【作者】 黄璐；

【导师】 夏黎明；

【作者基本信息】 华中科技大学 ， 影像医学与核医学， 2013， 博士

【摘要】 目的：探讨MRI对体外不同铁浓度模型定量测量的可行性，并比较1.5T和3.0T磁共振对体外铁浓度模型定量测定的准确性。材料和方法：制作2组不同铁浓度梯度的体外模型，一组铁浓度梯度范围为0-5.0mg/ml。另一组铁浓度梯度范围为0-1.00mg/ml。于同一天分别行1.5T和3.0TMR扫描，扫描序列包括短轴位FSE/T1WI、FRFSE/T2WI、T2map和T2*map，扫描范围从溶液顶部至底部。两位研究者分别独自测量体外铁浓度模型的MR图像，测量指标包括T1WI和T2WI信噪比、T2、R2、T2*和R2*值。测量指标的一致性采用组内相关系数分析。铁浓度与T1WI信噪比、T2WI信噪比、T2、R2、T2*和R2*值的关系使用Pearson相关分析。结果：两位研究者对体外铁浓度模型MR图像测量指标T1WI信噪比、T2WI信噪比、T2、R2、T2*和R2*值一致性评价的组内相关系数均大于0.900（P＜0.001）。T2WI信噪比、T2和T2*值与铁浓度在1.5T和3.0T MR上均呈非线性相关。在1.5T MR上，当铁浓度＜1.5mg/ml时，T1WISNR、R2与铁浓度呈显著性正相关（P＜0.05）；当铁浓度≥1.5mg/ml时，T1WI信噪比、R2与铁浓度呈显著性负相关（P＜0.001）。R2*则在铁浓度＜2.5mg/ml时，与铁浓度呈显著正性线性相关（P＜0.001）；铁浓度≥2.5mg/ml时，R2*与铁浓度呈明显负性线性相关（P=0.008）。而在3.0T MR上，当铁浓度＜1.5mg/ml时，T1WI信噪比与铁浓度无显著性线性关系（P=0.301）；当铁浓度≥1.5mg/ml时，T1WI信噪比与铁浓度呈显著性负相关（P=0.001）。当铁浓度＜0.90mg/ml时，R2和R2*值与铁浓度呈正相关（P＜0.05）；当铁浓度≥0.90mg/ml, R2和R2*值与铁浓度呈显著负相关（P＜0.05）。结论：MRI对体外不同铁浓度模型的定量测定准确性高、可重复性好和稳定性强，且1.5T MR在测定铁浓度上优于3.0T。目的：探讨磁共振T2*map成像定量测量兔心肌铁超负荷模型心肌铁负荷的可行性。材料和方法：日本长耳兔11只，10只作为心肌铁超负荷实验组，一只作为正常对照组。大腿深部肌肉注射含铁量为50mg/ml右旋糖酐铁剂（注射剂量为50mg/kg），每周注射一次，共注射12周。11只兔子在注射右旋糖酐铁前全部进行MRI检查，实验组兔注射右旋糖酐铁一周后复查MRI检查。MR扫描序列包括心脏短轴位的双反转恢复-FSE（DIR）、三反转恢复-FSE（TIR）和T2*map序列以及肝脏和双肾轴位FSPGR/T1WI、SSFSE/T2WI和T2*map序列。定量分析兔的心肌竖脊肌信号强度比、肝脏竖脊肌信号强度比、双肾竖脊肌信号强度比及心脏、肝脏、双肾、竖脊肌的T2*和R2*值。动物模型处死后取心脏、肝脏和双肾，固定后行离体标本MRI检查，之后取部分心肝肾组织进行铁沉积病理切片。所有兔子在行MRI检查前一天，抽取3ml静脉血查血清铁。测量参数之间的相关性分析使用Pearson相关。结果：心肌T2*、R2*值、DIR心肌/竖脊肌信号强度比与注入铁含量呈显著线性相关（P＜0.05），而TIR心肌/竖脊肌信号强度比与注入铁含量无显著性线性相关（P＞0.05）。心肌T2*、R2*值与肝T2*、R2*值相关性分析不具有显著性（P＞0.05）。肝T2*和R2*与注入铁含量无显著线性相关（P＞0.05）。肾T2*、R2*值与注入铁含量具有明显线性相关（P＜0.001），而T1WI和T2WI肾/竖脊肌信号强度比与注入铁含量无显著线性相关性（P＞0.05）。心肝肾MR测量结果与血清铁无显著相关（P＞0.05）。离体心、肾T2*和R2*值与注入铁含量成线性相关（P＜0.05），离体肝T2*和R2*值与注入铁含量无线性相关（P＞0.05）。心肌T2*的下降趋势与心肌病理学结果相一致。结论： MR-T2*map序列可以准确的测量心肌铁超负荷兔的心肌铁负荷，可以为临床铁超负荷患者提供一种集诊断和治疗后监测铁负荷的简便、安全、无创、全面的方法。目的：应用MR探讨兔心肌铁超负荷模型的心功能参数与心肌铁负荷程度之间的关系，并评价心功能参数与肝铁浓度之间的关系。材料和方法：日本长耳兔11只，10只作为心肌铁超负荷实验组，一只作为正常对照组。对实验组兔按照每只50mg/kg剂量，大腿深部肌肉注射含铁量为50mg/ml右旋糖酐铁，每周注射一次，共注射12周。11只兔子在注射右旋糖酐铁前全部行心脏和肝脏MRI检查，实验组兔开始注射右旋糖酐铁一周后复查MRI。扫描序列包括左心室连续短轴位cine、双反转恢复-FSE（DIR）、三反转恢复-FSE（TIR）和T2*map序列和横轴位肝脏FSPGR/T1WI、SSFSE/T2WI和T2*map序列。定量分析兔的心功能参数和心脏与竖脊肌T2*、R2*值、心肌/竖脊肌信号强度比以及肝与竖脊肌T2*、R2*、T1WI和T2WI肝/竖脊肌信号强度比。动物模型处死后取心脏和肝脏，固定标本后取部分心肝组织进行铁沉积病理切片。所有兔子在行MRI检查前一天，抽取3ml静脉血查血清铁。心功能参数与注入铁含量、血清铁、T2*map测量值的相关关系采用Pearson相关分析。结果：左心室心功能参数中LVEDD、LVEDV、LVESV和LVEF与注入铁含量具有显著性的线性相关（P＜0.05），右心室心功能参数中RVEDV、RVESV和RVEF与注入铁含量具有线性相关性（P＜0.05）。左心室心功能参数中LVSV和LVCO与血清铁具有中等线性相关性，而右心室心功能参数中RVEDV和RVESV与血清铁具有显著线性相关关系（P＜0.05）。LVESD、LVEDV、LVESV和LVEF与心肌T2*值呈显著线性相关关系，LVEDD、LVESD、LVEDV、LVESV和LVSV与心肌R2*值具有线性相关性（P＜0.05）。右心室心功能参数中，RVEDV和RVESV与心肌T2*呈中等负相关性（P＜0.05），而与心肌R2*呈正性线性相关（P＜0.05）。心功能参数与肝MR测量结果无显著性差异（P＞0.05）。结论：MRI是一种简单、方便、全面和有效的评估体内铁负荷的方法，且能同时评价心功能，LVESV较LVEF能更好的反映心肌铁超负荷时的心功能。目的：应用磁共振T2*map成像定量测定铁超负荷患者的心肌铁负荷情况，并分析T2*值和其心功能参数之间的关系。材料和方法：收集2010年10月至2012年10月在我院血液科住院的铁超负荷患者8例（铁超负荷组）。正常对照组均为健康志愿者，共22例。铁超负荷组和正常对照组均在1.5T磁共振扫描仪上检查，MR扫描序列包括左心室连续短轴位cine和T2*map序列以及肝胰脾T2*map序列。两位研究者分别独自测量所有的MR图像。测量参数包括左右心室心功能参数、左心室16个心肌节段T2*和R2*值、肝胰脾及竖脊肌T2*和R2*值。两位研究者测量的一致性使用组内相关系数评价。铁超负荷组和正常对照组的比较采用独立样本t检验，心、肝T2*值和心功能参数的关系采用Pearson相关分析。结果：两位研究者对MR T2*map的心肌、肝胰脾T2*和R2*值以及心功能参数测量的一致性均大于0.900（P＜0.001）。有4例铁超负荷患者存在心肌铁超负荷，心肌T2*＜20ms心肌节段共30段。8例铁超负荷患者中7例有肝铁超负荷和5例脾铁超负荷患者。铁超负荷组患者肝和脾的T2*值与正常对照组比较有显著差异（P＜0.05）。铁超负荷组胰腺和竖脊肌无铁超负荷。左心室心肌壁厚度和右心室心腔内径、双心室的每搏输出量及射血分数都在正常参考范围内。左心室的收缩末期内径较正常对照组明显扩大（P=0.005），而舒张末期内径（LVEDD）、收缩末期心腔容积（LVESV）和舒张末期心腔容积（LVEDV）虽然大于正常对照组，但是不具有统计学差异（P＞0.05）。左心室心腔的缩短率在正常值范围内，相比与正常对照组还是明显减低（P=0.004）。心肝T2*值与心功能参数的相关性分析均不具有显著性（P＞0.05），且心肝T2*值也不具有显著的相关性（P＞0.05）。结论：磁共振T2*成像是一种诊断和监测心肌铁超负荷快速、可靠、敏感、无创的方法。LVESD和LVFS有可能代替LVEF成为预测心肌铁超负荷患者心功能的新指标。更多还原

