Robust breast quantitative susceptibility mapping in the presence of silicone

• Published in: Magnetic resonance in medicine Vol. 90; no. 3; pp. 1209 - 1218
• Main Authors: Böhm, Christof, Stelter, Jonathan K., Weiss, Kilian, Meineke, Jakob, Komenda, Alexander, Borde, Tabea, Makowski, Marcus R., Fallenberg, Eva M., Karampinos, Dimitrios C.
• Format: Journal Article
• Language: English
• Published: United States Wiley Subscription Services, Inc 01.09.2023
• Subjects: Algorithms; Brain; Brain Mapping - methods; Breast; Breast - diagnostic imaging; Chemical equilibrium; Computer Simulation; Diamagnetism; Dipoles; Humans; Image Processing, Computer-Assisted - methods; In vivo methods and tests; Magnetic resonance imaging; Magnetic Resonance Imaging - methods; Mapping; Numerical methods; Robustness (mathematics); Silicones; Susceptibility; Water; silicone; quantitative susceptibility mapping; susceptibility; field map; QSM; breast disease; cancer; chemical shift encoding
• ISSN: 0740-3194; 1522-2594; 1522-2594
• DOI: 10.1002/mrm.29694

To (a) develop a preconditioned water-fat-silicone total field inversion (wfsTFI) algorithm that directly estimates the susceptibility map from complex multi-echo data in the breast in the presence of silicone and to (b) evaluate the performance of wfsTFI for breast quantitative susceptibility mapping (QSM) in silico and in vivo in comparison with formerly proposed methods. Numerical simulations and in vivo multi-echo gradient echo breast measurements were performed to compare wfsTFI to a previously proposed field map-based linear total field inversion algorithm (lTFI) with and without the consideration of the chemical shift of silicone in the field map estimation step. Specifically, a simulation based on an in vivo scan and data from five patients were included in the analysis. In the simulation, wfsTFI is able to significantly decrease the normalized root mean square error from lTFI without (4.46) and with (1.77) the consideration of the chemical shift of silicone to 0.68. Both the in silico and in vivo wfsTFI susceptibility maps show reduced shadowing artifacts in local tissue adjacent to silicone, reduced streaking artifacts and no erroneous single voxels of diamagnetic susceptibility in proximity to silicone. The proposed wfsTFI method can automatically distinguish between subjects with and without silicone. Furthermore wfsTFI accounts for the presence of silicone in the QSM dipole inversion and allows for the robust estimation of susceptibility in proximity to silicone breast implants and hence allows the visualization of structures that would otherwise be dominated by artifacts on susceptibility maps.