Automatic Spinal Cord Gray Matter Quantification: A Novel Approach.

Area of Science:

• Neuroimaging
• Medical Image Analysis
• Spinal Cord Anatomy

Background:

• Accurate spinal cord gray matter (GM) segmentation is challenging.
• A noninvasive, broadly accepted reference standard for GM measurements is still under discussion.

Purpose of the Study:

• To assess the reproducibility and accuracy of cervical spinal cord GM and white matter (WM) cross-sectional area measurements.
• To evaluate an automated postprocessing segmentation algorithm using averaged magnetization inversion recovery (MIR) acquisitions.

Main Methods:

• 24 healthy subjects underwent 3T MRI with averaged MIR acquisitions.
• Manual segmentation by two readers was compared with a 3D variational segmentation algorithm.
• Reproducibility and accuracy were assessed using coefficients of variation and Dice similarity coefficients.

Main Results:

• The automated method achieved high reproducibility (coefficients of variation ≤4.77%) and excellent accuracy (Dice similarity coefficients: 0.86 for GM, 0.90 for WM).
• The automated approach showed consistent performance across sessions and slices.
• Mean GM area was 17.20 ± 2.28 mm², and mean WM area was 72.71 ± 7.55 mm².

Conclusions:

• A novel approach using averaged MIR acquisitions and automated segmentation provides accurate and reproducible spinal cord GM and WM segmentation.
• This pipeline shows promise for longitudinal studies of GM changes and clinical applications in spinal cord disorders.