节点文献

磁共振功能成像在慢性肾病中的临床应用研究

The Clinical Study of Functional Magnetic Resonance Imaging in Chronic Kidney Disease

【作者】 李清海

【导师】 严福华; 丁小强; 林江; 徐鹏举;

【作者基本信息】 复旦大学 , 影像医学与核医学, 2009, 博士

【摘要】 第一部分前瞻性采集校正技术在肾脏弥散加权成像中的临床应用研究目的:探讨采用导航触发前瞻性采集校正技术的弥散加权成像(PACE-DWI)在肾脏ADC值测量中的可重复性,且就图像信噪比(SNR)及对比噪声比(CNR)和常规摒气DWI进行比较。方法:对15例健康志愿者进行了重复DWI检查(两次检查间隔时间为5—42天),每次DWI扫描均采用摒气单次激发的自旋回波平面回波(SS-SE-EPI)序列和采用PACE技术的SE-EPI序列,选用5个不同的弥散梯度因子(b值)为100、300、500、800及1000 s/mm2。PACE-DWI在肾实质ADC值测量中的重复性研究,采用配对t检验、Bland-Altman plot及一致性相关分析;两种成像序列间SNR及CNR的比较采用Mann-Whitney U检验;不同b值组问SNR及CNR的比较采用Kruskal-Wallis H检验;摒气DWI和PACE-DWI所测得ADC值的比较,采用Wilcoxon配对符号秩检验;不同b值组间ADC值的比较采用Friedman检验。结果:各b值条件下,采用PACE-DWI重复测得的双肾实质ADC值间均无统计学差异(P>0.05)。且除b=100 s/mm2条件外,PACE-DWI重复测得的肾实质ADC值的平均差为-4.4×10-5 mm2/s~6.3×10-5mm2/s,一致性限(平均差±1.96 SD)为±28.1×10-5mm2/s~±36.4×10-5mm2/s,重复测得的肾实质ADC值间具有很高的一致性(ρc=0.892,P=0.000)。相同b值条件下,PACE-DWI图像的SNR及CNR明显高于摒气序列,均存在统计学差异(P<0.05);但无论摒气DWI还是PACE-DWI,随着b值的增高,DWI图像的SNR及CNR均呈下降趋势,且在各b值组间存在显著的统计学差异(P<0.01)。另外,相同b值条件下,PACE-DWI所测得的ADC值均高于摒气序列,且除b=100 s/mm2条件下的右肾实质ADC值外,两者间均存在统计学差异(P<0.05);但无论摒气DWI还是PACE-DWI,肾实质ADC值随b值的增加均呈下降趋势,且各b值组间的ADC值均存在显著的统计学差异(P<0.01)。结论:PACE-DWI在肾脏ADC值的测量中具有较好的可重复性,且SNR和CNR均显著高于常规摒气DWI,适用于临床研究。但是,PACE-DWI与摒气DWI所测得的ADC值间存在差异。第二部分弥散加权成像在慢性肾病功能评估中的临床应用研究目的:探讨弥散加权成像(DWI)在慢性肾病(CKD)临床分期及分肾功能评估中的价值及肾体积对肾功能的影响,比较采用导航触发前瞻性采集校正技术的弥散加权成像(PACE-DWI)在分肾功能评估中是否存在优势。方法:对83例CKD病例及12例健康志愿者进行了DWI检查,分别采用摒气SS-SE-EPI序列及采用PACE技术的SE-EPI序列,弥散梯度因子(b值)分别为300 s/mm2及800 s/mm2。其中54例同时进行了同位素肾图检查,采用Gates法计算肾小球滤过率(GFR)。采用单因素方差分析来比较双肾ADC均值在不同肾功能临床分期间,及分肾ADC值、分肾体积及分肾ADC值体积乘积在不同分肾功能组间是否存在差异;采用控制年龄因素的偏相关分析及受试者工作曲线(ROC)来分析分肾ADC值、肾脏体积及分肾ADC值体积乘积在分肾功能评估中的价值,及在预测分肾功能降低及轻度降低中的诊断价值。结果:所测得的双肾ADC均值在各临床分期组间均存在显著的统计学差异(P<0.01),随CKD肾功能临床分期的增加呈下降趋势,且在b=800 s/mm2时,第1、2期与第3-5期间均存在统计学差异(P<0.05)。分肾ADC值与分肾GFR间存在轻度正相关性,以PACE-DWI在b=800 s/mm2时,两者的正相关性最高(r=0.471,P=0.000)。PACE-DWI所测得的分肾ADC值随分肾功能下降呈减小的趋势,在三组间存在显著的统计学差异(P<0.01),且以b=800 s/mm2条件下,在预测分肾功能轻度降低中的诊断效能最大(AUC=0.780,P=0.000)。另外,分肾体积与分肾GFR间存在低度正相关性(r=0.470,P=0.000)。分肾ADC值体积乘积与分肾GFR之间为中度正相关性(r=0.521-0.550,P=0.000),且在不同分肾功能组间均存在显著的统计学差异(P<0.01),在预测分肾功能轻度降低中的诊断效能进一步提高(AUC=0.824-0.841,P=0.000)。结论:DWI在CKD临床分期和分肾功能评估中具有一定的价值,尤其是采用PACE技术和高b值的DWI在分肾功能的评估中具有优势,且在诊断分肾功能轻度降低中存在一定的价值。另外,体积作为肾功能的影响因素不容忽视,同时结合ADC值和肾体积测量可以提高对分肾功能评估的准确性。