-
最小二乘法估计和不对称回波迭代水解和脂肪成像(iterative decomposition of water and fat with asymmetry and least squares estimation quantitative fat imaging,IDEAL-IQ)是一种无创的分析组织器官质子密度脂肪分数(proton density fat fraction,PDFF)和铁含量(R2*)的MRI技术,一般基于3.0 T MRI机型实现。目前,部分1.5 T MRI机型也开展了相关工作[1]。研究报道,不同场强的腰椎脂肪定量分析的一致性较好[1-2],而铁定量分析的一致性较差[3]。鉴于MRI机型、系统性能(硬、软件配置)等差异,不同场强MRI测量的PDFF、R2*值缺少扫描参数、ROI勾画、后处理算法等的标准化。临床上,同一受检者通过不同场强MRI测量的PDFF、R2*结果缺少直接比较的依据,受检者面临重复测量的问题。目前,基于IDEAL-IQ序列、不同场强MRI系统间PDFF、R2*测量值的对比研究较少。本研究以不同脂肪含量模型和健康志愿者于3.0 T和1.5 T MRI机型上分别用临床使用的IDEAL-IQ序列扫描,初步探讨2种场强MRI定量分析肝脏PDFF和R2*测量值的异同。
-
20名健康志愿者均未见明确的肝脏病变;模型和健康志愿者的图像质量均符合纳入标准。
-
A1组与B1组对应的5个模型的PDFF测量平均值分别为(20.59±14.39)%和(21.89±14.95)%,差异无统计学意义(Z=−1.550,P=0.121)。除试管1外,其他试管模型的B1测量值均高于A1,且差异有统计学意义(均P<0.05)。试管模型1的A1、B1测量值的差异无统计学意义(P>0.05)。A1组2~5号试管的PDFF测量值更接近试管模型的脂肪含量(表1)。
场强 试管编号 1 2 3 4 5 A1(3.0 T) 0.21±2.67 11.34±2.14 20.36±0.89 30.48±0.77 40.70±0.42 B1(1.5 T) 0.12±0.81 12.13±0.88 22.20±0.66 32.68±1.02 42.59±0.90 Z值 −1.464 −2.153 −4.329 −4.151 −4.979 P值 0.143 0.031 <0.001 <0.001 <0.001 注:1~5号试管模型的脂肪含量分别为0、10%、20%、30%、40%。MRI为磁共振成像;PDFF为质子密度脂肪分数 表 1 2种MRI场强模型PDFF测量值的比较(%)
Table 1. Comparison of models' proton density fat fraction measured values of two magnetic resonance imaging field strengths (%)
-
A2组与B2组对应的5个试管模型的R2*测量平均值分别为(84.86±116.43) Hz和(43.61±54.59) Hz,差异有统计学意义(Z=−3.448,P=0.001),且所有A2测量值均高于B2;试管模型1、3、4的A2、B2测量值的差异均有统计学意义(均P<0.05)(表2)。
场强 试管编号 1 2 3 4 5 A2(3.0 T) 107.76±185.71 90.31±118.00 62.28±77.58 96.40±98.26 67.52±71.61 B2(1.5 T) 39.66±50.30 43.19±47.37 35.83±33.20 43.07±43.23 56.30±86.99 Z值 −1.971 −1.241 −2.029 −2.175 −1.212 P值 0.049 0.215 0.042 0.030 0.226 注:1~5号试管模型的脂肪含量分别为0、10%、20%、30%、40%。MRI为磁共振成像;R2*为铁含量 表 2 2种场强MRI模型R2* 测量值的比较(Hz)
Table 2. Comparison of models' R2*measured values of two magnetic resonance imaging field strengths (Hz)
-
健康志愿者A组肝脏(A3)及皮下脂肪(A4)PDFF测量值与对应的B组(B3、B4)的差异均无统计学意义(均P>0.05);健康志愿者A组肝脏(A5)及皮下脂肪(A6)R2*测量值均高于对应的B组(B5、B6),且差异有统计学意义(均P<0.001)(表3)。健康志愿者PDFF和R2*测量的示意图见图2。
场强 PDFF(%) R2*(Hz) 肝脏 皮下脂肪 肝脏 皮下脂肪 3.0 T 3.33±2.95 81.78±6.33 52.42±12.22 50.88±10.32 1.5 T 4.39±2.80 81.54±5.53 32.73±5.62 39.41±9.11 检验值 Z=−1.867 t=−0.301 Z=−3.920 t=4.372 P值 0.062 0.767 <0.001 <0.001 注:MRI为磁共振成像;PDFF为质子密度脂肪分数;R2*为铁含量 表 3 2种MRI场强健康志愿者肝脏和皮下脂肪PDFF及 R2* 测量值的比较(n=20)
Table 3. Comparison of measured values of PDFF and R2* of liver and subcutaneous fat of volunteers with two MRI field strengths (n=20)
两种场强磁共振IDEAL-IQ序列对质子密度脂肪分数和铁含量定量分析的初步对比研究
A preliminary quantitative comparative study of proton density fat fraction and iron content on two different strength of magnetic resonance with IDEAL-IQ sequence
-
摘要:
