Sodium MRI of Healthy Human Skin

• Author / Creator

  Zhu, Jingxuan

• Measuring skin sodium content has attracted significant interest due to its potential as a biomarker for various diseases, including dermatological conditions, cardiovascular disease, kidney disease, and diabetes. Skin sodium content can also vary between males and females and increases with age in healthy individuals. Non-invasive quantification of skin tissue sodium concentration (TSC) in vivo using sodium magnetic resonance imaging (MRI) could provide valuable insights for monitoring disease progression and understanding sodium regulation in the body. Various studies have quantified human skin TSC using MRI. A commonly adopted method, first proposed in 2012 by Kopp et al, uses a volume coil, GRE sequence with TE=2 ms, and 3×3×30 mm³ voxels and has been widely adopted in skin sodium MRI studies. This method has limitations: TE=2 ms causes significant signal loss due to the rapid biexponential signal decay of sodium in the skin, and the large voxel size leads to partial volume effects because the skin is only 1-2 mm thick, affecting TSC quantification accuracy. This thesis aims to address these limitations by reducing signal loss and improving the accuracy of skin TSC measurements in healthy human skin using a surface coil and twisted-projection imaging (TPI) with smaller, reshaped voxels.
  Sodium images of calf skin were acquired on a 4.7T MRI scanner from 14 healthy adults using TPI with a short echo time (TE) of approximately 0.1 ms. Initially, images were captured with a volume radiofrequency (RF) coil using voxels measuring 1.5×1.5×15 mm³ (23Na MRI protocol denoted here as VolPencil), following the widely adopted skin imaging protocol. Subsequently, a 5 cm diameter surface RF coil was employed, allowing

  for five times smaller voxels (0.8×0.8×10 mm³) while maintaining a similar signal-to-noise ratio (SNR) within the same 12-minute scan time (SurfPencil). These ‘pencil-shaped’ voxels were replaced with ‘pancake-shaped’ voxels (0.4×4×4 =6.4 mm³), better matching the anatomy of pressed flat skin (SurfPancake). The B1 field of the surface coil was investigated using a novel spin-3/2 simulation. The relationship between TSC measurement and skin thickness was also investigated.
  Results indicated that the higher resolution SurfPencil approach yielded a 44±16% greater skin sodium image intensity compared to VolPencil, while SurfPancake provided an additional 20±9% increase (p < 1e-8), indicating reduced signal loss. Across participants with skin thicknesses ranging from 1.0 to 1.8 mm, sodium intensity significantly increased by 56%±19% and 44%±12% as a function of greater skin thickness for VolPencil and SurfPencil, respectively (p < 0.003), but not for SurfPancake, indicating reduced bias when pancake-shaped voxels are used. Imaging with SurfPancake measured a skin sodium concentration of 34±5 mM, greater than the ~20 mM measured using the widely adopted protocol but still much lower than the 85 mM measured by atomic absorption spectroscopy. Simulation, which matched experimental trends, identified a remaining 64% signal loss from signal loss during RF pulse and point spread function (PSF) smearing. Relaxation-based compensation yielded a concentration of 95±15 mM, similar to the value obtained by atomic absorption spectroscopy.
  These findings suggest that previously published studies using volume RF coils with suboptimal spatial resolution may have underestimated skin sodium content due to partial

  volume effects from adjacent tissues like adipose, muscle and air. Consequently, observed sex differences or disparities between healthy and diseased states in earlier studies could be partially attributed to varying skin thickness rather than true physiological differences.
  By adopting TPI sequence with 0.1 ms TE, surface coil and 6.4 mm³ voxels with only 0.4 mm voxel dimensions across the skin, these errors have been mitigated, providing more accurate measurements of skin sodium concentration. Further, relaxation-based signal compensation demonstrated that it is feasible to compensate for signal loss with T2 relaxation parameters, suggesting a future direction of measuring individual relaxation values for more accurate skin TSC estimation. It is important to acknowledge the limitations of this approach, as measurements are confined to a small region of skin, and the entire cross-section of the leg is not imaged. Despite these limitations, this study offers a promising tool for overcoming challenges in sodium quantification and could lead to more precise assessments of sodium regulation and disease progression in the human body.

• Subjects / Keywords

  • MRI
  • skin
  • sodium MRI

• Graduation date

  Fall 2024

• Type of Item

  Thesis

• Degree

  Master of Science

• DOI

  https://doi.org/10.7939/r3-zptk-s373

• License

  This thesis is made available by the University of Alberta Library with permission of the copyright owner solely for non-commercial purposes. This thesis, or any portion thereof, may not otherwise be copied or reproduced without the written consent of the copyright owner, except to the extent permitted by Canadian copyright law.