Optimization and application of zero echo time sequence in MRI of sacroiliac joints: a preliminary study

This study aimed to optimize and apply ZTE sequence parameters for magnetic resonance imaging of sacroiliac joints.

A healthy volunteer underwent ZTE imaging with different slice thicknesses (0.8, 1, 1.2, 1.4 mm) and flip angle (FAs) (1°, 2°, 3°). Two observers independently measured the bone marrow signal-to-noise ratio (SNR) to identify the optimal sequences. The top two sequences were applied to 43 axial spondyloarthritis patients (Oct-Dec 2024). Observers scored image quality (1-5 scale: very poor to excellent), SNR, and contrast-to-noise ratio (CNR) independently. Inter-observer agreement was assessed by weighted Kappa and Intraclass Correlation Coefficient (ICC). Statistical significance was tested using paired t-tests or Wilcoxon tests.

A 1.4 mm slice thickness maximized SNR for all FAs. The 2°/1.4 mm and 3°/1.4 mm sequences performed best. Inter-observer agreement for image quality scores showed excellent weighted Kappa (0.872 vs. 0.862). ICCs for SNR were 0.960 vs. 0.970, CNR ICCs were 0.782 vs. 0.897. The 2°/1.4 mm sequence achieved superior image quality (5 vs. 4; P < 0.05), lower SNR (97.09 ± 12.9 vs. 107.77 ± 15.34; P < 0.05), and better CNR (2.85 ± 4.29 vs. 1.89 ± 4.27; P < 0.05).

A 1.4 mm slice thickness maximized SNR across FAs. The 2°/1.4 mm sequence provided superior image quality and CNR for sacroiliac joint evaluation.