Interscanner Reproducibility of Cardiovascular Magnetic Resonance T2* Measurements of Tissue Iron in Thalassemia

Abstract

Purpose: To assess interscanner reproducibility of tissue iron measurements in patients with thalassemia using gradient echo T2* measurements on two different MRI scanners. Materials and Methods: Twenty-five patients with thalassemia major had liver and myocardial T2* assessment using a Picker Edge 1.5T Scanner and a Siemens Sonata 1.5T scanner, with similar gradient echo sequences. In a subset of 13 patients, two scans on the Siemens scanner were performed to assess interstudy reproducibility. Results: There was a highly significant, linear correlation between T2* values obtained for both the heart (r = 0.95) and the liver (r = 0.99) between scanners. The mean difference, coefficient of variability, and 95% confidence intervals between scanners were 0.8 msec, 9.4% and -5.0 to 6.7 msec for the heart; and 0.9 msec, 7.9% and -2.0 to 3.9 msec for the liver. The interstudy mean difference and coefficient of variability on the Siemens scanner was 0.3 msec and 4.80% (r = 0.99) for the heart, and 0.04 msec and 1.9% (r = 0.99) for the liver. Conclusion: The T2* technique for measuring tissue iron is reproducible between the two manufacturers' scanners. This suggests that the widespread implementation of the technique is possible for clinical assessment of myocardial iron loading in thalassemia. © 2003 Wiley-Liss, Inc.