Active Contour-Based Segmentation of Head and Neck with Adaptive Atlas Selection

This paper presents automated segmentation of structures in the Head and Neck (H&N) region, using an active contour-based joint registration and segmentation model. A new atlas selection strategy is also used. Segmentation is performed based on the dense deformation field computed from the registration of selected structures in the atlas image that have distinct boundaries, onto the patient's image. This approach results in robust segmentation of the structures of interest, even in the presence of tumors, or anatomical differences between the atlas and the patient image. For each patient, an atlas image is selected from the available atlas-database, based on the similarity metric value, computed after performing an affine registration between each image in the atlas-database and the patient's image. Unlike many of the previous approaches in the literature, the similarity metric is not computed over the entire image region; rather, it is computed only in the regions of soft tissue structures to be segmented. Qualitative and quantitative evaluation of the results is presented.

Dense Deformation Field Estimation for Atlas Registration using the Active Contour Framework

Meritxell Bach Cuadra, Xavier Bresson, Valérie Duay, Jean-Philippe Thiran

In this paper, we propose a new paradigm to carry out the registration task with a dense deformation field derived from the optical flow model and the active contour method. The proposed framework merges different tasks such as segmentation, regularization, incorporation of prior knowledge and registration into a single framework. The active contour model is at the core of our framework even if it is used in a different way than the standard approaches. Indeed, active contours are a well-known technique for image segmentation. This technique consists in finding the curve which minimizes an energy functional designed to be minimal when the curve has reached the object contours. That way, we get accurate and smooth segmentation results. So far, the active contour model has been used to segment objects lying in images from boundary-based, region-based or shape-based information. Our registration technique will profit of all these families of active contours to determine a dense deformation field defined on the whole image. A well-suited application of our model is the atlas registration in medical imaging which consists in automatically delineating anatomical structures. We present results on 2D synthetic images to show the performances of our non rigid deformation field based on a natural registration term. We also present registration results on real 3D medical data with a large space occupying tumor substantially deforming surrounding structures, which constitutes a high challenging problem.

INTEL2006

Segmentation of head and neck lymph node regions for radiotherapy planning, using active contour based atlas registration

Meritxell Bach Cuadra, Valérie Duay, Nawal Houhou, Jean-Philippe Thiran, Subrahmanyam Venkata Ravi Mohana Sai Gorthi

In this paper, we present the segmentation of the head and neck lymph node regions using a new active contour based atlas registration model. We propose to segment the lymph node regions without directly including them in the atlas registration process; instead, they are segmented using the dense deformation field computed from the registration of the atlas structures with distinct boundaries. This approach results in robust and accurate segmentation of the lymph node regions even in the presence of significant anatomical variations between the atlas-image and the patient's image to be segmented. We also present a quantitative evaluation of lymph node regions segmentation using various statistical as well as geometrical metrics: sensitivity, specificity, dice similarity coefficient and Hausdorff distance. A comparison of the proposed method with two other state of the art methods is presented. The robustness of the proposed method to the atlas selection, in segmenting the lymph node regions, is also evaluated.

2009

Geodesic Active Fields - A Geometric Framework for Image Registration

Xavier Bresson, Jean-Philippe Thiran, Dominique Zosso

In this paper we present a novel geometric framework called geodesic active fields for general image registration. In image registration, one looks for the underlying deformation field that best maps one image onto another. This is a classic ill-posed inverse problem, which is usually solved by adding a regularization term. Here, we propose a multiplicative coupling between the registration term and the regularization term, which turns out to be equivalent to embed the deformation field in a weighted minimal surface problem. Then, the deformation field is driven by a minimization flow toward a harmonic map corresponding to the solution of the registration problem. This proposed approach for registration shares close similarities with the well-known geodesic active contours model in image segmentation, where the segmentation term (the edge detector function) is coupled with the regularization term (the length functional) via multiplication as well. As a matter of fact, our proposed geometric model is actually the exact mathematical generalization to vector fields of the weighted length problem for curves and surfaces introduced by Caselles-Kimmel-Sapiro. The energy of the deformation field is measured with the Polyakov energy weighted by a suitable image distance, borrowed from standard registration models. We investigate three different weighting functions, the squared error and the approximated absolute error for monomodal images, and the local joint entropy for multimodal images. As compared to specialized state-of-the-art methods tailored for specific applications, our geometric framework involves important contributions. Firstly, our general formulation for registration works on any parametrizable, smooth and differentiable surface, including non-flat and multiscale images. In the latter case, multiscale images are registered at all scales simultaneously, and the relations between space and scale are intrinsically being accounted for. Secondly, this method is, to the best of our knowledge, the first re-parametrization invariant registration method introduced in the literature. Thirdly, the multiplicative coupling between the registration term, i.e. local image discrepancy, and the regularization term naturally results in a data-dependent tuning of the regularization strength. Finally, by choosing the metric on the deformation field one can freely interpolate between classic Gaussian and more interesting anisotropic, TV-like regularization.

Institute of Electrical and Electronics Engineers2011

Geodesic Active Fields - A Geometric Framework for Image Registration

Xavier Bresson, Jean-Philippe Thiran, Dominique Zosso

Institute of Electrical and Electronics Engineers2011