A chemical shift encoding (CSE) approach for spectral selection in fluorine‐19 MRI

• Published in: Magnetic resonance in medicine 2018-04, Vol.79 (4), p.2183-2189
• Main Authors: Ludwig, Kai D., Hernando, Diego, Roberts, Nathan T., van Heeswijk, Ruud B., Fain, Sean B.
• Format: Article
• Language: English
• Subjects: Animals; chemical shift encoding; Chemical speciation; Coding; Computer Simulation; Contrast Media - chemistry; Crown Ethers - chemistry; Feasibility studies; Fluorination; Fluorine; Fluorine-19 Magnetic Resonance Imaging; fluorine‐19; Humans; Image acquisition; Image processing; Image Processing, Computer-Assisted - methods; Image reconstruction; Isoflurane; Isoflurane - chemistry; Linear Models; Lower bounds; Magnetic Resonance Imaging; Male; Mice; Mice, Inbred C57BL; Models, Statistical; Noise; Perfluorocarbons; Phantoms, Imaging; Separation; Signal-To-Noise Ratio; Spectra
• ISSN: 0740-3194; 1522-2594
• DOI: 10.1002/mrm.26874

Purpose To develop a chemical shift encoding (CSE) approach for fluorine‐19 MRI of perfluorocarbons in the presence of multiple known fluorinated chemical species. Theory and Methods A multi‐echo CSE technique is applied for spectral separation of the perfluorocarbon perfluoro‐15‐crown‐5‐ether (PFCE) and isoflurane (ISO) based on their chemical shifts at 4.7 T. Cramér‐Rao lower bound analysis is used to identify echo combinations with optimal signal‐to‐noise performance. Signal contributions are fit with a multispectral fluorine signal model using a non‐linear least squares estimation reconstruction directly from k‐space data. This CSE approach is tested in fluorine‐19 phantoms and in a mouse with a 2D and 3D spoiled gradient‐echo acquisition using multiple echo times determined from Cramér‐Rao lower bound analysis. Results Cramér‐Rao lower bound analysis for PFCE and ISO separation shows signal‐to‐noise performance is maximized with a 0.33 ms echo separation. A linear behavior (R2 = 0.987) between PFCE signal and known relative PFCE volume is observed in CSE reconstructed images using a mixed PFCE/ISO phantom. Effective spatial and spectral separation of PFCE and ISO is shown in phantoms and in vivo. Conclusion Feasibility of a gradient‐echo CSE acquisition and image reconstruction approach with optimized noise performance is demonstrated through fluorine‐19 MRI of PFCE with effective removal of ISO signal contributions. Magn Reson Med 79:2183–2189, 2018. © 2017 International Society for Magnetic Resonance in Medicine.