A penalized weighted least-squares image reconstruction based on scatter correction methods for X-ray CT

Long Chen (1) , Thomas Rodet (1) , Nicolas Gac (1)

1 L2S - Laboratoire des signaux et systèmes

For metal artifact reduction (MAR) in X-ray conebeam computed tomography (CBCT), we propose a penalized weighted least-squares (PWLS) image reconstruction method [1] based on scatter correction (SC) method. Our main contribution is to incorporate the inaccuracy of the scatter correction data in PWLS method, and to accelerate the PWLS image reconstruction with a preconditionned conjugate gradient (PCG) method [2]. The reconstruction results of both the simulation and experimental data show that our proposed method effectively reduces the metal artifacts (MA). Compared to the state of art, normalized metal artifact reduction (NMAR) method [3], our method prevents the introduction of some new artifacts due to an inapporiate interpolation possible used in NMAR.

Mots clés: X-ray tomography; scatter correction; metal artifact reduction; penalized weighted least-squares

Domaines: Traitement du signal et de l'image [eess.SP]; Traitement du signal et de l'image [eess.SP]

Long Chen, Thomas Rodet, Nicolas Gac. A penalized weighted least-squares image reconstruction based on scatter correction methods for X-ray CT. 2013 IEEE Nuclear Science Symposium and Medical Imaging Conference, Oct 2013, Seoul, South Korea. pp.2. ⟨hal-00843508⟩