Musculoskeletal MRI Relaxometry Phantom for T1 , T2 , and T1ρ Measurements

Abstract

ABSTRACT Background Quantitative MRI (qMRI) requires quality assurance methods that reflect in vivo conditions to ensure consistent measurements with low bias and high reproducibility over time, across multiple sites, and across different MRI scanner manufacturers and models. Purpose To develop a musculoskeletal (MSK) relaxometry phantom that spans the T1 and T2 (and T1ρ as an exploratory aim) ranges of articular cartilage. Study Type Longitudinal phantom study. Phantom A cylindrical phantom designed to fit within multiple knee coil configurations using dedicated support stands. The phantom contains 12 gel‐filled vials spanning T1 values of 300–1200 ms, T2 values of 10–100 ms, and T1ρ values of 20–120 ms, and an MR‐visible thermometer. Field Strength/Sequence 3T; single‐echo spin‐echo T2; inversion‐recovery T1; spin‐lock prepared T1ρ for NMR (spin‐lock frequencies 300–5000 Hz) and MAPSS T1ρ for MRI (spin‐lock frequency 500 Hz). Assessment NMR and two MRI systems from different vendors were used. A prototype phantom ( n = 1), production phantoms ( n = 12), and matched NMR samples were evaluated. NMR data acquired at multiple temperatures served as reference and enabled temperature correction of MRI measurements. The prototype phantom was scanned longitudinally (15 measurements over 725 days) to assess stability, while production phantoms were scanned at a single timepoint to evaluate batch consistency. Statistical Analysis Measurement variability and agreement were assessed using coefficients of variation (CVs), concordance correlation coefficients (CCCs), and Bland–Altman analysis for NMR–MRI comparison, longitudinal stability, batch consistency, and inter‐vendor reproducibility. Results After temperature correction, longitudinal CVs were 1.1% (T1), 1.8% (T2), and 1.6% (T1ρ). Production phantoms showed inter‐phantom CVs of 0.8% (T1), 1.2% (T2), and 0.9% (T1ρ). NMR and MRI measurements demonstrated strong agreement (CCCs > 0.99 for T1, T2, and T1ρ) with biases < 6%. Inter‐vendor MRI comparisons yielded CCCs > 0.99 for T1 and T2. Data Conclusion The phantom demonstrated excellent stability and reproducibility with temperature correction, supporting its use for quality assurance in longitudinal and multisite qMRI studies. Evidence Level 1. Technical Efficacy Stage 1.