ABSTRACT
MRI-R2* has been accepted as a clinical tool for monitoring iron overload in thalassemia patients, especially for monitoring liver iron concentration [LIC]. The most optimal and practical method of analysis however, is still open to further investigations. Our objective was to investigate intra- and intersite observer variability of the pixel-wise method for liver R2* analysis in thalassemia patients using a monoexponential with a constant offset model. We performed 88 liver R2* measurements on 72 thalassemia major patients. A single breath-hold multi-echo gradient-echo sequence was acquired and analyzed at both the reference [REF] and local [LOC] sites. The analysis defined the region of interest in the whole liver parenchyma, excluding the great vessels, and were reported as median values. The R2* values from the REF and LOC were statistically comparable for all comparisons. The intrasite and intersite observer variation were 0.75% [less than 0.9%] and 2.5%, respectively, both of which are comparable to previous reports, but substantially lower than conventional region-based approaches. The low variation of the R2* also yielded excellent variation in the tabulated hepatic iron content. However, caution is required when comparing the results to different implementation methods and appropriate evaluation and validation of methodology for any new scan site is essential before its clinical use