第三部分弥散加权成像评价慢性肾病病理损害的临床-病理对照研究目的:探讨弥散加权成像(DWI)在慢性肾病(CKD)病理损害评估中的临床应用价值,及采用导航触发前瞻性采集校正技术的弥散加权成像(PACE-DWI)是否存在优势。方法:83例拟行右肾穿刺的CKD病例及12例健康志愿者进行了DWI检查,分别采用摒气SS-SE-EPI序列和采用PACE技术的SE-EPI序列进行,选取b值为300s/mm2及800 s/mm2。由两位病理科医生共同对肾穿刺标本进行病理损害评分,做为评价慢性肾病病理损害的标准。最后共71例CKD病例及12例健康志愿者纳入分析。采用Pearson相关分析、单因素方差分析及ROC曲线分析ADC值在评估CKD病理损害程度及在预测中重度病理损害程度中的价值,并探讨是否PACE-DWI存在优势;采用两个独立样本的t检验和单因素方差分析来比较不同CKD病理类型及不同IgA肾病Lee氏分级间的ADC值是否存在差异;采用多重线性回归模型初步分析影响ADC值的独立因素。结果:随CKD病理损害程度的加重,右肾ADC值呈现下降的趋势,与病理积分间存在显著的负相关性,以PACE-DWI在b=800 s/mm2时,两者的负相关性最大(r=-0.632,P=0.000),且在预测中重度病理损害中的诊断效能最大(AUC=0.806,P=0.000)。PACE-DWI在b=800 s/mm2时,原发性肾小球肾炎四种病理类型间的ADC值存在统计学差异(F=4.164,P=0.016),LSD两两分析的结果表明,硬化性肾小球肾炎的ADC值显著低于膜性肾病及局灶性节段性肾炎(P<0.05)。IgA肾病不同Lee氏分级组间ADC值亦存在显著的统计学差异(P<0.01),LSD两两比较的结果表明,4级病例的ADC值显著低于2级及3级病例组(P<0.01)。多重线性回归分析的结果表明,只有病理损害积分与ADC值间存在显著的线性回归关系(P<0.05),而右肾GFR及年龄与ADC值间均不存在线性回归关系。结论:DWI能够在一定程度上反映CKD的病理损害程度,且采用PACE技术和高b值的DWI在病理损害评估中具有优势,有望成为一种无创性评价CKD病理损害程度、指导治疗和随访的手段。但DWI在CKD病理类型诊断方面的价值受限,有待于进一步研究。第四部分3.OT血氧水平依赖性磁共振成像在慢性肾病中的初步临床应用研究目的:探讨3.0T血氧水平依赖性磁共振成像(BOLD-MRI)最大TE时间的选择是否对肾脏R2*值的测量存在影响,以及初步探讨其在CKD肾脏氧合状态评估中的价值。方法:在GE 3.0T Signa HDx磁共振扫描仪上,对13例健康志愿者及61例拟行肾穿刺检查的CKD患者进行了BOLD-MRI,采用多梯度重聚回波序列,分别采用6个回波链及12个回波链,TE时间分别为2.8-22.4 ms及2.8-45.9 ms。最后13例健康志愿者及44例CKD病例纳入分析,其中共39例CKD病例进行了肾实质病理损害评分,20例同时进行了同位素GFR检查。由两位放射科医师共同对图像进行分析,通过直方图分析的方法,在R2*图像上分别测得采用不同最大TE时间条件下双肾皮髓质的R2*值,并采用采用Wilcoxon配对符号秩检验比较两者间是否存在差异。对于CKD患者,只测量6个回波链条件下双肾皮髓质的R2*值。正常肾脏皮髓质R2*值的比较,采用Wilcoxon配对符号秩检验;采用6个回波链和12个回波链的3.0T BOLD-MRI在测量双肾皮髓质R2*间是否存在差异的比较,采用配对t检验、Bland-Altman plot及一致性相关分析;正常对照组与CKD两组间双肾皮质、髓质R2*均值的比较,采用Mann-Whitney U检验;不同肾功能临床分期间的双肾皮质、髓质R2*均值的比较,采用单因素方差分析;分肾R2*与分肾GFR的相关性分析及与病理损害积分问的相关性分析,采用spearman相关性分析。结果:在R2*彩图上,正常肾脏的皮髓质分界相对清楚,呈现从皮质到髓质色彩逐渐由蓝色到绿色、黄色、红色的过渡变化。肾脏髓质的R2*值明显高于皮质,两者间存在显著的统计学差异(P<0.01)。6个回波链(最大TE:22.4 msec)和12个回波链(最大TE:45.9 msec)条件下,所测得的左右肾皮髓质的R2*值间均不存在显著性差异(P>0.05)。并且,不同回波链条件下重复测得的双肾皮质及髓质R2*的平均差为-0.3 sec-1,一致性限(平均差±1.96 SD)为-1.9~1.4 sec-1,重复测得的双肾皮质及髓质R2*值具有很高的一致性(pc=0.983,P=0.000)。CKD组的双肾皮质、髓质R2*均值明显高于正常对照组,两者间存在显著的统计学差异(P<0.01)。随肾功能临床分期的增加,双肾髓质R2*均值呈逐渐增大的趋势,且各组间存在统计学差异(P<0.05)。LSD两两比较的结果表明,正常对照组与2-4期间均存在统计学差异(P<0.05)。另外,分肾髓质R2*值与分肾功能GFR间存在显著的负相关性(r=-0.415,P=0.008),与病理损害积分间存在显著的正相关(r=0.450,P=-0.004)。结论:在3.0T BOLD-MRI检查中,采用6个回波链(最大TE 22.4 msec)就能够满足肾脏R2*值的测量,且能够反映CKD中存在的缺氧状态,有望为肾脏氧合状态的评估提供一种无创性的手段。

