Measuring tissue sodium concentration: Cross‐vendor repeatability and reproducibility of 23Na‐MRI across two sites

• Published in: Journal of magnetic resonance imaging Vol. 50; no. 4; pp. 1278 - 1284
• Main Authors: Riemer, Frank, McHugh, Damien, Zaccagna, Fulvio, Lewis, Daniel, McLean, Mary A., Graves, Martin J., Gilbert, Fiona J., Parker, Geoff J.M., Gallagher, Ferdia A.
• Format: Journal Article
• Language: English
• Published: Hoboken, USA John Wiley & Sons, Inc 01.10.2019; Wiley Subscription Services, Inc
• Subjects: biomarker; Biomarkers; brain; Calibration; Cerebrospinal fluid; Coefficient of variation; Cones; Females; Field of view; Field strength; Image processing; Image segmentation; Magnetic resonance imaging; Males; Medical imaging; Neuroimaging; Original Research; repeatability; Reproducibility; Scanners; Sodium; sodium MRI; Statistical analysis; Statistical tests; Substantia alba; Substantia grisea; Tumors
• ISSN: 1053-1807; 1522-2586
• DOI: 10.1002/jmri.26705

Background Sodium MRI (23Na‐MRI)‐derived biomarkers such as total sodium concentration (TSC) have the potential to provide information on tumor cellularity and the changes in tumor microstructure that occur following therapy. Purpose To evaluate the repeatability and reproducibility of TSC measurements in the brains of healthy volunteers, providing evidence for the technical validation of 23Na‐MRI‐derived biomarkers. Study Type Prospective multicenter study. Subjects Eleven volunteers (32 ± 6 years; eight males, three females) were scanned twice at each of two sites. Field Strength/Sequence Comparable 3D‐cones 23Na‐MRI ultrashort echo time acquisitions at 3T. Assessment TSC values, quantified from calibration phantoms placed in the field of view, were obtained from white matter (WM), gray matter (GM), and cerebrospinal fluid (CSF), based on automated segmentation of coregistered 1H T1‐weighted images and hand‐drawn regions of interest by two readers. Statistical Tests Coefficients of variation (CoVs) from mean TSC values were used to assess intrasite repeatability and intersite reproducibility. Results Mean GM TSC concentrations (52.1 ± 7.1 mM) were ∼20% higher than for WM (41.8 ± 6.7 mM). Measurements were highly repeatable at both sites with mean scan–rescan CoVs between volunteers and regions of 2% and 4%, respectively. Mean intersite reproducibility CoVs were 3%, 3%, and 6% for WM, GM, and CSF, respectively. Data Conclusion These results demonstrate technical validation of sodium MRI‐derived biomarkers in healthy volunteers. We also show that comparable 23Na imaging of the brain can be implemented across different sites and scanners with excellent repeatability and reproducibility. Level of Evidence: 1 Technical Efficacy: Stage 2 J. Magn. Reson. Imaging 2019;50:1278–1284.