目的 比较3.0 T和1.5 T MR最小二乘法估计和不对称回波迭代分解水和脂肪成像(IDEAL-IQ)序列质子密度脂肪分数(PDFF)和铁含量(R2*)测量值的异同。 方法 于2019年12月选取佛山市第一人民医院健康志愿者20名[其中男性13名、女性7名,年龄(23.7±2.5)岁]并用不同脂肪含量的塑料试管模型进行前瞻性研究,分别在3.0 T(A组)和1.5 T(B组)2台 MRI设备上应用IDEAL-IQ序列进行扫描。在自动生成的FatFrac图和R2*图上自动勾画感兴趣区(ROI),分别测量模型、健康志愿者的肝脏和皮下脂肪的PDFF、R2*值。符合正态分布的计量资料采用配对样本t检验(方差齐)和Mann-Whitney U检验(方差不齐)进行比较。 结果 模型A1组和B1组PDFF的测量平均值分别为(20.59±14.39)%和(21.89±14.95)%,差异无统计学意义(Z=−1.550,P=0.121);A1组和B1组R2*的测量平均值分别为(84.86±116.43) Hz和(43.61±54.59) Hz,差异有统计学意义(Z=−3.448,P=0.001)。健康志愿者3.0 T和1.5 T MRI肝脏、皮下脂肪PDFF测量平均值分别为(3.33±2.95) %和(4.39±2.80) %、(81.78±6.33) %和(81.54±5.53) %,差异均无统计学意义(Z=−1.867、t=−0.301,均P>0.05);A组肝脏、皮下脂肪的R2*测量平均值为(52.42±12.22) Hz、(50.88±10.32) Hz,分别高于对应的B组[(32.73±5.62) Hz、(39.41±9.11) Hz],且差异均有统计学意义(Z=−3.920、t=4.372,均P<0.001)。 结论 基于IDEAL-IQ序列的3.0 T和1.5 T MR模型、健康志愿者肝脏和皮下脂肪的PDFF测量值的差异无统计学意义,但3.0 T MRI的PDFF测量值可能比1.5 T MRI更接近临床实际情况;2种场强获得的R2*值的差异有统计学意义,其影响因素需进一步研究。 Abstract:Objective Aimed to compare similarities and differences of the quantitative measurements of liver proton density fat fraction (PDFF) and iron content (R2*) between 3.0 T and 1.5 T MRI with iterative decomposition of water and fat with asymmetry and least squares estimation quantitative fat imaging (IDEAL-IQ) sequences. Methods Twenty healthy volunteers [13 males and 7 females, aged (23.7±2.5) years] and plastic test tube models with different fat contents were scanned with IDEAL-IQ sequence for prospective research on two MRI equipment of 3.0 T (group A) and 1.5 T (group B) in December 2019 in First People's Hospital of Foshan. The FatFrac images and R2* images automatically generated were used to draw the region of interests (ROIs) automatically. The PDFF and R2* values of the models and the liver and subcutaneous fat of the healthy volunteers were measured and compared between groups A and B. Measurement data that conformed to normal distribution were compared using paired sample t test (equal variances assumed) and Mann-Whitney U test (equal variances not assumed). Results The measured mean PDFF values of the model of groups A1 and B1 were (20.59±14.39)% and (21.89±14.95)%, respectively, with no significant difference (Z=−1.550, P=0.121). The measured mean R2* values of the model of groups A1 and B1 were (84.86±116.43) Hz and (43.61±54.59) Hz, respectively, with a statistically significant difference (Z=−3.448, P=0.001). No significant difference was found in the mean PDFF values of liver and subcutaneous fat of healthy volunteers [(3.33±2.95)% vs. (4.39±2.80)%, (81.78±6.33) Hz vs. (81.54±5.53) Hz)(Z=−1.867, t=−0.301; both P>0.05] with 3.0 T and 1.5 T MRI, respectively. The average R2* measured values of group A of the liver and subcutaneous fat of the healthy volunteers were higher than those of group B, with statistical significantly differences [(52.42±12.22) Hz vs. (32.73±5.62) Hz, (50.88±10.32) Hz