Xfct inversion by generalized ridge functions

Abstract

X-Ray fluorescence computed tomography (xfct) aims at reconstructing fluorescence density from emission data given the measured x-ray attenuation. In this paper, inspired by the classical results from Logan & Shepp [3], we briefly discuss the existence of generalized ridge functions providing the minimal norm solution of the inverse problem. An algorithm to construct such functions is presented, based on results from Kazantsev [4]. Numerical results are also shown, with real and simulated data.