直肠癌磁共振体素内不相干运动成像高b值优化的初步研究

Optimization of High B Values for Intravoxel Incoherent Motion Imaging of Rectal Cancer:A Pilot Study

目的对直肠癌高b值(>200 s/mm^2)进行初步的优化,观察高b值变化对直肠癌体素内不相干运动成像(IVIM)参数的影响。方法前瞻性研究22例直肠腺癌患者。由一名从事消化道肿瘤影像诊断专业10年的医生,选取全部16个b值(A组:0、10、20、30、40、60、80、100、150、200、400、800、1000、1200、1500、2000)后,在b值=800的IVIM序列图像上选取肿瘤实性成分最大的层面用于测量。保持感兴趣区(ROI)不变,分别再次重新选取14个b值(B组:0、10、20、30、40、60、80、100、150、200、400、800、1000、1200)和12个b值(C组:0、10、20、30、40、60、80、100、150、200、400、800),软件分别自动生成相应的IVIM参数图及参数,同样的步骤间隔一周后由同一医生分别再重复测量一次。取每次测量的平均值用于分析。采用Bland-Altman图对同一医生间隔1周前后两次测量数据的一致性及可重复性进行分析。对A、B、C三组IVIM参数的平均值进行单因素方差分析(ANOVA),同时进行Bonferroni组间比较,以P<0.05认为差异有统计学意义。结果三组间比较,D*值95%一致性界限在A、B组和A、C组组间波动相对较大。ADC、D值95%一致性界限组间波动均不明显。ADC、D、D*值的单因素方差分析结果 P值均<0.05,差异具有统计学意义。f值A、B、C三组单因素方差分析P值>0.05,差异没有统计学意义。Bonferroni组间比较,D、D*、f值在A组和B组间均没有统计学差异。随着高b值数目的减少,ADC、D、D*值及标准误增加,f值变化不明显。结论直肠癌IVIM成像高b值变化对参数测量的可重复性影响不明显。b值>1500的选取价值有待进一步研究。

Objective To optimize the high b values（ 200 s/mm^2） for intravoxel incoherent motion imaging of rectal cancer and to observe the effect of high b values variation on IVIM parameters. Methods Twenty-two patients with rectal adenocarcinoma were included in the prospective study. Three groups（ A group with all 16 b values： 0,10,20,30,40,60,80,100,150,200,400,800,1000,1200,1500,2000,B group with 14 b values： 0,10,20,30,40,60,80,100,150,200,400,800,1000,1200 and C group with 12 b values： 0,10,20,30,40,60,80,100,150,200,400,800） were selected respectively for measurement by a radiologist（ with 10 years of experience in gastrointestinal cancer imaging）. The largest slice of the solid components of the tumor was selected for measurement in the IVIM metrics map（ b = 800）. Monoexponential ADC and biexponential IVIM metrics maps and IVIM parameters were generated automatically by the software. The same measurements were repeated one week apart. The average values of each measurement were used for statistical analysis. The consistency and reproducibility of the two measurements were analyzed by Bland-Altman plots. One-way analysis of variance（ ANOVA） and post-hoc test were performed on the mean values of IVIM parameters in groups A,B,and C,with a significance level of P〈0. 05. Results The D＊ values variations of 95% limits of agreement（ Lo A） between groups A,B and A,C were relatively large. The 95% Lo A of ADC、D values between groups were not obvious. The p values of ANOVA results in A DC、D、D＊values were less than 0. 05,the differences were statistically significant. The p value of ANOVA results in f values was greater than 0. 05,the difference was not statistically significant. The p values of Bonferroni post-hoc test in D、D＊、f values did not have statistically significant differences in group A and B. With the number of high b values decreasing,the values of ADC、D、D＊values and standard error increased,while those of f values did not change significantly. Conclusion In our study,the reproducibility of the IVIM parameters caused by high b value variation was not significant. The value of selected b 1500 needs to be further studied.