vs. (39.41±9.11) Hz)(Z=−3.920, t=4.372; both P<0.001]. Conclusions Results showed no significant differences in the PDFF values of models and the liver and subcutaneous fat of the volunteers measured between 3.0 T MR and 1.5 T MR based on IDEAL-IQ sequences, but the PDFF values of 3.0 T MRI may be closer to the clinical actual situation than that of 1.5 T MRI. The R2* values obtained based on 3.0 T MRI field strength were significantly different from that of 1.5 T MRI, and its influencing factors need to be further studied. -
表 1 2种MRI场强模型PDFF测量值的比较(%)
Table 1. Comparison of models' proton density fat fraction measured values of two magnetic resonance imaging field strengths (%)
场强 试管编号 1 2 3 4 5 A1(3.0 T) 0.21±2.67 11.34±2.14 20.36±0.89 30.48±0.77 40.70±0.42 B1(1.5 T) 0.12±0.81 12.13±0.88 22.20±0.66 32.68±1.02 42.59±0.90 Z值 −1.464 −2.153 −4.329 −4.151 −4.979 P值 0.143 0.031 <0.001 <0.001 <0.001 注:1~5号试管模型的脂肪含量分别为0、10%、20%、30%、40%。MRI为磁共振成像;PDFF为质子密度脂肪分数 表 2 2种场强MRI模型R2* 测量值的比较(Hz)
Table 2. Comparison of models' R2*measured values of two magnetic resonance imaging field strengths (Hz)
场强 试管编号 1 2 3 4 5 A2(3.0 T) 107.76±185.71 90.31±118.00 62.28±77.58 96.40±98.26 67.52±71.61 B2(1.5 T) 39.66±50.30 43.19±47.37 35.83±33.20 43.07±43.23 56.30±86.99 Z值 −1.971 −1.241 −2.029 −2.175 −1.212 P值 0.049 0.215 0.042 0.030 0.226 注:1~5号试管模型的脂肪含量分别为0、10%、20%、30%、40%。MRI为磁共振成像;R2*为铁含量 表 3 2种MRI场强健康志愿者肝脏和皮下脂肪PDFF及 R2* 测量值的比较(n=20)
Table 3. Comparison of measured values of PDFF and R2* of liver and subcutaneous fat of volunteers with two MRI field strengths (n=20)
场强 PDFF(%) R2*(Hz) 肝脏 皮下脂肪 肝脏 皮下脂肪 3.0 T 3.33±2.95 81.78±6.33 52.42±12.22 50.88±10.32 1.5 T 4.39±2.80 81.54±5.53 32.73±5.62 39.41±9.11 检验值 Z=−1.867 t=−0.301 Z=−3.920 t=4.372 P值 0.062 0.767 <0.001 <0.001 注:MRI为磁共振成像;PDFF为质子密度脂肪分数;R2*为铁含量 -
[1] 宋宇, 宋清伟, 张楠, 等. 不同场强下磁共振IDEAL-IQ技术对腰椎椎体脂肪含量定量对比分析[J]. 中国临床医学影像杂志, 2019, 30(9): 656−659. DOI: 10.12117/jccmi.2019.09.012.
Song Y, Song QW, Zhang N, et al. Quantitative comparison and analysis of lumbar vertebral fat content by IDEAL-IQ under different field intensities[J]. J Chin Clin Med Imaging, 2019, 30(9): 656−659. DOI: 10.12117/jccmi.2019.09.012.[2] 李承启, 曹代荣, 邢振, 等. MRI场强及设备类型对同反相位成像定量分析健康志愿者腰椎脂肪含量的影响[J]. 中国中西医结合影像学杂志, 2015, 13(2): 122−125. DOI: 10.3969/j.issn.1672-0512.2015.02.002.
Li CQ, Cao DR, Xing Z, et al. Influence of different field strength and different MR scanner with same field on quantitative analysis for lumbar bone marrow fat content with in- and opposed -phase in healthy volunteers[J]. Chin J Integr Tradit West Med, 2015, 13(2): 122−125. DOI: 10.3969/j.issn.1672-0512.2015.02.002.[3] 彭鹏, 杨光鑫, 张灵, 等. 3.0 T和1.5 T MRI定量心肌铁沉积的体外水模对比研究[J]. 实用放射学杂志, 2017, 33(12): 1951−1954. DOI: 10.3969/j.issn.1002-1671.2017.12.037.
