Liver iron quantification by 3 tesla MRI: calibration on a rabbit model.

PURPOSE: To determine the feasibility of liver iron quantification by 3 Tesla (T) MRI using a novel iron overload rabbit model. MATERIALS AND METHODS: Forty-two rabbits underwent iron dextran loading from 1 to 15 weeks. MRI signal intensity ratio (SIR) was measured using a gradient-echo sequence, and R2(1/T2) measured using an eight-echo spin-echo sequence at 3T. Ex vivo hepatic pathology was obtained for all rabbits studied. Postmortem assessments of liver iron concentration (LIC) were conducted in an atomic absorption spectrophotometer. MRI measures were fitted against LIC using linear regression for 30 of the iron-loaded rabbits. The remaining 12 iron-loaded rabbits were used to test the prediction accuracy of the derived models. RESULTS: LIC was linearly correlated to both liver-to-muscle SIR (r = -0.845) and R2 (r = 0.965) in a range achieved in this study (LIC < 10 mg/g dry tissue) at 3T. By regression, the linear equations were determined as: Y1 = 10.581-5.924X1 (Y1 : LIC, X1 :SIR); Y2 = -1.273+0.103X2 (Y2 :LIC, X2 :R2). In the 12 test rabbits, the predicted LICs using the derived equations agreed well with the results obtained using the spectrophotometer. CONCLUSION: Both SIR and R2 are highly correlated with LIC in a novel rabbit model. MRI quantification of liver iron overload is feasible at 3T.