Attenuation Correction of L-shell X-ray Fluorescence Computed Tomography Imaging

X-ray fluorescence computed tomography (XFCT) is an experimental technique that can reconstruct the distribution of elements within the sample from the measurement of fluorescence stimulated from the sample[1]. The sample is irradiated with a x-ray beam, of which the energy is greater than the K-shell energy or L-shell energy of the interest elements. These x rays undergo photoelectric interaction and stimulate fluorescence of atoms. An energy-discriminating detector is placed at 90-deg. to detect the fluorescence undergoing attenuation for minimizing the effect by Compton Scattering[2]. Each element has its own fluorescence, and intensity of fluorescence can reflect content of element. The distribution of interest elements in sample can be reconstructed with measurement of fluorescence when the sample is scanned and rotated if attenuation can be neglected[3]. To reconstruct more accurate element distribution image, attenuation correction is necessary[4, 5]. Hogan presents a method about FBP with attenuation correction in the XFCT reconstruction, in which attenuation coefficient distribution at incident energy and fluorescence energy must be known[6]. Then Golosio comes up with a method that solves the attenuation problem combining x-ray fluorescence, Compton and transmission tomography[7]. La Riviere develops an alternating-update iterative reconstruction algorithm based on maximizing a penalized Poisson likelihood objective function[8]. In this work, the