【Abstract】 PartⅠThe clinical application of diffusion-weighted imaging with navigator-triggered prospective acquisition correction in kidney:initial experienceObjective:To evaluate the Reproducibility of renal apparent diffusion coefficient (ADC) measured by diffusion-weighted imaging with navigator-triggered prospective acquisition correction(PACE-DWI),and to compare the signal to noise ratio(SNR) and contrast to noise ratio(CNR) of DWI images with conventional DWI with breathholding.Materials and methods:Fifteen healthy volunteers accepted a repetitive DWI examination within a interval of 5 to 42 days.Two DWI sequences were performed on every subject at b values of 100,300,500,800 and 1000 s/mm2,including a single-shot spin-echo echo-planar sequence within a single breath-hold and a spin-echo echo-planar sequence with PACE.The repetitive measured ADCs by PACE-DWI were compared using the matched-pairs T test,the Bland-Altman plot and Concordance correlation coefficient.The differences of SNR and CNR between DWI with breathholding and PACE-DWI was analyzed using the Mann-Whitney U test.The Kruskal-Wallis H test was used to compare the difference of SNR and CNR among DWI with different b values.The influence of different DWI sequences on ADC measurement was analyzed by the Wilcoxon matched-pairs signed-ranks test. The Friedman test was used to analyze the influence of b value on ADC measurement.Results:There was no significant statistical difference between the ADCs of renal parenchyma repetitive measured by PACE-DWI(P>0.05).The absolute differences of repetitive measured ADCs of renal parenchyma and the limits of agreement by PACE-DWI were -4.4×10-5 mm2/s to -6.3×10-5 mm2/s and±28.1×10-5 to±36.4×10-5 mm2/s except when b=100 s/mm2.In addition,there was a high concordance between repetitive measured ADCs of renal parenchyma by PACE-DWI(ρc=0.892,P=0.000). Higher SNR and CNR were obtained by PACE-DWI than those by conventional DWI group at the same b Value,and the differences were statistically significant(P<0.05). The SNR and CNR of DWI images decreased as the b Value increased in both conventional DWI and PACE-DWI,and there was significant statistical difference among the same group with five b Values(P<0.01).However,higher ADCs were obtained by PACE-DWI than those by conventional DWI at the same b Value,and the differences were statistically significant except the ADCs of right renal parenchyma with b value of 100 s/mm2(P<0.05).The measured ADCs of renal parenchyma decreased as the b Value increased in both conventional DWI and PACE-DWI,and there was significant statistical differences among the same group with five b Values (P<0.01).Conclusion:PACE-DWI has a good reproducibility in the measurement of ADCs of renal parenchyma and will benefit clinical researchs due to its higher SNR and CNR than conventional DWI with breathholding.However,there is still some differences between ADCs measured by PACE-DWI and conventional DWI.PartⅡClinical study of diffusion-weighted imaging in renal function assessment of chronic kidney diseaseObjective:To investigate the value of DWI in the evaluation of clinical staging and split renal function of(chronic kidney disease) CKD,and to discuss whether or not DWI with navigator-triggered prospective acquisition correction(PACE-DWI) is superior to conventional DWI with breathholding.In addition,the influence of renal volume on renal function is discussed.Materials and Methods:Eighty-three CKD patients and 12 healthy volunteers underwent DWI examination using both a single-shot spin-echo echo-planar sequence within a single breath-hold and a spin-echo echo-planar sequence with PACE at b values of 300 and 800 s/mm2.Among them,54 patients received radionuclide renography examination,and the glomerular filtration rate(GFR) is calculated based on the Gates Method.One-way analysis of variance(ANOVA) was used to compare the average apparent diffusion coefficient(ADC) of both kidney among different clinical stages,the ADCs of split kidney,the volume of split kidney and the split volume-corrected renal ADCs among the three groups with different split renal function.The partial correlation test with the control of age and receiver operationg characteristics analysis(ROC) were used to assess the performance of split renal ADCs,split