2022/08/10 by Fredrik Grönberg, Mats Persson, Grönberg, Fredrik +3
Engineering · Medicine · #Advanced X-ray and CT Imaging #FOS: Physical sciences #Medical Imaging Techniques and Applications #Medical Physics (physics.med-ph) #Radiomics and Machine Learning in Medical Imaging
paper · pdf · doi:10.48550/arxiv.2208.05362
openalex publication_date 2022/08/10 · openalex created_date 2022/08/12 · openalex updated_date 2026/07/28
Purpose: The set of linear attenuation coefficients that belong to materials in the human body is commonly assumed to be spanned by two basis functions in the range of clinical x-ray energies, even though there is evidence that the dimensionality of this set is greater than two. It has not yet been clear that the use of a third basis function could be beneficial in absence of contrast agents. Approach: In this work, the choice of the number of basis functions used in the basis decomposition method is studied for the task of producing an image where a third material is separated from a background of two other materials, in a case where none of the materials have a K-edge in the range of considered x-ray energies (20-140 keV). The case of separating iron from mixtures of liver and adipose tissue is studied with a simulated phantom which incorporates random and realistic tissue variability. Results: Inclusion of a third basis function improves the quantitative estimate of iron concentration by several orders of magnitude in terms of mean squared error in the resulting image. Conclusions: The inclusion of a third basis function in the basis decomposition is essential for the studied imaging task and could have potential application for quantitative estimation of iron concentration from material decomposed images.