A statistical framework for material decomposition using multi-energy photon counting x-ray detector

Recently, material decomposition using multi energy photon counting x-ray detector (PCXD) have been one of the very active areas of researches. Even with a few success, decomposition of three materials, including malignant tissue, has been still considered a difficult problem and systematic studies are still required. In this paper, we propose a statistical framework for decomposition of three material using PCXD. In particular, we derive an energy level optimization algorithm using the theory of minimum variance unbiased estimator (MVUE), and iterative update of material composition as well as system parameter estimation under a unified statistical estimation framework. Experiment results using physical breast phantom as well as real ex vivo breast data confirm that the proposed algorithm can accurately decompose three materials including malignant tissue.