renal volume and split volume-corrected renal ADCs in the evaluation of split renal function and predicting the reduction or mild reduction of split renal function.Results:The average ADCs of both kidneys decreased as the clinical stage of CKD progressed.There was a statistically significant difference among the five clinical stages(P<0.01),and between stage one to two and stage three to five at b value of 800 s/mm2(P<0.05).There was a weak positive correlation between the split ADCs and split GFR.The best correlation was observed for PACE-DWI with b value 800 s/mm2(r=0.471,P=0.000).The split renal ADCs measured by PACE-DWI were significant different among these three groups with different GFR(P<0.01) and was also a significant predictor of mild split renal function reduction for PACE-DWI with b value 800s/mm2,with an area under the curve of 0.780(P=0.000).There was a weak positive correlation between split renal volume and renal GFR(r=0.470, P=0.000).However,there was a moderate positive correlation between the split volume-corrected renal ADC and renal GFR(r=0.521 to 0.550,P=0.000).The split volume-corrected renal ADCs were significant different among these three groups with different GFR(P<0.01) and the diagnosis potency predicting mild split renal function reduction was improved too(AUC=0.824 to 0.841,P=0.000).Conclusion:Renal ADC measured by DWI can be used in the assessment of clinical stage and split renal function of CKD.PACE-DWI with b value 800s/mm2 is the best in assessing split renal function,and has potential value in predicting mild reduction of split renal function.In addition,renal volume can not be ignored in the evaluation of renal function,and the ADC measurement combined with renal volume measurement would be a clinical feasible modality in the evalution of renal function.PartⅢComparative study of diffusion-weighted imaging in assessing pathological lesion of chronic kidney disease with pathologyObjective:To investigate the value of apparent diffusion coefficient(ADC) measured by DWI in assessing the pathological injures of chronic renal disease(CKD),and to discuss whether or not DWI with navigator-triggered prospective acquisition correction(PACE-DWI) is superior to conventional DWI with breathholding.Materials and methods:Eighty-three patients with CKD intended for renal biopsy and 12 healthy volunteers underwent the DWI examination using both a single-shot spin-echo-planar sequence within a single breath-hold and a spin-echo echo-planar sequence with PACE at b values of 300 and 800 s/mm2.Renal biopsy results were reviewed together by two experienced renal pathologists to score the pathological injures of CKD.Finally,71 patients and 12 healthy volunteers were enrolled into further analysis of this study.The bivariate correlation test of Pearson,One-way analysis of variance(ANOVA) and receiver operationg characteristics analysis were used to discuss the value of ADCs in assessing the degree of renal pathological lesions,as well as predicting the moderate and severe degree of renal pathological lesions.Besides,two independent-samples T test and ANOVA were used to compare the ACDs among different pathological types of CKD and different Lee’s grades of IgA nephropathy.In addition,multiple linear regression analysis was used to find out the independent influential factors of ADC.Results:The ADCs of right renal parenchyma decreased as the aggravation of renal pathological lesions.There was a significant negative correlation between the ADCs of right renal parenchyma and its pathological scores.The largest negative correlation coefficient was obtained for PACE-DWI with b value 800 s/mm2(r=-0.632,P=0.000). The renal ADC was also a significant predictor of moderate and severe renal pathological lesions for PACE-DWI with b value 800 s/mm2,with an area under the curve of 0.806(P=0.000).The ADCs of right renal parenchyma among four different primary glomerulonephritis types were significant different for PACE-DWI with b value 800 s/mm2(F=4.164,P=0.016).The ADCs of sclerotic glomerulonephritis was significant lower than those of membranous nephropathy and focal segment nephritis (P<0.05).There was a statistically significant difference among the ADCs of different Lee’s grades(P<0.01).The ADC of fourth grade was significant lower than