The Experiment Study Of MRI-R₂^* Value For Noninvasive Measurement Of Liver Iron Content In Iron Overload Rabbit

Objective To study the feasibility of the use of R₂* for measurement of liver iron content(LIC) in iron overload rabbit compared with the pathological findings and laboratory data, and explore the relationship between R₂ * and LIC.Materials and methods 32 adult healthy New Zealand rabbits were randomly divided into four groups,control group(n=8),mild iron overload group(n=8),moderate iron overload(n=8) and severe iron overload group(n=8). The three iron overload group received intravenous injection of Iron Sucrose of different oncentrations every one week for four weeks.All experimental animals underwent liver scan with 2D MFGRE sequence by using GE Signa HDx 3.0T MR.The raw data were postprocessed to obtained R₂*Map by functool software on workstation.The R₂* value were measured on R₂*Map with whole liver and small region-of interest(ROI) methods by two observers who had been trained to use the technology proficiently.The liver of experimental animals were sended for LIC assessment immediately after MR scan was finished.Use SPSS 13.0 for Windows statistical software to analyze the data and compare the differences in the R₂* value intra- and inter-viewer.The R₂* value with whole liver and the R₂* value small ROI methods of different dosages were also compared respectively.The relationship between R₂* value with whole liver and LIC and R₂* value with small ROI methods and LIC were explored respectively.Results (1)The LIC increase with the increased in total dose of iron.(2)The inter-observer agreement was good for both whole liver and small ROI methods(P>0.05,ICC=0.99). (3)The R₂* value of small ROI methods were higher than that of whole liver methods in all the groups except the mild iron overload group.(4)The differences of R₂* value between whole liver and small ROI methods for each observer were significant(P<0.05),the R₂* value of whole liver methods were lower than that of small ROI methods.(5)There was a significant exponent curve association both between the R₂* value of whole liver methods and LIC and the R₂* value of small ROI methods and LIC.Curve regression equation of the R₂* value of whole liver methods and LIC were:Y^=EXP(1.418+20.039X),R=0.905 (observer 1);Y^ =EXP(1.416+ 20.117X),R=0.909(observer 2).Curve regression equation of the R₂* value of whole liver methods and LIC were: Y^=EXP(1.950+ 16.200X) , R=0.894 (observer 1) ; Y^ =EXP(1.958+ 16.265X) ,R=0.891(observer 2).But,a weak correlation (R=0.201,0.174, 0.186,0.206,respectively) was displayed in animals with high LIC(above 20.0 mg/g dw).Conclusion (1)R₂* value can accurately predicting LIC with a significant exponent curve association.But,the correlation become weak high LIC(above 20.0 mg/g dw).(2)The whole ROI method can reduce the observer dependence and with a higher reproducibility .