纤维束示踪成像技术的定量测量:可重复性研究

Quantitative Measurement of Corticospinal Tracts by MR Fiber Tractography:A Reproducibility Study

目的：利用白质纤维束示踪成像技术对神经纤维条目数这个定量参数进行可重复性研究，探讨该指标运用于临床的可靠性。方法：对21例健康志愿者行头颅MR—DTI检查，应用日本东京大学Volume One1．64和diffusion Tensor Visualizer（dTV）软件进行白质纤维束的三维重建，将由软件计算获得的皮质脊髓束（CST）纤维束条目数作为定量指标，分别由同一评定者于3次不同时间及3位不同评定者对同一资料进行重复测量，利用相关系数分析同一评定者3次之间及不同评定者之间的数据，评估其可重复性。而后由同一评定者测量双侧CST的各向异性分数（FA），从而探讨双侧CST条目数及FA值之间有无显著性差异。结果：同一评定者不同3次数据之间的相关系数（r值）范围在0．737～0．939之间，3位不同评定者所得数据之间的r值范围在0．062～0．727之间；由此可以得出同一评定者之间的相关系数高于3位不同评定者之间的相关系数，并且同一评定者测量的双侧CST条目数及FA值无显著性差异（P值均〉0．05）。结论：纤维束示踪成像定量的可重复性不佳，同一评定者的可重复性要高于不同评定者之间。

Purpose： To study the reproducibility of the quantitative measurement of number of white matter fiber bundles with fiber tractography, and to discuss the clinical reliability of the measurement. Methods： 21 healthy volunteers received magnetic resonance diffusion tensor imaging （DTI） with GE 1.5T nuclear magnetic resunance system. The corticospinal tract was used for this study with Volume - One1.64 and diffusion Tensor Visualizer（dTV） software from Tokyo University for three - dimensional reconstruction of white matter fiber bundles. The software calculates the number of corticospinal tract （CST） as quantitative indicator. The measurement was repeated by a rater three times at different time and by three different raters at the same time. The measurement results were analyzed and the correlation coefficients were evaluated to represent reproducibility. After that, the same rater measures fractional anisotropy （FA） values of both sides of CST, and to study the significant difference of number and FA values of both sides of CST. Results： The range of the correlation coefficients （r value） by the same rater at three different time is from 0. 737 to 0. 939, and the conterpart is from 0. 062 to 0. 727, it can be concluded that the correlation coefficient by the same rater is higher than that by three different raters. There is no significant difference about the number and the FA values by the same rater. Conclusion： The reproducibility of the quantitative measurements derived from tractograpy was not good, which by the same rater is better than by different raters.