Feasibility of in vivo, multicontrast-weighted MR imaging of carotid atherosclerosis for multicenter studies

Purpose To test the image quality (ImQ) and interscan coverage of MRI for measuring carotid atherosclerosis across multiple centers. Materials and Methods Thirty-nine subjects from five clinical sites (site 1: n = 11; site 2: n = 16; site 3: n = 2; site 4: n = 3; site 5: n = 7) were imaged on GE 1.5T scanners using a standardized carotid imaging protocol with five weightings (T1, proton density (PD), T2, time-of-flight (TOF), and contrast-enhanced (CE) T1). MR technologists from the five sites received comprehensive protocol training. A maximum coverage of 24 mm (12 slices) was designed for each of four scans (baseline and at four, eight, and 13 weeks). The adequacy of coverage was calculated as the percentage of arteries with at least six slices matched across all four scans. ImQ was evaluated using an established five-point scale for each image. ImQ ≥ 3 was considered acceptable for image analysis. Results Across five sites, the mean ImQ was 3.4–4.2 for T1W, 3.6–4.4 for CE-T1W, 3.4–4.2 for PDW, 3.3–4.2 for T2W, and 3.4–4.0 for TOF. The mean ImQ per site was 3.5–4.2. All sites generated at least six-slice coverage (mean = 8.0–9.1) for all index carotid arteries. Conclusion The ImQ and coverage values were comparable among clinical sites using a standardized carotid imaging protocol. With comprehensive protocol training, carotid MRI is technically feasible for use in multicenter studies. J. Magn. Reson. Imaging 2005;21:809–817. © 2005 Wiley-Liss, Inc.