Robustness of Texture Features in Quantitative Susceptibility Mapping: Influence of Background Field Removal and Inversion Algorithms

Motivation: The smoothness of quantitative susceptibility mapping (QSM) images varies with reconstruction method used, which may impact the reproducibility of texture features. Goal(s): To assess the robustness of texture features in QSM images across different background field removal and inversion algorithms used in QSM reconstruction Approach: Three background field removal algorithms and two inversion algorithms were applied for QSM reconstruction. The inter-algorithm reproducibility of texture features was then assessed in a cohort of 40 subjects. Results: The reproducibility of texture features varied across different combinations of background field removal and inversion algorithms in QSM reconstruction. Impact: The robustness of texture features varies across different background field removal and inversion algorithms in QSM reconstruction. The impact of the QSM reconstruction algorithm on texture features should be carefully considered before clinical implementation.