A Parametric Level Set Approach to Simultaneous Object Identification and Background Reconstruction for Dual Energy Computed Tomography

摘要

Dual energy computerized tomography has gained great interest because of itsability to characterize the chemical composition of a material rather thansimply providing relative attenuation images as in conventional tomography. Thepurpose of this paper is to introduce a novel polychromatic dual energyprocessing algorithm with an emphasis on detection and characterization ofpiecewise constant objects embedded in an unknown, cluttered background.Physical properties of the objects, specifically the Compton scattering andphotoelectric absorption coefficients, are assumed to be known with some levelof uncertainty. Our approach is based on a level-set representation of thecharacteristic function of the object and encompasses a number ofregularization techniques for addressing both the prior information we haveconcerning the physical properties of the object as well as fundamental,physics-based limitations associated with our ability to jointly recover theCompton scattering and photoelectric absorption properties of the scene. In theabsence of an object with appropriate physical properties, our approach returnsa null characteristic function and thus can be viewed as simultaneously solvingthe detection and characterization problems. Unlike the vast majority ofmethods which define the level set function non-parametrically, i.e., as adense set of pixel values), we define our level set parametrically via radialbasis functions (RBF's) and employ a Gauss-Newton type algorithm for costminimization. Numerical results show that the algorithm successfully detectsobjects of interest, finds their shape and location, and gives a adequatereconstruction of the background.