Peng P, Yang GX, Zhang L, et al. The study of quantitative myocardial iron deposition in water phantoms at 3.0 T in comparison with 1.5 T MRI[J]. J Pract Radiol, 2017, 33(12): 1951−1954. DOI: 10.3969/j.issn.1002-1671.2017.12.037.[4] Reeder SB, Cruite I, Hamilton G, et al. Quantitative assessment of liver fat with magnetic resonance imaging and spectroscopy[J]. J Magn Reson Imaging, 2011, 34(4): 729−749. DOI: 10.1002/jmri.22775. [5] Pecorelli A, Franceschi P, Braccischi L, et al. MRI appearance of focal lesions in liver iron overload[J/OL]. Diagnostics (Basel), 2022, 12(4): 891[2022-08-06]. https://www.mdpi.com/2075-4418/12/4/891. DOI: 10.3390/diagnostics12040891. [6] França M, Carvalho JG. MR imaging assessment and quantification of liver iron[J/OL]. Abdom Radiol (NY), 2020, 45(11): 3400−3412[2022-08-06]. https://link.springer.com/article/10.1007/s00261-020-02574-8. DOI: 10.1007/s00261-020-02574-8. [7] 安梓铭, 冯琴. 磁共振质子密度脂肪分数——一种具有潜力的非酒精性脂肪性肝炎临床试验替代终点[J]. 临床肝胆病杂志, 2021, 37(6): 1445−1448. DOI: 10.3969/j.issn.1001-5256.2021.06.047.
An ZM, Feng Q. Magnetic resonance imaging-proton density fat fraction: a potential surrogate endpoint for nonalcoholic steatohepatitis clinical trials[J]. J Clin Hepatol, 2021, 37(6): 1445−1448. DOI: 10.3969/j.issn.1001-5256.2021.06.047.[8] Starekova J, Reeder SB. Liver fat quantification: where do we stand?[J/OL]. Abdom Radiol (NY), 2020, 45(11): 3386−3399[2022-08-06]. https://link.springer.com/article/10.1007/s00261-020-02783-1. DOI: 10.1007/s00261-020-02783-1. [9] Stine JG, Munaganuru N, Barnard A, et al. Change in MRI-PDFF and histologic response in patients with nonalcoholic steatohepatitis: a systematic review and meta-analysis[J]. Clin Gastroenterol Hepatol, 2021, 19(11): 2274−2283.e5. DOI: 10.1016/j.cgh.2020.08.061. [10] 洪居陆, 贺小红, 陈婉雯, 等. 对比剂对MRI非对称回波三点法水脂分离测量肝脏脂肪和铁的影响[J]. 放射学实践, 2019, 34(3): 328−331. DOI: 10.13609/j.cnki.1000-0313.2019.03.017.
Hong JL, He XH, Chen WW, et al. A preliminary study on the effect of contrast agents on the quantification of liver fat and iron using MRI IDEAL-IQ sequence[J]. Radiol Pract, 2019, 34(3): 328−331. DOI: 10.13609/j.cnki.1000-0313.2019.03.017.[11] 吴小伟, 陈自谦. 磁共振IDEAL-IQ技术对肝脏脂肪定量的应用研究进展[J]. 功能与分子医学影像学: 电子版, 2017, 6(1): 1159−1162. DOI: 10.3969/j.issn.2095-2252.2017.01.015.
Wu XW, Chen ZQ. Progress in the application of magnetic resonance IDEAL-IQ technique in liver fat quantification[J]. Funct Mol Med Imaging (Electron Ed), 2017, 6(1): 1159−1162. DOI: 10.3969/j.issn.2095-2252.2017.01.015.[12] 胡磊, 查云飞, 林苑, 等. IDEAL-IQ定量评价兔糖尿病模型椎体骨髓脂肪含量的可行性研究[J]. 磁共振成像, 2015, 6(12): 941−946. DOI: 10.3969/j.issn.1674-8034.2015.12.012.
Hu L, Zha YF, Lin Y, et al. The feasibility of IDEAL-IQ quantitative evaluation of vertebral fat fraction content in rabbit models of diabetes mellitus[J]. Chin J Magn Reson Imaging, 2015, 6(12): 941−946. DOI: 10.3969/j.issn.1674-8034.2015.12.012.[13] 朱文雁, 刘朋, 杜婧, 等. 3台不同厂家3.0 T MR设备测量非酒精性脂肪性肝病患者肝脏质子密度脂肪分数的一致性[J]. 中国医学影像学杂志, 2022, 30(2): 111−115, 128. DOI: 10.3969/j.issn.1005-5185.2022.02.004.