【Abstract】 Objective: To investigate the feasibility of quantitative assessment of iron concentration invitro model using MRI and compare the accuracy between1.5T and3.0T MR.Materials and Methods: Two sets of iron concentration in vitro models, ranging from0mg/ml to5.0mg/ml and from0mg/ml to1.00mg/ml, respectively. The two sets modelswere performed both on1.5T and3.0T MR in the same day. The scanning protocolincluded FSE/T1WI, FRFSE/T2WI, T2map and T2*map on axial section, and scanningrange were from top to bottle of the solution. Two radiologists independently evaluated theMR images of the models.T1WI signal to noise ratio (SNR), T2WI SNR, T2, R2, T2*andR2*values were measured. Intraclass correlation coefficient (ICC) was used to evaluate theagreements of the parameters and Pearson correlation analysis was used to calculate thecorrelation between iron concentration and MR measurement indexes.Results: The ICCs of T1WI SNR, T2WI SNR, T2, R2, T2*and R2*values evaluated bytwo radiologists were all more than0.900(P＜0.001). The relation between ironconcentration and T2WI SNR, T2and T2*values demonstrated function relation. Wheniron concentration was less than1.5mg/ml, T1WI SNR and R2correlated positively withiron concentration in1.5T MR(P＜0.05); while iron concentration was more than1.5mg/ml,T1WI SNR and R2correlated negatively with it(P＜0.05). When iron concentration wasless than2.5mg/ml, R2*had a positive correlation with it（P＜0.001）; while more than 2.5mg/ml, R2*had a negative correlation with it（P=0.008）.When iron concentration wasless than1.5mg/ml, T1WI SNR had no significant correlation with it(P=0.301) in3.0T MR;while iron concentration was more than1.5mg/ml, T1WI SNR had a significantly positivecorrelation with it(P=0.001). When iron concentration was less than0.900mg/ml, R2andR2*were positively correlated with it(P＜0.05）; while it was more than0.900mg/ml, R2and R2*were negatively correlated with it(P＜0.05).Conclusion: MRI is able to quantitatively evaluate iron concentration in vitro modelaccurately, reproducibly and stably, and it is better on1.5T MR. Objective: To discuss whether it is feasible for magnetic resonance T2*map to evaluatemyocardial iron concentration of myocardial iron overload rabbit model.Materials and Methods: Eleven rabbits were divided into two groups, myocardial ironoverload group (n=10) and control group (n=1). Iron dextrin (concentration of50mg/ml，dose of50mg/kg) was injected in muscles of thigh once a week, totally12weeks. Beforeiron dextrin ejection, all the11rabbits were performed MRI. The myocardial iron overloadgroup underwent MRI1week after iron dextrin injection. MRI scanning protocol includedcardiac short axial Double inversion recovery-FSE (DIR), Triple inversion recovery-FSE(TIR) and T2*map, as well as liver and kidneys FSPGR/T1WI, SSFSE/T2WI and T2*mapon axial section. Myocardial to erector spinae signal intensity ratio, liver to erector spinaesignal intensity ratio, kidney to erector spinae