Using a geometric formulation of annular-like shape priors for constraining variational level-sets

• Published in: Pattern recognition letters 2011-07, Vol.32 (9), p.1240-1249
• Main Authors: Alessandrini, M., Dietenbeck, T., Basset, O., Friboulet, D., Bernard, O.
• Format: Article
• Language: English
• Subjects: Annular shapes; Applied sciences; Biological and medical sciences; Computer Science; Computerized, statistical medical data processing and models in biomedicine; Exact sciences and technology; Image processing; Information, signal and communications theory; Least square fitting; Level-sets; Medical Imaging; Medical management aid. Diagnosis aid; Medical sciences; Segmentation; Signal processing; Telecommunications and information theory
• ISSN: 0167-8655; 1872-7344
• DOI: 10.1016/j.patrec.2011.03.018

► Segmentation of images exhibiting annular like shapes which may be approximated by two elliptical contours. ► Design a new level-set framework specifically dedicated to the detection of annular shapes. ► Thanks to a fast solution to the least-squares fitting problem of similar patterns, our model handles the segmentation task efficiently with a single level-set function. In this paper we address the segmentation of images exhibiting annular like shapes which may be approximated by two elliptical contours. Such patterns are indeed recurrent in many image processing applications. In this context, we develop a level-set framework specifically dedicated to the detection of annular shapes. Thanks to a fast solution to the least-squares fitting problem of similar patterns, our model handles the segmentation task efficiently with a single level-set function. The behavior of this approach is illustrated on images from various fields. An evaluation is then performed for the myocardium detection in MRI and ultrasound cardiac images.