Zhu WY, Liu P, Du J, et al. Agreement of proton density fat fraction measurements of 3.0 T MR platforms from three different venders in patients with nonalcoholic fatty liver disease[J]. Chin J Med Imaging, 2022, 30(2): 111−115, 128. DOI: 10.3969/j.issn.1005-5185.2022.02.004.[14] Zhu AT, Hernando D, Johnson KM, et al. Characterizing a short T2* signal component in the liver using ultrashort TE chemical shift-encoded MRI at 1.5 T and 3.0 T[J]. Magn Reson Med, 2019, 82(6): 2032−2045. DOI: 10.1002/mrm.27876. [15] Roberts NT, Hinshaw LA, Colgan TJ, et al. B0 and B1 inhomogeneities in the liver at 1.5 T and 3.0 T[J]. Magn Reson Med, 2021, 85(4): 2212−2220. DOI: 10.1002/mrm.28549. [16] Serai SD, Dillman JR, Trout AT. Proton density fat fraction measurements at 1.5- and 3-T hepatic MR imaging: same-day agreement among readers and across two imager manufacturers[J]. Radiology, 2017, 284(1): 244−254. DOI: 10.1148/radiol.2017161786. [17] Liau J, Shiehmorteza M, Girard OM, et al. Evaluation of MRI fat fraction in the liver and spine pre and post SPIO infusion[J]. Magn Reson Imaging, 2013, 31(6): 1012−1016. DOI: 10.1016/j.mri.2013.01.016. [18] 中国研究型医院学会肝病专业委员会, 中国医师协会脂肪性肝病专家委员会, 中华医学会肝病学分会脂肪肝与酒精性肝病学组, 等. 脂肪性肝病诊疗规范化的专家建议(2019年修订版)[J]. 临床肝胆病杂志, 2019, 35(11): 2426−2430. DOI: 10.3969/j.issn.1001-5256.2019.11.007.
Committee of Hepatology, Chinese Research Hospital Association, Fatty Liver Expert Committee, Chinese Medical Doctor Association, National Workshop on Fatty Liver and Alcoholic Liver Disease, Chinese Society of Hepatology, et al. Expert recommendations on standardized diagnosis and treatment for fatty liver disease (2019 revised edition)[J]. J Clin Hepatol, 2019, 35(11): 2426−2430. DOI: 10.3969/j.issn.1001-5256.2019.11.007.[19] 郭若汨, 唐文杰, 朱叶青, 等. 磁共振IDEAL-IQ序列对肝脏脂肪变性和铁过载的诊断价值[J]. 中山大学学报: 医学科学版, 2015, 36(5): 689−692. DOI: 10.13471/j.cnki.j.sun.yat-sen.univ(med.sci).2015.0111.
Guo RM, Tang WJ, Zhu YQ, et al. Diagnostic value of MRI IDEAL-IQ sequence to hepatic steatosis and hepatic iron-overloaded[J]. J Sun Yat-Sen Univ (Med Sci), 2015, 36(5): 689−692. DOI: 10.13471/j.cnki.j.sun.yat-sen.univ(med.sci).2015.0111.[20] 黄璐, 韩瑞, 夏黎明. 分析1.5 T与3.0 T MR定量评价体外铁浓度模型比较研究[J]. 放射学实践, 2017, 32(10): 1014−1017. DOI: 10.13609/j.cnki.1000-0313.2017.10.004.
Huang L, Han R, Xia LM. Quantitative measurement of iron concentration in vitro model: a comparative study between 1.5 T and 3.0 T MRI[J]. Radiol Pract, 2017, 32(10): 1014−1017. DOI: 10.13609/j.cnki.1000-0313.2017.10.004.[21] Hernando D, Levin YS, Sirlin CB, et al. Quantification of liver iron with MRI: state of the art and remaining challenges[J]. J Magn Reson Imaging, 2014, 40(5): 1003−1021. DOI: 10.1002/jmri.24584. [22] Kirk P, Roughton M, Porter JB, et al. Cardiac T2* magnetic resonance for prediction of cardiac complications in thalassemia major[J]. Circulation, 2009, 120(20): 1961−1968. DOI: 10.1161/CIRCULATIONAHA.109.874487. [23] Karlsson M, Ekstedt M, Dahlström N, et al. Liver R2* is affected by both iron and fat: a dual biopsy-validated study of chronic liver disease[J]. J Magn Reson Imaging, 2019, 50(1): 325−333. DOI: 10.1002/jmri.26601.