signal intensity ratio, as well as T2*and R2*of myocardial, liver, kidney and erector spinae were evaluated. After animal models killed,heart, liver and kidneys were fixed in order to undergo MRI and excise for pathologicalslices. All rabbits were withdrawn blood to test serum iron, one day before performed MRI. Pearson correlation analysis was used to evaluate relations among measurement indexes.Results: Myocardial T2*, R2*values and DIR myocardial to erector spinae signal intensityratio had linear correlation with injecting iron content(P＜0.05), while TIR myocardial toerector spinae signal intensity ratio had no significant correlation with injecting ironcontent. Myocardial T2*and R2*values had no significant correlation with liver T2*andR2*values(P＜0.05). Liver T2*and R2*values were not significantly correlated withinjecting iron content(P＞0.05). Kidney T2*, R2*had significant correlation with injectingiron content(P＜0.001),while T1WI and T2WI kidney to erector spinae signal intensityratio had no significant correlation(P＞0.05). MR measurement indexes were notsignificantly correlated with serum iron(P＞0.05). T2*and R2*of heart and kidney in vitrohad linear correlation with injecting iron content(P＜0.05), while T2*and R2*of liver hadno significant correlation with injecting iron content(P＞0.05). The change of myocardialT2*value was consistent with myocardial pathological resultsConclusion: MR T2*map is able to evaluate myocardial iron concentration of myocardialoverload rabbit model accurately, and to provide a simple, safe, noninvasive and completemodality for iron load diagnosis and follow-up supervision. Objective: To determine the relationship between cardiac function parameters andmyocardial iron load of myocardial iron overload rabbit model, and the relationshipbetween cardiac function and hepatic iron concentration.Materials and Methods: Eleven rabbits were divided into two groups, myocardial ironoverload group (n=10) and control group (n=1). Iron dextrin (concentration of50mg/ml,dose of50mg/kg) was injected in muscles of thigh once a week, totally12weeks. Before iron dextrin ejection, all the11rabbits were performed MRI. The myocardial iron overloadgroup underwent MRI1week after iron dextrin injection. MR scanning protocol includedcardiac short axial cine, Double inversion recovery-FSE (DIR), Triple inversionrecovery-FSE (TIR) and T2*map, as well as liver FSPGR/T1WI,SSFSE/T2WI andT2*map on axial section. Cardiac function parameters, myocardial to erector spinae signalintensity ratio and liver to erector spinae signal intensity ratio, as well as T2*and R2*ofmyocardial, liver and erector spinae were evaluated. After animal models killed, heart andliver were fixed to excise for pathological slices. All rabbits were withdrawn blood to testserum iron, one day before performed MR. Pearson correlation analysis was used toevaluate relations among cardiac function indexes, injecting iron content, serum iron andT2*map measurement indexes.Results: LVEDD, LVEDV, LVESV and LVEF had significant linear correlation withinjecting iron content(P＜0.05) and LVSV as well as LVCO has moderate correlation withserum iron(P＜0.05). RVEDV, RVESV and RVEF were correlated with injecting ironcontent significantly(P＜0.05), as well as RVEDV