Sodium magnetic resonance imaging (23Na-MRI) provides access to tissue sodium content and related biophysical parameters that are not visible on conventional proton MRI, but its clinical use remains limited because of low intrinsic sensitivity, very short T2* components, and constraints on spatial resolution and scan time. In this review, we first summarize the nuclear magnetic properties of 23Na and its position among other X-nuclei in terms of NMR receptivity and in vivo signal strength, and then outline representative acquisition strategies such as ultrashort echo time (UTE) and zero echo time (ZTE) imaging, together with non-Cartesian reconstruction and practical workflows for quantitative imaging. Particular emphasis is placed on B0/B1 inhomogeneity correction, phantom-based calibration for tissue sodium concentration (TSC), and compartmental modeling to derive indices such as cellular volume fraction and fixed charge density. We then briefly review preclinical and clinical studies in the brain, articular cartilage, musculoskeletal system, and kidneys, highlighting situations in which 23Na-MRI has provided complementary information on cell viability, cartilage matrix integrity, ionic homeostasis, and tumor composition. Finally, we discuss future directions, including physics-guided deep learning for reconstruction and denoising of low-SNR 23Na data, and integration with quantitative proton MRI and radiomics-based analysis. These developments are expected to improve feasibility and reproducibility and to clarify the potential clinical value of 23Na-MRI in neurological, musculoskeletal, and renal diseases.
Yasuhiko TERADA (2026) studied this question.