those of second grade and third grade(P<0.01).The multiple linear regression analysis showed that the significant linear regression was only observed between pathological scores and ADCs(P<0.05),not for age and GFR.Conclusion:DWI can be used to assess the degree of pathological injures in CKD. Moreover,PACE-DWI with high b value is superior to conventional DWI in the evaluation of pathological lesions of CKD.It provides a non-invasive modality for the evaluation of pathological lesions,guiding therapy and follow up of CKD.However, the value of DWI in differentiating pathological types of CKD is limited and need further researches. PartⅣClinical study of 3.0T blood oxygenation level-dependent MRI in chronic kidney disease:preliminary experienceObjective:To investigate the advisable maximum TE for 3.0T blood oxygenation level-dependent MRI(BOLD-MRI) in renal R2* mearsurement and to further discuss its value in assessing renal oxygenation of chronic renal disease(CKD).Materials and methods:Sixty-one patients with CKD intended for renal biopsy and 13 healthy volunteers accepted the BOLD MRI examination on a GE 3.0T Signa HDx scanner.Multi gradient recalled echo with the echo train length of 6 and 12 was used for BOLD-MRI.The maximum TE were 22.4 and 45.9 ms respectively.Finally, thirteen healthy volunteers and 44 patients were enrolled into further analysis.Among them,the biopsy samples of thirty-nine patients were reviewed together by two experienced renal pathologists to score the pathological injures of CKD.In addition, twenty patients received radionuclide renography examination,and the glomerular filtration rate(GFR) is calculated based on the Gates Method.All images were reviewed by two experienced residents together.The R2* values of renal cortex and medulla with the different maximum TE were measured on the R2* images using the analysis of histogram.The Wilcoxon matched-pairs signed-ranks test was used to compare the difference of R2* between renal cortex and medulla in normal kidneys. The repetitive measured R2* by 3.0T BOLD-DWI with different maximum TE were compared using the matched-pairs T test,the Bland-Altman plot and Concordance correlation coefficient.R2* was measured only with the echo train length of 6 in patients with CKD.The Mann-Whitney U test was used to compare the R2* difference between normal group and CKD group.One-way analysis of variance(ANOVA) was used to compare the difference of R2* among different stages of renal function.The bivariate correlation test of Spearman was used to analyze the correlation of R2* and scores of pathological lesions of right kidney as well as the split R2* and GFR.Results:On the R2* images,renal cortex could be clearly differentiated from medulla with a color change from blue to green,yellow till red.The R2* of renal medulla was remarkably higher than that of renal cortex.There was a significant difference between the R2* of renal cortex and medulla(P<0.01)while there was no significant difference between the R2* of renal cortex with two maximum TE(P>0.05).Besides, the absolute differences and the limits of agreement of repetitive measured R2* by 3.0T BOLD-MRI with two maximum TE were -0.3 sec-1 and -1.9~1.4 sec-1,and there was a high concordance between repetitive measured R2*(ρc=0.983,P=0.000).The average R2* of renal cortex of both sides in the normal was significantly higher than that in CKD,as well as for the average R2* of renal medulla(P<0.01).Moreover,the average R2* of renal medulla increased with the clinical stage of CKD,and there was a statistical difference among different stages of CKD.The average R2* of renal medulla of stage 1 is significant higher than that of stage 2 to 4(P<0.05).A significant negative correlation between R2* of renal medulla and GFR(r=-0.415, P=0.008) and a significant positive correlation between R2* of renal medulla and scores of pathological lesions were found(r=0.450,P=0.004).Conclusion:3.0T BOLD-MRI with the echo train length of 6(the maximum TE of 22.4 ms) is adequate for the renal R2* measurementand can detect hypoxia of CKD, which have potential advantages in the evaluation of renal oxygenation as a noninvasive modality.

  • 【网络出版投稿人】 复旦大学
  • 【网络出版年期】2010年 04期
  • 【分类号】R692;R445.2
  • 【被引频次】2
  • 【下载频次】407
  • 攻读期成果
节点文献中: 

本文链接的文献网络图示:

本文的引文网络