Inferring Preoperative Reconstructed Spine Models to Volumetric CT Data through High-Order MRFs
نویسندگان
چکیده
In this paper, we introduce a novel approach based on higher order energy functions which have the ability to encode global structural dependencies to infer articulated 3D spine models to CT volume data. A personalized geometrical model is reconstructed from biplanar X-rays before spinal surgery in order to create a spinal column representation which is modeled by a series of intervertebral transformations based on rotation and translation parameters. The shape transformation between the standing and lying poses is then achieved through a Markov Random Field optimization graph, where the unknown variables are the deformations applied to the intervertebral transformations. Singleton and pairwise potentials measure the support from the data and geometrical dependencies between neighboring vertebrae respectively, while higher order cliques are introduced to integrate consistency in regional curves. Optimization of model parameters in a multi-modal context is achieved using efficient linear programming and duality. A qualitative evaluation of the vertebra model alignment obtained from the proposed method gave promising results while the quantitative comparison to expert identification yields an accuracy of 1.8±0.7mm based on the localization of surgical landmarks. Key-words: Registration, physical modeling, image segmentation, articulated 3D spine model, Higher-order MRFs ∗ Laboratoire MAS, Ecole Centrale de Paris, Grande Voie des Vignes, 92295 ChatenayMalabry, France in ria -0 04 42 04 8, v er si on 1 18 D ec 2 00 9 Inférence de Modèles 3D Préopératoire de la Colonne Vertébrale aux Données Volumétriques par des MRFs de Haut-Niveau Résumé : Ce papier présente une méthode d’inférence d’un modèle personnalisé de la colonne vertébrale en 3D à partir de données tomodensitométriques en exploitant des fonctions d’énergie de haut niveau incorporant des dépendances géométriques. Une reconstruction 3D précise à partir d’images radiographiques standards est exploitée afin d’obtenir une représentation du rachis modélisée par une série de transformations intervertébrale. Ces transformations sont basées sur les paramtres de rotation et de translation. La transformation de la forme du rachis entre les positions couchées et debout est atteinte grâce a une optimisation un Markov Random Field (MRF), où les variables inconnues sont les déformations appliquées aux transformations intervertébrales. Des valeurs potentiels unitaires et binômes mesurent le lien entre les images et les contraintes géométriques entre les vertèbres, alors que des fonctions de haut niveau introduisent des contraintes d’alignement des régions anatomiques. L’optimisation des paramètres dans un contexte multi-modale est effectuée par une approche de programmation linéaire et par dualité. Nous présentons des résultats prometteurs pour le recalage d’images à partir d’une comparaison avec une identification manuelle d’un expert qui offre une précision de 1.8 ± 0.7mm basée sur la localisation de repères chirurgicaux. Mots-clés : Recalage d’images, modélisation physique, segmentation, modèle 3D articulé de la colonne vertébrale, MRF haut-niveau in ria -0 04 42 04 8, v er si on 1 18 D ec 2 00 9 High-Order MRF 3D Spine Registration 3
منابع مشابه
Nonlinear Embedding towards Articulated Spine Shape Inference Using Higher-Order MRFs
In this paper we introduce a novel approach for inferring articulated spine models from images. A low-dimensional manifold embedding is created from a training set of prior mesh models to establish the patterns of global shape variations. Local appearance is captured from neighborhoods in the manifold once the overall representation converges. Inference with respect to the manifold and shape pa...
متن کاملAn Estimation of Multiphase Relative Permeabilities in Reservoir Cores from Micro-CT Data
With significant increase of tomographic equipment power, demand for Prediction relative permeability prediction Predicting in porous media from digital image data. In this work, it is predicted three -phase relative permeabilities with co-applying Darcy’s and Stokes equations in two case studies, namely Bentheimer sandstone and Estaillades limestone which their micro-CT data files were downloa...
متن کاملMarkov Random Field Based Structure Localisation of Vertebrae for 3D-Segmentation of the Spine in CT Volume Data
Medical Image Processing is a growing field in medicine and plays an important role in medical decision making. Computer-based segmentation of anatomies in data made by imaging modalities supports clinicians and speeds up their diagnosis making compared to doing it manually. Computed Tomography (CT) is an imaging modality for slice-wise three dimensional reconstruction of the human body in the ...
متن کاملPerformance of a 2D-3D Image Registration System using (Lossy) Compressed X-ray CT
In 2D-3D Medical Image Registration (MIR), preoperative 3D CT images are registered with intraoperative 2D X-rays images obtained by fluoroscopy or Electronic Portal Imaging Devices (EPID). 2D-3D MIR is established by computing Digitally Reconstructed Radiograph (DRR) images rendered from the volumetric data and is useful in many medical procedures where surgical planning is undertaken using vo...
متن کاملIntegrative Construction and Analysis of Condition-specific Biological Networks
Condition-specific biological networks are networks of genes and proteins induced in cells in response to different environmental conditions. The dependency structure of these networks provides important insight into how cells respond to healthy and stressful conditions. Unfortunately, for most conditions the network structure is unknown and must be inferred from condition-specific measurements...
متن کاملذخیره در منابع من
با ذخیره ی این منبع در منابع من، دسترسی به آن را برای استفاده های بعدی آسان تر کنید
عنوان ژورنال:
دوره شماره
صفحات -
تاریخ انتشار 2009