. DUNCAN, «3D image segmentation of deformable objects with joint shape-intensity prior models using level sets », Medical Image Analysis, vol. 8, p. 285-294, 2004.

[15] A. PITIOT, H. DELINGETTE, P. THOMPSON et N. AYACHE, «Expert knowledge-guided segmentation system for brain MRI», NeuroImage, vol. 23, p. S85-S96, 2004.

[16] B. FISCHL, D. SALAT, E. BUSA, M. ALBERT, M. DIETERICH, C. HASEL-GROVE, A. V. D. KOUWE, R. KILLIANY, D. KENNEDY, S. KLAVENESS, A. MONTILLO, N. MAKRIS, B. ROSEN et A. DALE, « Whole brain segmentation : Automated labelling of neuroanatomical structures in the human brain », Neuron, vol. 33, p. 341-355, 2002.

[17] J. ZHOU et J. RAJAPAKSE, «Segmentation of subcortical brain structures using fuzzy templates », NeuroImage, vol. 28 #4, p. 915-924, 2005.

[18] B. DAWANT, S. HARTMANN, J.-P. THIRION, F. MAES, D. VANDERMEULEN et P. DEMAEREL, « Automatic 3-D segmentation of internal structures of the head in MR images using a combination of similarity and free-form transformations: Part I, methodology and validation on normal subjects », IEEE Transactions on Medical Imaging, vol. 18 #10, p. 909-916, 1999.

[19] S. DUCHESNE, J. PRUESSNER et D. COLLINS, «Appearancebased segmentation of medial temporal lobe structures », NeuroImage, vol. 17 #2, p. 515-531, 2002.

[20] B. VEMURI, J. YE et C. LEONARD, «Image registration via levelset motion: Applications to atlas based segmentation», Medical Image Analysis, vol. 7, p. 1-20, 2003.

[21] J. HALLER, A. BANERJEE, G. CHRISTENSEN, M. GADO, S. JOSHI, M. MILLER, Y. SHELINE, M. VANNIER et J. CSERNANSKY, « Three-dimensional hippocampal MR morphometry with high-dimensional transformation of a neuroanatomic atlas », Radiology, vol. 202, p. 504-510, 1997.

[22] R. HOGAN, K. MARK, L. WANG, S. JOSHI, M. MILLER et R. BUCHOLZ, « Mesial temporal sclerosis and temporal lobe epilepsy: MR imaging deformation based segmentation of the hippocampus in five patients », Radiology, vol. 216, p. 291-297, 2000.

[23] O. CARMICHAEL, H. AIZENSTEIN, S. DAVIS, J. BECKER, P. THOMPSON, C. MELTZER et Y. LIU, « Atlas-based hippocampus segmentation in Alzheimer's Disease and Mild Cognitive Impairment», NeuroImage, vol. 27 #4, p. 979-990, 2005.

[24] I. BLOCH, O. COLLIOT, O. CAMARA et T. GÉRAUD, «Fusion of spatial relationships for guiding recognition, example of brain structure recognition in 3D MRI », Pattern Recognition Letters, vol. 26, p. 449-457, 2005.

[25] V. BARRA et J.-Y. BOIRE, «Automatic segmentation of subcortical brain structures in MR images using information fusion », IEEE Transactions on Medical Imaging, vol. 20 #7, p. 549-568, 2001.

[26] J.-F. MANGIN, F. TUPIN, V. FROUIN, I. BLOCH, R. ROUGETET, J. RÉGIS et J. LÓPEZ-KRAHE, « Deformable topological models for segmentation of 3D medical image », IPMI 95, Brest, France, p. 153-164, 1995.

[27] F. POUPON, J. MANGIN, D. HASBOUN, C. POUPON, I. MAGNIN et V. FROUIN, « Multi-object deformable templates dedicated to segmentation of brain deep structures », MICCAI 1998, vol. 1496, p. 1134-1143, 1998.

[28] J. BESAG, « Towards bayesian image analysis », Journal of Applied Statistics, vol. 16 #3, p. 395-407, 1989.

[29] G. MALANDAIN, G. BERTRAND et N. AYACHE, «Topological segmentation of discrete surfaces », International Journal of Computer Vision, vol. 10 #2, p. 183-197, 1993.

[30] S. GEMAN et D. GEMAN, «Stochastic relaxation, Gibbs distributions, and the Bayesian restoration of images », IEEE Transactions on Pattern Analysis and Machine Intelligence, vol. 6 #6, p. 721-741, 1984.

[31] D. RIVIÈRE, D. PAPADOPOULOS, C. POUPON, F. POUPON, O. COULON, J.-B. POLINE, V. FROUIN, J. RÉGIS et J.-F. MANGIN, « A structural browser for human brain mapping », HBM 2000, San Antonio, Texas USA, vol. 11 #5, p. 912, 2000.

[32] G. GERIG, M. JOMIER et M. CHAKOS, « Valmet: a new validation toll for assessing and improving 3D image segmentation», MICCAI 2001, vol. 2208, p. 516-523, 2001.

[33] M. CHUPIN, D. HASBOUN, A.R. MUKUNA BANTUMBAKULU, E. BADINET, S. BAILLET, S. KINKINGNÉHUN, L. LEMIEUX, B. DUBOIS et L. GARNERO, «Competitive Segmentation of the Hippocampus and the Amygdala from MRI Data: Validation on Young Healthy Controls and Alzheimer's Disease Patients », SPIE Medical Imaging-2006, vol. 6144, 2006.