Spectral fitting strategy to overcome the overlap between 2-hydroxyglutarate and lipid resonances at 2.25 ppm.

Autor:	Askari P; Advanced Imaging Research Center, University of Texas Southwestern Medical Center, Dallas, Texas, USA.; Joint Graduate Program in Biomedical Engineering at University of Texas Arlington and University of Texas Southwestern Medical Center, Texas, USA., Dimitrov IE; Advanced Imaging Research Center, University of Texas Southwestern Medical Center, Dallas, Texas, USA.; Philips Healthcare, Gainesville, Florida, USA., Ganji SK; Philips Healthcare, Andover, Massachusetts, USA.; Department of Radiology, University of Texas Southwestern Medical Center, Dallas, Texas, USA., Tiwari V; Advanced Imaging Research Center, University of Texas Southwestern Medical Center, Dallas, Texas, USA., Levy M; Department of Neurological Surgery, University of Texas Southwestern Medical Center, Dallas, Texas, USA., Patel TR; Department of Neurological Surgery, University of Texas Southwestern Medical Center, Dallas, Texas, USA.; Department of Neurology and Neurotherapeutics, University of Texas Southwestern Medical Center, Dallas, Texas, USA.; Department of Radiation Oncology, University of Texas Southwestern Medical Center, Dallas, Texas, USA., Pan E; Department of Neurology and Neurotherapeutics, University of Texas Southwestern Medical Center, Dallas, Texas, USA.; Harold C. Simmons Cancer Center, University of Texas Southwestern Medical Center, Dallas, Texas, USA., Mickey BE; Department of Neurological Surgery, University of Texas Southwestern Medical Center, Dallas, Texas, USA.; Harold C. Simmons Cancer Center, University of Texas Southwestern Medical Center, Dallas, Texas, USA.; Annette G. Strauss Center for Neuro-Oncology, University of Texas Southwestern Medical Center, Dallas, Texas, USA., Malloy CR; Advanced Imaging Research Center, University of Texas Southwestern Medical Center, Dallas, Texas, USA.; Department of Radiology, University of Texas Southwestern Medical Center, Dallas, Texas, USA.; Veterans Affairs North Texas Health Care System, Dallas, Texas, USA., Maher EA; Department of Neurology and Neurotherapeutics, University of Texas Southwestern Medical Center, Dallas, Texas, USA.; Harold C. Simmons Cancer Center, University of Texas Southwestern Medical Center, Dallas, Texas, USA.; Annette G. Strauss Center for Neuro-Oncology, University of Texas Southwestern Medical Center, Dallas, Texas, USA.; Department of Internal Medicine, University of Texas Southwestern Medical Center, Dallas, Texas, USA., Choi C; Advanced Imaging Research Center, University of Texas Southwestern Medical Center, Dallas, Texas, USA.; Department of Radiology, University of Texas Southwestern Medical Center, Dallas, Texas, USA.; Harold C. Simmons Cancer Center, University of Texas Southwestern Medical Center, Dallas, Texas, USA.
Jazyk:	angličtina
Zdroj:	Magnetic resonance in medicine [Magn Reson Med] 2021 Oct; Vol. 86 (4), pp. 1818-1828. Date of Electronic Publication: 2021 May 12.
DOI:	10.1002/mrm.28829
Abstrakt:	Purpose: 1 H MRS provides a noninvasive tool for identifying mutations in isocitrate dehydrogenase (IDH). Quantification of the prominent 2-hydroxyglutarate (2HG) resonance at 2.25 ppm is often confounded by the lipid resonance at the same frequency in tumors with elevated lipids. We propose a new spectral fitting approach to separate these overlapped signals, therefore, improving 2HG evaluation. Methods: TE 97 ms PRESS was acquired at 3T from 42 glioma patients. New lipid basis sets were created, in which the small lipid 2.25-ppm signal strength was preset with reference to the lipid signal at 0.9 ppm, incorporating published fat relaxation data. LCModel fitting using the new lipid bases (Fitting method 2) was conducted along with fitting using the LCModel built-in lipid basis set (Fitting method 1), in which the lipid 2.25-ppm signal is assessed with reference to the lipid 1.3-ppm signal. In-house basis spectra of low-molecular-weight metabolites were used in both fitting methods. Results: Fitting method 2 showed marked improvement in identifying IDH mutational status compared with Fitting method 1. 2HG estimates from Fitting method 2 were overall smaller than those from Fitting method 1, which was because of differential assignment of the signal at 2.25 ppm to lipids. In receiver operating characteristic analysis, Fitting method 2 provided a complete distinction between IDH mutation and wild-type whereas Fitting method 1 did not. Conclusion: The data suggest that 1 H MR spectral fitting using the new lipid basis set provides a robust fitting strategy that improves 2HG evaluation in brain tumors with elevated lipids. (© 2021 International Society for Magnetic Resonance in Medicine.)
Databáze:	MEDLINE
Externí odkaz:	Zobrazit plný text záznamu Plný text