Work

Short echo time first-echo images of T2* mapping for detecting structural sacroiliac joint lesions in axial spondyloarthritis: a CT-referenced study

Year: 2026

Type: article

Abstract: Objective: To assess the CT-referenced diagnostic performance of short-echo time (TE) first-echo images derived from a T2* mapping acquisition for detecting sacroiliac joint (SIJ) structural lesions in patients evaluated for axial spondyloarthritis (axSpA), compared with conventional T1-weighted turbo spin-echo (T1-TSE) MRI. Methods: In this study, 54 patients underwent SIJ CT and 3T MRI; 52 were included in the final analysis. MRI included T1-TSE and a T2* mapping acquisition; the first-echo magnitude images (TE = 4.36 ms) were used for structural assessment. Two blinded readers scored CT and MRI using the Berlin and SPARCC structural scoring systems. Patient-level sensitivity (SE), specificity (SP), and diagnostic accuracy (DA) were calculated using CT as the reference. Correlation with CT was assessed using Pearson's r, and inter-rater agreement was evaluated (κ for Berlin binary positivity; ICC for continuous sum scores). Results: Forty-three of 52 patients fulfilled ASAS 2009/2011 spondyloarthritis classification criteria. Compared with T1-TSE, short TE-T2* images showed higher diagnostic performance for erosions (SE 88.6% vs 51.4%; SP 70.6% vs 82.4%; DA 82.6% vs 61.7%) and sclerosis (SE 86.2% vs 34.5%; SP 95.7% vs 100%; DA 90.3% vs 62.4%), while both sequences achieved high accuracy for joint space changes (DA 98.2%; T2*: SE/SP 98%/100%; T1: 100%/66.7%). Agreement with CT was stronger for short TE-T2* than for T1-TSE for Berlin erosion and sclerosis sum scores (r=0.913 and 0.921 vs 0.646 and 0.758), with substantial reader agreement for erosion and sclerosis positivity (κ=0.805 and 0.731 vs 0.564 and 0.598). Conclusion: Short TE first echo images from a T2* mapping acquisition provide an accessible, radiation-free, erosion-sensitive approach that demonstrates higher CT-referenced detection of erosions and sclerosis compared with conventional T1-TSE imaging, with strong CT agreement and robust inter-rater reliability.

Source: Frontiers in Immunology

DOI: 10.3389/fimmu.2026.1946946