MEDLINE/PubMed Journal Browser Search

Pubmed for Handhelds

PUBMED FOR HANDHELDS

Search MEDLINE/PubMed

Title: [Measurement reproducibility of parameters derived from introvoxel incoherent motion diffusion-weighted MRI imaging of rectal cancer].
Author: Meng YK, Zhang CD, Zhang HM, Ye F, Ouyang H, Zhao XM, Xu K, Zhou CW.
Journal: Zhonghua Zhong Liu Za Zhi; 2017 Dec 23; 39(12):910-915. PubMed ID: 29262507.
Abstract:
Objective: To study the measurement reproducibility of parameters derived from introvoxel incoherent motion (IVIM) diffusion-weighted imaging (DWI)-MRI of rectal cancer between- and within- radiologists. Methods: Clinical data of 34 patients with rectal cancer were prospective analyzed. Conventional MRI sequences, IVIM DWI-MRI with sixteen b values and dynamic contrast enhancement (DCE)-MRI sequences of rectum were acquired by GE 3.0-T MRI imager. The IVIM sequence images with b value=1000 sec/mm(2) were selected to measure the maximum axial section of tumor by a radiologist with 15 year-experiences in gastrointestinal cancer imaging.Two radiologists (radiologist 1 and radiologist 2 with 2 and 10 years of experience in gastrointestinal cancer imaging, respectively) independently draw a freehand region of interest (ROI) that contained the largest available tumor area on the selected section. Monoexponential apparent diffusion coefficient (ADC) and biexponential IVIM metrics maps and IVIM parameters were generated automatically by the software. The repeated measurement was performed at an interval of one week. The average values of each measurement were used for statistical analysis. ADC values and IVIM parameters obtained between- and within- radiologists were analyzed by Wilcoxon signed-rank test. Intraclass correlation coefficients (ICC) and Bland-Altaman plots were used to analyze the parameter reproducibility of two measurements between- and within- radiologists. Results: The first and second measured ADC (×10(-3)mm(2)/s), true diffusivity (D, ×10(-3)mm(2)/s), false diffusivity (D(*,) mm(2)/s) and perfusion fraction (f, %) by radiologist 1 were 0.997, 0.692, 0.043, 34.6 and 0.993, 0.691, 0.038, 32.8, respectively. The first and second measured ADC (×10(-3)mm(2)/s), D (×10(-3)mm(2)/s), D(*) (mm(2)/s), f (%) by radiologist 2 were 0.987, 0.651, 0.046, 32.8 and 0.996, 0.689、0.041, 32.7, respectively. No statistically significant differences were observed in ADC and IVIM parameters obtained between- and within- radiologists (P>0.05). The ADC values and the f values of two times were significantly correlated between- and within- radiologists. The D values were significantly correlated within a radiologist, and the correlation of D(*) values within a radiologist was significantly higher than that between radiologists. The 95% limits of agreement (LoA) of ADC values and f values were smaller than those of D values and D(*) values between- and within- radiologists. The 95% LoA of ADC values was the least, while that of D(*) values varied most.The 95% LoA of f values and D values kept steady within a radiologist, and 95% LoA of f values was slightly smaller than that of D values. The 95% LoA of IVIM parameters (ADC, D, f, D(*) values) within radiologists 2 were better than those within radiologist 1. Conclusions: The reproducibilities of ADC and f values are the best, while the reproducibility of D(*) values is relatively poorer in rectal cancer. Measurement reproducibility of parameters derived from IVIM may be improved by increasing radiologists' experiences in drawing ROI. 目的： 研究不同医师和同一医师之间直肠癌体素内不相干运动参数测量的可重复性。 方法： 回顾性分析34例直肠腺癌患者的临床资料，所有患者均在磁共振扫描仪上行直肠MRI常规序列、双指数体素内不相干运动模型(IVIM)和动态对比增强(DCE)-MRI序列扫描。由1位从事消化道肿瘤影像诊断专业15年的医师，在b值为1 000的IVIM序列图像上选定用于测量的肿瘤最大层面。由2位医师(医师1和医师2，分别从事消化道肿瘤影像诊断专业2年及10年)分别在选定的图像上手动勾画感兴趣区，软件自动生成单指数的表观扩散系数(ADC)图、双指数的IVIM参数图及相关参数。间隔1周后重复测量1次。取每次测量的平均值用于统计学分析。对2位医师之间及同一医师自身前后2次测量数据行Wilcoxon符号秩和检验。利用组内相关系数(ICC)和Bland-Altaman图比较2位医师之间及同一医师自身前后2次测量数据的可重复性。 结果： 医师1第1次测量的ADC、真性扩散系数(D)、假性扩散系数(D(*))和灌注分数(f)值分别为0.997×10(－3)mm(2)/s、0.692×10(－3)mm(2)/s、0.043 mm(2)/s和34.6%，第2次测量的ADC、D、D(*)和f值分别为0.993×10(－3)mm(2)/s、0.691×10(－3)mm(2)/s、0.038 mm(2)/s和32.8%。医师2第1次测量的ADC、D、D(*)和f值分别为0.987×10(－3)mm(2)/s、0.651×10(－3)mm(2)/s、0.046 mm(2)/s和32.8%，第2次测量的ADC、D、D(*)和f值分别为0.996×10(－3)mm(2)/s、0.689×10(－3)mm(2)/s、0.041 mm(2)/s和32.7%。2位医师之间及同一医师自身前后2次测量数据的差异均无统计学意义(均P>0.05)。2位医师之间及同一医师自身前后2次测量的ADC值、f值均显著相关；同一医师自身前后2次测量的D值显著相关；医师2自身前后2次测量的D(*)值的相关性明显高于医师1。无论是2位医师之间，还是在同一医师自身前后2次测量的ADC值和f值的95%一致性界限(LoA)均低于D值和D(*)值，其中ADC值最小，D(*)值波动较大；同一医师自身前后2次的f值、D值的95% LoA均较好，f值略小于D值；医师2自身前后2次测量的ADC、D、f和D(*)值的95% LoA均低于医师1。 结论： 在直肠癌IVIM定量参数测量的可重复性研究中，ADC值、f值的可重复性最好，D(*)值相对较差；测量者经验的增加可以提高IVIM参数测量的可重复性。.

[Abstract] [Full Text] [Related] [New Search]