and RVESV were correlated with serumiron significantly(P＜0.05). LVESD, LVEDV, LVESV and LVEF had a significant linearcorrelation with myocardial T2*, as well as LVEDD, LVESD, LVEDV, LVESV and LVSVhad a linear correlation with myocardial R2*(P＜0.05). RVEDV and RVESV weremoderate correlated with myocardial T2*and R2*(P＜0.05).Cardiac function parametershad no significant correlation with liver T2*and R2*(P＞0.05).Conclusion: MRI is a simple, convenient, complete and effective modality to evaluate ironload in vivo, and can evaluate cardiac function simultaneously. LVESV may be better thanLVEF to predict cardiac function of myocardial iron overload. Objective: To evaluate myocardial load of patients with iron overload by MR T2*map, andrelationship with their cardiac function.Materials and Methods: Eight patients with iron overload (iron overload group) wereenrolled in this study from October2010through October2012. The control group werehealthy volunteers (n=22). Iron overload group and control group both underwent theexamination in1.5T MR scanner. The MR protocol included short axial cine, T2*map ofleft ventricle and axial T2*map of abdomen. Two radiologists independently measured allthe MR images. Cardiac function parameters, T2*and R2*values of left ventricle16segments, as well as T2*and R2*values of liver, pancreas, spleen and erector spinae weremeasured. The agreements of two radiologists were evaluate by intraclass correlationcoefficient (ICC). The differences between iron overload group and control group werecalculated by independent sample t-test, as well as relationship between T2*values ofmyocardial and liver and cardiac function parameters were calculated by Pearsoncorrelation analysis.Results: The ICCs of T2*and R2*of myocardial, liver, pancreas, spleen and erectorspinae and cardiac function parameters measured by two radiologists were all more than0.900(P＜0.001). Four iron overload patient with myocardial iron overload, and T2*valuesof30myocardial segments were less than20ms. Of eight patients with iron overload, sevencases had hepatic iron overload and five cases had spleen iron overload. T2*va lues of liverand spleen between iron overload group and control group had significant differences (P＜0.05). Pancreas and erector spinae had no iron overload in iron overload group. Leftventricle wall thickness, dimensions of right ventricle, stroke volume of chambers andejection fraction were all within the reference ranges. Left ventricle end-systolic dimensionwas significantly larger than the control group one (P=0.005). End-diastolic dimension, end-systolic volume and end-diastolic volume were larger than the control group ones, butthe differences had no significance (P＞0.05). Left ventricle shortening fraction was inreference range, though it was significantly lower than the control group one (P=0.004). Ithad no significant correlations between myocardial T2*and hepatic T2*(P＞0.05), andalso no significant correlations between cardiac function parameters and T2*values ofmyocardium and liver(P＞0.05).Conclusion: MR T2*is able to diagnose and supervise myocardial iron load promptly,reliably, sensitively and noninvasively. LVESD ad LVFS may substitute LVEF to predictcardiac function of patients with myocardial overload.更多还原