Logo image
157 - Repeatability of knee cartilage T1ρ and T2 mapping: A multi-site multi-vendor study by QMIC
Abstract   Peer reviewed

157 - Repeatability of knee cartilage T1ρ and T2 mapping: A multi-site multi-vendor study by QMIC

Zhiyuan Zhang, Jennifer Jian, Jeehun Kim, Richard Lartey, Kihwan Kim, Patrick Y. Yeh, Mei Li, Nancy Obuchowski, Carl Winalski, Brian J. Soher, …
Osteoarthritis and cartilage, Vol.34(Supplement), pp.S132-S133
04/2026
DOI: 10.1016/j.joca.2026.01.164

View Online

Abstract

Purpose (the aim of the study): To evaluate the repeatability of T1ρ and T2 imaging, promising biomarkers for cartilage composition in osteoarthritis (OA), using the Musculoskeletal (MSK) Committee of the Quantitative Medical Imaging Coalition (QMIC; formerly the RSNA/QIBA-MSK Committee) recommended protocol across multiple sites and MR platforms. Methods: Twelve sites using MRI platforms from 3 vendors participated in this study (Table 1). Healthy volunteers without diagnosed OA were recruited, and their knees were scanned. T1ρ and T2 images were acquired using the prototype 3D Magnetization-Prepared Angle-Modulated Partitioned k-Space Spoiled Gradient Echo Snapshots (MAPSS) sequence. 3D MAPSS was scanned with an acquisition matrix of 320x160x24 voxels, FOV of 140x140x96 mm3, time of spin lock (TSL) of 0,10,30,70ms with 500Hz spin lock frequency for T1ρ, TE of 0,20,40,60ms for T2. A 3D dual echo steady state (DESS) sequence was acquired for cartilage auto-segmentation (Figure 1) with an acquisition matrix of 384x307x160 voxels, FOV of 140x140x112 mm3, and TR/TE of 16.6/6.2ms. In total, 76 volunteers were scanned and re-scanned with off-table-on-table repositioning between scans at all sites. T1ρ and T2 values were estimated by fitting a mono-exponential decay using nonlinear least-squares. CVs were calculated between scan and rescan, and RMS CVs were computed across different subjects for each compartment for each site. RMS CVs were also calculated to combine different sites or compartments. Results: Excellent scan-rescan repeatability measures were calculated, with an overall scan-rescan RMS CV of 2.9% and 3.4% for T1ρ and T2, respectively, across all compartments and sites. The scan-rescan RMS CV for T1ρ and T2 varied between 1.7%-7.1% across different sites, and RMS CV of each cartilage compartment ranged from 2.5%-4.1% (Table 2). We will continue collecting volunteer data to update these results (target enrolment: 5-10 per site), and investigate factors contributing to performance variability between sites, including auto-segmentation, knee position, and B0/B1 inhomogeneities. Conclusions: This study demonstrated excellent repeatability of knee cartilage T1ρ and T2 imaging using the QMIC MSK committee-recommended protocols in a multi-site multi-vendor setting. These techniques hold strong potential for translation into clinical practice to improve early diagnosis and risk stratification of OA, and to support large-scale clinical trials in accelerating the development and efficacy testing of novel OA drugs.

Details

Metrics

1 Record Views
Logo image